OXXus.net: Web Hosting » 2007

Archive for October, 2007

im trying to use an old ibook hard drive for a linux installation qtparted is telling me the partition changes

October 31, 2007 at 11:17 pm · Filed under Tomcat

yeah well there are no devices created which I can mount

LOL I had training on Linux LVM a week ago

_2intense – http://pastebin.ca/667307

pending analysis, 1 sec
which devs are you trying to mount?
I mean the dev names

they are called /dev/var/www and /dev/var/toplevel

you have to DECOMPRESS the kernel headers…..
OK, let me check

might be vgchange

_2intense – can you give me a hint how to do that?

hi
i have problem
i am installing Etch on my PC

nils don’t change the vgs if you only want to mount

and he never wanted me to tell him what IDE driver my disk wnats

follow this example

it’s vgchange -ay

bu today ? i do not know what to select

mount /dev/VolGroup00/LogVol00 /var/www or what have you (use the right values from lvdisplay)
drew 1 sec mate I’m multitasking

yeah well I know that, I just didn’t know how to make that device appear. But now I know.

does anybody know ?

drew you downloaded the headers with aptget didn’t you?

SKOM, are you done with the install?

intense yes – apt-get install linux-headers-(version)

thanks

thanks anyways

cd /usr/src/linux-headers-2.6.18-5-686
what’s in there

arch block crypto drivers fs include init ipc kernel lib Makefile mm Module.symvers net scripts security sound usr

hm, it seems that there is no way, that a client can list all exported directories of a nf-server, is this correct?

Yegods, I just looked at aptitude for the first time in I don’t know. We’re not seriously suggesting newbies use that? It’s extemely scary.

no i’m not

Anybody here recommend Totem as a music player?

it appeared during instalation
i have WD Caviar 84AA

Spiritus55, no, not rally.

I would either 1.) not worry about it (it didn’t ask me, or if it did, I don’t know how it will affect you) or 2.) ask someone else.
Recommend one to me then?

or is there a way to do it?

it sounds all right, are you compiling as root?
huh you are

Spiritus55, QuodLibet

can it decode m4a’s?

_2intense – yes

Spiritus55, if you have the right codec installed, everything should be able to play m4as

I apparently don’t have the right codec, then.

Spiritus55, the codec does the decoding

_2intense – i am logged in via openssh, dunno if that makes any difference

and now time to build a new kernel on a system which I can only access via ssh, aka russian roulette

no it’s all right, are you sure that module is not available in a debianized format? It seems like you’re missing some libs somewhere

What codec would I need to decode MP3’s?

Spiritus55, possibly libavcodec0d

worst case scenario: install gstreamer0.10-plugins-really-bad

can i install the 32bit java plugin on my amd64 system with dpkg and set it up to run with nspluginwrapper?

I tried that, and it said It wasn’t there.

to make the zaptel module you have to use make linux26 instead of just make

Couldn’t find package gstreamer0.10-plugins-really-bad

Spiritus55, or libmpeg4ip-0: if not there install them

it’s on debian-multimedia
mmm let’s find the factoid ….
!debian-multimedia

from memory, debian-multimedia is a repository of unofficial Debian packages, not in Debian for various legal reasons, maintained by Christian Marillat. For instructions to add APT sources, ask me about debian-multimedia.etch.

Couldn’t find package libmpeg4ip-0

!tell Spiritus55 about debian-multimedia.etch

_2intense – yeah i tried that – “make: *** No rule to make target `linux26′. Stop.”

hello

i’m just unsure how to actually get the plugin binary on my box

let’s recap. configure and make clean work right?

i am getting an error aclocal not found

zomg. enough

_2intense – configure works – neither make, make clean nor make linux26 work – see pastebin link i posted above for output of make clean

alright i got talked into it. i’m going to d/l testing with jigdo and try it out

what version of GCC do you have? you need 3.3 along 4.0 for the module to work

dunno, whatever apt-get gave me…

query mate, will ya
aptitude and look at the version

coldsteal, what are you doing with automake and autoconf packages?

yoi need to install (since you have aptit open.. gcc 3.3)

“gcc -v”: gcc version 4.1.2 20061115 (prerelease) (Debian 4.1.1-21)

I think we’ve solved the prob

so i need an older gcc…

from zaptel doc: ‘make CC=gcc-3.3 linux26′ if you have gcc-4.0 as default, since some modules use depreciated c extentions.
and of course install gcc 3.3

Spiritus55, http://wiki.debian.org/MultimediaCodecs?highlight=%28codec%29

i have a problem i have two interface ( subnet ) 192.168.1.0 (eth0) and 192.168.2.0 (eth1) .. i have ACCEPT FORWARD between eth0 and eth1 but when i want to ping from 192.168.1.2 to 192.168.2.2 it failed …
nobody have idea ?
it’s a gateway

how do I get a repos’ gpg key if there’s no data on the website?

is ip forwarding enabled on /proc?

yes, I have Internet on both sides
yes and*

hueeee
i had tried installation again

from which machine are you trying to ping?

and that problem with ide driver is here again

how do i find a file from command line

fas3r, echo “1″ /proc/sys/net/ipv4/ip_forward

find / -name thename

it’s enable

thx

does exist some universal driver?

is there any way to reset the permissions on system files and directories quickly to their defaults?

from 192.168.2.2
to 192.168.1.2

you aren’t filtering ICMP are you?

nop

fas3r, routing table is properly set?

it’s the problem i have no firewall, i have just masquerade

check the def gw settings on 2.2

for connect to internet
and echo “1″ in ip_forward

check the settings with sysctl -a | grep ipv4 or icmp

the def gw in host 2.2 should be 2.1 or whatever eth1 is

i think i solved it

Dexter_F, have you looked in the repo for a ‘keyring’ package?

my music keeps droping out and in how i fix it?

hmm, no. will do.

it’s good i have look

have you checked the default gw on host 2.2?
traceroute?

yep it’s good
and tracert stop at the gateway ..

which is 2.1 right?
check the other chains on iptables on the gw, reset the counters and see where the connections stop

yes
the gateway for 192.168.1.0 is 192.168.1.1
and for 192.168.2.0 it’s 192.168.2.1

it’s probably the output chain defaulting to drop

the two interface on my gateway
no i have look …

ok, followed instructions on website, said install multimedia-keyring

i’m trying to find the .jigdo file for amd64/testing but the url on the debian site just hangs

did that, now on aptitude update the multimedia repos are ignored

Dexter_F, that’s odd, i found once i installed that keyring, i don’t get the message about GPG key unknown, but i still get a warning about every package i install

aptitude update

i can’t find the url for amd64/testing to enter into jigdo
the link on the debian jigdo page won’t open so i can get the url

karsten, i did that, i still get the warnings, the one about ‘unverified source’ or something like that

free up the other chains on the gw as a test

Do you have keys for _all_ your package sources?

allow everything then close down one chain at a time

i have just one chain …

karsten, i don’t know, i just added the debian-multimedia repo to sources.list, and installed the keyring from their repo

input output forward

and it’s for masquerade other computer

can i just hover over the link and copy it?

karsten, that’s not a big issue for me, but Dexter_F says when he did aptitude update he doesn’t get any of the repo anymore

yep but policy are at ACCEPT ..

wanted to add a debian-multimedia mirror

fas ok now it’s a bit clearer

wait, keyring from *their* repo?

so you get to the internet from the .1.x subnet and .2.x but NOT from 1.x to 2.x

Dexter_F, yes, i think it’s called debian-multimedia-keyring or something like that

yeah
problem solved

how.. I mean, I can’t even update with the repo
i got that, thought it came from etch itself

yes it’s that

Dexter_F, no, it is in the repo, so you where able to see the packages at that point

_2intense – can i change the defalut compiler? it just doesnt seam to work

you are MASQing between subnets and you get to 1.x from 2.x with a nat address
Drew have you installed a 3.x compiler?

Dexter_F, look in aptitude search for something like libdvdcss, that is not in etch repos, so if you can see that, then you have the repo working ok

_2intense – yes i did apt-get on gcc-3.3 and it’s dependancies

make CC=gcc-3.3 linux26

well, on aptitude upgrade nothing happens, 0 Byte. thought a few libs were at least substituted. il check something…

Well, boys, I’m off to church. Bbl.

when i ping from 1.x or 2.x to 1.x or 2.x you think i use public adress ?

_2intense – yeah, it scrolls a couple of screens and then: “make: *** No rule to make target `linux26′. Stop.”

sorry english and me …

try “make CC=gcc-3.3″

ah. works.
Bob knows why.

yes, I think so, I has the same problems with a network of a customer

make[2]: Entering directory `/usr/src/linux-headers-2.6.18-5-686′
make[2]: Makefile: No such file or directory
make[2]: *** No rule to make target `Makefile’. Stop.
make[2]: Leaving directory `/usr/src/linux-headers-2.6.18-5-686′
make[1]: *** [modules] Error 2

check with tcpdump

hmm does grub start counting kernels with 0 or 1…

sorry, i guess that was too much

how do i turn iceweasel into firefox

anyway, that doesnt work either

IW is firefox only w/o the graphics

ok just i need to make NAT redirection from eth1 to eth2 and vice versa, it’s that ?

TurboBee, in what way do you want to change it?

or SNAT / DNAT ?

yes, that is basically it

basically i want it to say firefox
and use firefox icon
etc

between local network you need to forward

oki thank _2intense ^^ i go to test

between LAN and inet you need to nat

Drew__, apt-get install linux-tree && cd /usr/src/ && tar -xvzf linux-2.6tab

TurboBee, you would need to get the firefox from the official site, and install manually i guess, it’s not in the debian repos

you’re very wemlcome fas3r
mbroeker thanks for the help

TurboBee, but you should probably remove iceweasel first, i would expect them to clash

hehe

I’m trying to help everyone but I only have 1 brain… but it’s fun

just use both halves of your brain symaltaneously

mbroeker – E: Couldn’t find package linux-tree

er simoltaneously
er

Drew__, ok take the linux-source then: linux-tree is specific to lenny

forget it
i can’t spell

I can’t, the other half is currently in use by another process that I can’t kill -9 without going to jail

hey heycan i use an etch package in lenny ?

hehe

but i haven’t forward tables in -t nat ?

don’t masq your individual subnets, masq the default gw

hum ..

Drew__, which version? 2.6.x.y? what is x what is y

otherwise you would have to do some prerouting checks and the likes

Hey, can I use etch packages in lenny?
Like, just one set of packages. (OpenVZ Set)

mbroeker – i just downloaded linux-source-2.6.18

Drew__, ok. and uname -r says something like 2.6.18

linuxtk11 and UptimeFactor-Jus you probably can, i have, but sometimes i’ve had dependency nightmares with mixing packages like that

if your uname -a reports 2.6.18-something you’re ok

uname -r: 2.6.18-5-686

oops sorry mbroeker he’s all yours

Drew__, ok thats looks fine

sorry i don’t understand .. i go to google .. I don’t want to abuse

Like, its OpenvZ (just needs its kernel) thats the only dependency, do yout hink I’d have much of a prob?

Drew__, now cd /usr/src/ and untar the fedora linux source tar.bz2

use normal forwards, not masq, between your two subnets and only masquerade your default gw towards the internet

ln -s linux-source-2.6.18 linux

UptimeFactor-Jus, i don’t know what OpenVZ is, but i guess the only way to find out is to try

Drew__, say when you are done and ready for the next step…

stdin: not in gzip format ??

Drew__, tar -xvjf

if it’s a bz2 is -xvbf mate

i’ll have to write myself a cheatsheet for that sometime

-b has nothing to do with bzip2

Drew__, all done?

I beg your pardon?

@link

if it’s a bz2 is -xvbf mate

the -b option to tar has nothing at all to do with bzip2

ok all done

hi everybody

Drew__, and you have made the symlink ?

yes

you’re right, I meant j LOL

Drew__, if so, cd linux and make menuconfig

heh

if that works, you are fine

somebody have a socks install on his debian ?

it’s midnight in my timezone SerajewelKS, bear w/me LOL

i get a menu

!tell greenlynx about poll

in blue, Drew__ `?

jep

yes, it works

ok sorry
i don’t know it

does tor count as socks?

Drew__, try a fresh “zcat /proc/config.gz

that’s ok, just ask your question

Drew__, does it work ?

well my tor won’t work

you might also try #debian-fr if you speak french

i can’t open the port

_2intense does it provide a socks interface? I think it does

mbroeker – no

excuse for my english i’m french ^^

Yes kevin, tor does

ok Drew__ look in /boot/config*

this ? http://pastebin.com/m612cc1f1

Drew__, is there a config-2.6.18-something?

SerajewelKS there nobody on debian-fr ^^

yes

you’re right, as you can see yourself you’re MASQing everything FROM .1.x OR 2.x, no matter the destination

i confirm what greenlynx say :p

cp /boot/config-2.6.18-xyz /usr/src/linux/.config

hello, I sometimes have lags when I start applications (e.g. terminal) with gnome , how can I diagnose what the problem is ?

ok thank i come back i go to test

lol thx fas3r ^^

done

de rien :p

Drew__, cd /usr/src/linux/ && make modules_prepare

tiens que utilizer dedicated ip forwarding entre .2.x and 1.x et nat pour l’internet

Drew__, after that step, you will be able to build your kernel module outside the tree

does someone know a webbased user management except webmin ? something php based ?

hello folks.. if i got a x86_64 processor arch, should i be downloading the AMD64 version of debian?

good _2intense ..

depends.

*** No rule to make target `modules_prepare’. Stop.

depends on what?

i think it’s same for me ..

I’ve been to canada LOL

lol oki ^^

what will you use the system for and how many ram you have.

if somebody know a good proxy software it’s ok for me

err how much
what do you want to proxy?

squid?

squid
:p

not just a http one

Drew__, make prepare then?

it is not a problem to run 32 bit applications on 64 bit systems as i understand it, in chmod’ed env

i want to tunnel all my network

squid does have plugins for other protocols

2GB ram btw

Drew__, anyway, try to build your kernel module right now outside this tree

*** No rule to make target `include/config/auto.conf’, needed by `include/config/kernel.release’. Stop.

for a tunnel you should use openssh

tunnel whats that

yeah well it’s possible but why would you want 64 bit if you don’t have applications benefiting from it.
it is still a large bit of hassle to deal with the chmoded stuff

sorry

Drew__, ok cd /usr/src/linux && make oldconfig

aren’t most(tm) applications ported to 64 bit yet?

i don’t know the english word

mbroeker – that worked

ok try to compile your kernel module

tunneling ??

yes, most… if you don’t use proprietary software you might as well use 64 bit

64 bit in user space is useless, we sparc folks avoid it since linux was ported LOL

Drew__, if it works, you are fine

but it doesn’t help either so stick with 32 bit if you have less than 3GB RAM… I even use PAE on this system to avoid 64 bit hassle

well, only prop. software i use will be the nvidia drivers (sorry, but the open drivers just aren’t good enough.. )

Drew__, if not, then you need to build and install this kernel from source and reboot with this new kernel

not flash, acrobat reader or stuff like that?

if you want nvidia drivers stick to a 32bit userspace

well I think there are 64 bit drivers now but I don’t know

not really.. well i guess the occasional flash vid will appear on my browser..

google earth is not in any deb repos, right?

there’s no real advantage to 64bit nowadays (in user space, in kernel space there actually is)

mbroeker, ok ill try that, thanks for the help

i will just follow your advice and stick to 32bit

no 64 bit flash plugin available I think…

Dexter_F, i think an installer is in the non-free, same as flash, but last time i tried, it was broken, so i installed manually

just seems a shame to get a 64 bit cpu and not being able to fully utilize it

non-free…. hmm.. forgot that… need to touch up sources.list some more

oh well, it’s just 32 bits… you can still use all the cores and stuff

do you have more than 3GB?

should my etch work with ipv6 by default, or do i need to change something? i’m wondering here why test sites don’t work, if it’s just settings, or my ISP

okay then.. next question: will i have to reinstall my current system (i386) to cope with the hardware upgrade?
no, i will have 2GB of ram

686 no?

well depends… 2.4.20 was the first kernel which was fully stable for my system.

what do you want with 64bit dists then?

intel quad core if you must know.. yeah i686 ..

there’s a -k7
erm k8

i wasn’t aware that 3GB was a minimum req.

We in the Sparc camp had mixed 32/64 releases (both of linux and solaris, and apps such as oracle) for years, it didn’t matter all that much
32 bits can’t address 3 gb of ram, that’s why it’s quoted.

it isn’t. but the only advantage of 64bit over 32 is that you can adress 4GB and more. you don’ gain the slightest bit of speed.

i see

not true… depends on the app

example?

you don’t want to go down the PAE road…

most apps aren’t optimized for 64b, on every opsys

databases
i wasn’t going to

i guess i will just keep my current system then.. saves a lot of reinstalling anyways. .

we’re talking enterprise grade dbs, not your amateur site with mysql here anyway

I sense a flamewar

nah we’re all friends

still the same premise

well, I see how a db profits from directly adressing 4GB+, but that’s because of the address space, which still comes down to: if you want 4gig+ go 64.

I had a 333-Mhz 64 bit cpu

that wasn’t what i meant

well to take php mysql web hosting as example, there is PAE support in it but it isn’t stable or enabled…

ok, how else?

Agreed completely, but do you really need 4gb dbs these days other than in data centres?

and I don’t know of other means to access more than ~3GB in a single process

just my point

IIRC PAE also allows up to 64GB

yeah but it sucks.

PAE is *ugly*

the amount of data moved per op

physical address extension sucks
oops sorry I was preceded LOL

and most apps don’t support pae

oh well fella, thanks for the chat, I am off to bed..

make them almost all apps

the… you can’t cram 2×32 into 64 bit ops. that’s not how it works.

you’re still missing it ..

from a coding standpoint PAE is really really ugly
and mind you I’m a shitty coder LOL
I do storage stuff for a living though

so am I. still I’d seriuosly liek to know what other point there is in 64b…

well if storage doesn’t work out for you, you can still join the typo3 developer team

but first I need to tinker with the sources…

well I prefer the hybrid approach, 64b kernel (with mmap and all) and 32bit userspace (proven by decades of coding by now)

also don’t know of any other point other than allocating large amounts of RAM in a single process

Unless they need ASM guys I can’t help much LOL

hehe

I work for a fibre channel storage company, it’s all quite machine oriented

i’m sure a google of “32 bit vs 64 bit” would show mixed results .. as i said .. depends on the app

Is the current 2.6.22 in Debian tickless? (Wondering about tickless stability/worthiness…)

ill sure look into that. but not today, local time says “way too late”

oh well abrotman, every debate in the linux hosting world is awesome, starting from torvalds vs. tanenbaum

I recently installed postfix and when it asked if I wanted to use Local Only or Internet Site I chose Local Only and I was supposed to choose Internet Site. How can I figure out what I need to change to get it back to Internet Site?

it always does .. damn silly 24hour days!

it’s half past midnight in my lovely TZ

dpkg-reconfigure postfix (IIRC)

happen to know ananke from #slackware? man had I darn nerdy geek arguments with that guy

jae, cheers

ah, back then…

no, I never logged into #slackware myself, too busy chasing girls on the internet when I was young

LOL

catched any?

yes sir, as a matter of fact

MB or GB?

but of course not on IRC

hehe

girls were in p2p nets
downloading backstreet boys and shit
one day I said “sure I’ll fix your computer”
and I ended up fixing all the rest LOL

after sharing some trojans with them?
(hmm well even that could be a sexual metaphor)

LOL that brand is not published overseas, it’s not mirrored on non-US LOL

if I want to rework a .tif image…would gimp work best, or would scribus, or inkscape ? ideas?

bonez39, for tiff files I always used gimp and it works

scribus and inkscape are vector based IIRC

isn’t tif a bitmap format?

Is there a way to launch gnome on only one screen? (I am not using the xinerama extension)

Let’s not forget xara either.

corel xara. yeah baby.

It’s by Corel?

it was in 1995

uhm .. one monitor? or one viewport ?

it’s work but how to add internet for 1.x and 2.x ? what i need to masquerade ? what go out from eth0 (internet interface) ?

and Gnome or X itself?

hey guys, i just completely removed mplayer and mozilla-mplayer and reinstalled the totem-mplayer plugin but it still plays with mplayer, any ideas?

Xara Xtreme. It was in Debian, but I can’t find it right now.

I told you that was the problem

what is “it”?

now, use destination routing instead of masquerade

mpeg, avi, wmv, etc.

masq it’s only good for a single host

hello my favorite peoples in the worlds

germans also included?

sure

got it, nevermind

I hate everyone equally

in other words, use your gateway as a router rather than masqing

hehe

and bear with me, it was three years ago last time I used iptables in a prod network
I was so happy

ok..gimp then

i need to use FORWARD new time ?

LOL

FORWARD with destination option

one screen. I have one display with two screens.

… :p

uhm .. your monitor has two screens ?

smoke break

Oh, right. xaralx
!info xaralx

you’ll find examples on the net, but the matter is, use forwarding between the local nets (like a local router) and upstream from the default gw towards the internet. a forward rule will do, check the howtos

(Heavyweight vector graphics, illustration and DTP Program), section non-free/graphics, is optional. Version: 0.7r1692-2 (sid), Packaged size: 4783 kB, Installed size: 11820 kB

Each screen is attached to one monitor”

It’s in non-free right now, but I think they’re working out some chinks so they can move it down to main.

don’t even mention smoking, I quit 10 days ago for the kids LOL

by “screen” I am referring to what X calls a “screen”, which is different (usually part of) a “display”

hrm .. i’d guess you’d have to monkey with your X config ..
maybe ask #xorg
i dont’ think it’s a gnome issue

ok thank for your help i have better understand ” I will come back ” after my congratulation :p
see you soon everybody

s/is different/is different from/g

I’m really honoured to help
really
that makes me happy

not certain .. :p
thank

that’s probably why I’ve been in support for the last 10 yrs lol

see you soon :p
+++

see ya
ATH+++
LOL
if someone still remembers that

It’s not an “issue”, it works great for the normal use-case (shows up automagically on both screens), I am just trying to find a way to make it only start on one of them (DISPLAY :0.1) and leave the other one unmanaged.

it’s an “issue” for you

is anyone here aware of a method to achieve that?
werd.

that’s why i mentioned #xorg

let me understand, do you want not to initialize a separate monitor on the same vga as part of the same display config?

!savage

http://www.probo.com/timr/savage40.html

gotta run

no, I just want to have gnome not draw anything on one of my screens.

The etch installer seems to have the annoying feature that you cannot easily select all locales to be installed.

why would you want all locales?

I’m on my eighth computer now where I spend about two minutes quickly hitting the down and space key to get them all selected. :-/

so you want X, just not gnome?

Why not?
In case we ever get some Zimbabwian prof here, man, I’m gonna give him Debian in his language.

dpkg, savage is also if it locks up when 3d is enabled, try Option “DmaMode” “Vertex” in the Device section of xorg.conf

okay

I don’t use any locale LOL

good point.

and I’m not a native Eng speaker

personally I hate localized linux

i’ve seen a ’select all locales’ option
i don’t remember where, but probably dpkg-reconfigure locales

I hate locales, another dispersion of work. learn english. i did. LOL

Lo Im trying to get htaccess files to work Ive set AllowOverdie to All in the default file and corrected the DocumentRoot

October 31, 2007 at 9:29 pm · Filed under Tomcat

AuthenNTLM on cpan

what brand is your bot?

its a “got at” wollabot
removed text lookups added db etc

well, thanks for the information, too bad it’s not useful, since mod_auth_sspi is a windows-only module, and apache2

Permalink Comments off

hmm im trying to half-ass tidy up this old php4 piece of shit site that i dont really care about too much if i

October 31, 2007 at 8:29 pm · Filed under Tomcat

is there a separate channel for the development of php itself or can stuff like that just be discussed here?
uis there a separate channel for the development of php webhosting itself or can stuff like that just be discussed here?/u

and sorry ari-_-e, i have no idea

ari-_-e: you would want to contact the mailing list for that kind of stuff

Sno, ok, thanks

ari-_-e: http://us3.php.net/mailing-lists.php check that out and go to the Internals list

okee, thanks

hmm… im trying to half-ass tidy up this old php4 piece of shit site that i dont really care about too much… if i make a Base class which has a Base-DBConnect property, that the other classes extend from… Should I also make a Base-User property (User being ab object with all
roles/permissions/user details etc) which other objects can extend from as well? or is it better to have User a seperate object?
just wondering what’s more common practise
to have the User object inherited into every class, or to have every page load a seperate standalone user object for use in the page.

Sno http://discuss.pastebin.com/d3901e129 does it make any sense to have the class?

I don’t see anything

ugh
http://discuss.pastebin.com/d4901e129
i have to manually type in urls because this is on a different computer… dedicated irc machine

lol
there is no real common practice with that
its basically do what you think is best

so does it really make any sense to have the class or am i doing something right?

give me a sec

‘postgresql’,

hey all

If that’s some PEAR package, you’d have better luck in #pear

DB Error: not found

wb j2daosh

!+pear

lol

Sno ok thanks

your still here?
we have to get lives

hmm…where’d the bot go?

you are missing a few { } go find where to put them

j2daosh this is a dedicated irc machine. it’s on 24/7

ok.. thanks

Sno i’m not missing any. why do you say that?

ok thanks

you are go find them

well since i already know you, what is that thing called that people put on their site to stop people from brute forcing logins? it makes you verify characters in an image

CAPTCHA

missing 3

Sno oh the tabs in front of the second bit are just because i pasted it from a different file and was too lazy to remove them
j2daosh check out http://www.recaptcha.net

know of a good tutorial on that?
lol one step ahead of me… i like that

Google probably knows of several

so you mean that $_SESSION[user_id] has {}
and you should be using “” or ” in $_SESSION[user_id]

It’s not missing any {}… it would be borked if it were
oho
that might be doing it

…

does it make sense to have the class in there though?
i mean, technically i could just put a while in a while xD

you got a nested looping query in there
might cause some resource issues
ok restate that
it WILL cause resource issues
nested queries bad
nested looping queries worse

then how do you suggest that i accomplish what this does without doing that?

hrm
let me open my handy notepad++

go ahead

ok the short answer is do a while in a while
long answer is figure out another way to do this

crap, i have to install libraries and stuff

I’m trying to run a mysql thing but it won’t run and I want to output the error. I’m assuming I should be using a php function to do this?

i don’t know how to do taht yet

lol yea
you need a quick captcha?
http://captchas.net/

j2daosh recaptcha shouldn’t require you to install libraries

its a remote system

well i signed up at the link he gave me

j2daosh you copy one file over and read their documentation

but i have to “Download the reCAPTCHA Library, extract recaptchalib.php in the directory where you your forms live.”

it’s really very self-explanatory and you basically require_once the recaptchalib.php
yup

but i dont know where my forms live lol.

its possible to do it in one query

j2daosh where are your forms based out of?

how do i append an array to another one in php?

do i put them in the main site directory? or am i supposed to drop taht in like the cgi-bin directory thing?

$array1[].append($array2)?

lol no
array_merge()
http://php.net/array_merge

i’m fairly certain there is a way to do so, but i never thought about it

my forms are all in the main site directory (www/files_are_here)

cool.

j2daosh so put recaptchalib.php into www

oh, ok

$result=$connection-query($query) .

i think i can manage that lol
oh… duh

What is $connection?

sno i’m wondering how i managed to write those two bits of code so poorly. the rest of my site is rather sexy and refined

its just like any other .php file i have made… god i feel dumb now

TML, that’s my new mysqli(blahblabhablh) connection.

$result = mysql_query($query) or die(“bA fatal php mysql web hosting error occured/b.\nbr /Query: ” . $query . “br /\nError: (” . mysql_errno() . “) ” . mysql_error());
its late?

php.net/mysqli-error

yea that
worry
thought you were talking about mysql
lol

Sno, what mysql command am I looking at to put that into one query?

well
its still joining technically
but you just don’t request any info

oh. but that won’t separate it based on category :-
it’ll just make it one huge list of rows

hrm

$result=$connection-query(“INSERT INTO sip (id,name,host,nat,context,mailbox,qualify,username) VALUES(NULL,’$username’,'dynamic’,$nat,’$username’,'$extension’,$nat,’$username’”) Why is that receiving: You have an error in your sql syntax blah near
”;

the reason why this is messy code is because its difficult situation

anyone familiar with cURL here?

:/
it is possible though
mysql can do alot of things that most people don’t expect it to do

My guess would be that you failed to escape one of those values

TML, what do you mean?

there should be some info on the mysql docs examples

empty variables?

If $username was the string “‘foo’”, for example, you would have “… VALUES(NULL, ”foo”, …”
That would be a parse error

I’ll take a look and ask in #mysql. I don’t see exactly how I’d manage to return the information separate enough that I could split it up into different modules, though….
It’s really intended to be displayed in different tables

I know

You almost certainly can do what you need by learning how to use GROUP BY

OH RIGHT!!!
that is it
damn its 2am

i insterted the php code into my site but its not showing the captcha… is there something else i’m supposed to be doing?

I should go do sleep

$result=$connection-query(“INSERT INTO sip (id,name,host,nat,context,mailbox,qualify,username) VALUES(NULL,$username,’dynamic’,$nat,$username,$extension,$nat,$username”); TML, that still shows the same error.

lol

j2daosh you read the documentation?

thanks for reminding me j2daosh
I do
its how you become a great programmer

*sigh*

thanks for what?

php docs = god
mysql docs = jesus
lol

php.net/mysqli-escape-string

yeah i am reading the docs now… its talking about the recaptcha_get_html function and the other ones… but its not telling me if i should be using them. it seems like they are just explaining how it works in the code they told me to paste in my site…

j2daosh http://recaptcha.net/plugins/php
read the quickstart
i literally copy/paste my code from there every time and just add in the keys
might i get an example from you of how i might split up mysql results using GROUP BY into different tables? keeping in mind that i have 2 table styles, a thin one and a thick one that’s specified in the modules table

give me a few min and I’ll try to figure it out

thats exactly what i did, but it didn’t show the image on the site… it fails me logining in though

oh, lol
j2daosh you put the keys in? both?

oops, nevermind
my browser was caching the site

lol

god i hate when that happens

i had to disable the squid cache on my firewall box for dev. it was sucking as well

lol
that must have been fun

Instead of running 2 queries, you’d look at each row as you were “printing” to decide whether it belongs in the table you’re currently printing, or not.

HAHAHA I just realized that you had limits for each category
*sigh*
hrm

that works even if, say, the user has two tables from the same category with different limits, different order by’s, etc?

one odd thing though… its showing the captcha image hosting above my html forms on the page… i put the code below those though… does php on an .html page supercede the html code?

No, if you want to change the query itself, you’ll have to do a different query

Ah. That’s why I did it how it is, then.
because this way the user can create a certain number of modules (up to a limit, say 6) on the page with whatever his heart desires, no?
except for maybe the limit, was thinking of capping that at either 10 or 25 rows per table

oh right now I remember
this is the classic nested query problem
this will take 2 queries

which means it has a classic nested query solution? right?

not really
the first query will build the second query

what’s what’s happening… no?

that or your stuck
well
never mind
I’ll figure it out
lol

i know i’m a pain in the ass, but really, i’m worth it!

your worth it?

Sno, damn straight.

“you’re”, not “your”.

too tired
3am

lol

give me a break
long day and problems with girlfriend

None of which are excuses to get sloppy with your spelling

so what’s new?
lol

no but the interenet is!

rarely do i not have problems with women.

all the womenz love me

far too confusing

lol

just accept my invitation to a nice dinner and be happy with it.
lol
the problem is that they all have jobs which seem to consume their lives except for when i’m working

anyone familiar with cURL here?

every day, at the exact hours that i work)

I think KairuMaiku is right it has to be nested

Just ask your question, if someone can help, they will

its because of the limit
it it wasn’t for the limiit it would work

if i were to get rid of the limit?
oh wait
nevermind
i’m not getting rid of that

yea then it would just spew them out
yea exactly

i don’t need a user fetching 6,000,000 rows every time they view the index page

it would spew them out in order! but yea
lol
so you got 1 + number of modules queries on your hands
I am sure you could design the DB in a way
to do it in one query
but its too much work

is it too much work if i ask you about it in say, 12 hours?

…
ass

???

do your own work

that’s a subtitle format y’know
lol i’m just kidding

I’m gonna go to sleep
3am is time programming make no sense nor does english

indubitaly.

its been good talking to you

you too. thanks for your help

I forgot how nice #php was
I’ll be sure to come back soon

with my unescapable curse of loops in loops it feels like a rollercoaster

lol
fine ask me in 12 hours
night

ok, odd problem

using IE, i can pass the recaptcha correctly and login… with firefox, i can’t

noscript?

ahh
that could be it
ill check

c/could be/is

nope, its set to allow

you know i don’t know why i did that, because i don’t understand how to use the /// system to correct mistakes
j2daosh on both your site and recaptcha.net?

im not sure about the recaptcha.net, ill check that too

just use s where you put c.
lol

ahh ok

???

i have noscript gobally allowed and still dont work

is anyone familliar with NSSBoard?

j2daosh what is it doing / not doing?

it is supposed to create a table in my database, but it does not

DARKGirl just ask your question.
ah

instead, it gives me error

what error?

it fails to login with my recaptcha error message in firefox, but works perfectly in IE

mysql_fetch_assoc(): supplied argument is not a valid mysql result resource in …/logon_user.php on line 21

j2daosh what’s the error message

im actually not using it for its forum capabilities, but its username/password login capabilities

the one i made for it (“you entered the words wrong.”)
what is the mysql query your performing?

means what it says. you entered the words wrong. link me to the page

…it looks VERY promising, if I can get it to operate properly
mysql_fetch_assoc()
oh wait, you mean the code

might be a good idea

well, lets see…
if($row = mysql_fetch_assoc($result)) {
but I know what its not doing
its not creating a table in my database for some reason

is it code from nssboard ?

aye

ok dark… what is the mysq query your doing

i dont understand ‘query’

your code you just postd is just saying “if i have rows returned, do this

are you a real girl DARKGirl?

some poeple still not use db absttraction layer ? at least for mysql/msqli …

The question of the centruy.

lol

Yes

superb answer.

heh

query would be something like, $query = “SELECT * FROM ‘test’”;

lets see

or whatever variable you named it

$ss-Destroy();
$ss-Open()
$_Session['logged_in'] = true;
etc, etc, etc
and then an else statment

why are you pasting those things?

echo “Unknown username or password”
because j2daosh wanted to know what the rest of the code was
and mind you,

paste the code in www.pastebin.com and then show us the link

j2daosh doesn’t matter horribly because it’s under $1000 usually

yeah, paste it

in addition to the mysql error, I am receiving unknown username and password

j2daosh i’d have to have him web design 20+ pages to bust a couple grand.

it sounds like you’ve brokened the intertron.

erp that was meant for /msg lol

badly. and may need to buy a new one

lol i know

http://pastebin.com/d6b586df2
there it is

ok, gimme a sec

after line 19, put echo $sql; exit;
then you’ll see the sql being sent. try and run it in your phpMyAdmin or mysql or sqlyog.
lawl!

would it be a bad idea for me to boot my comp into ms-dos 3.0?

never used mysql_unbuffered_query() before. weird.

um
I DO NOT have phpmyadmin
I do have navicat though

well try that. see what its querying.
hmm should have session_start(); at the top of the page i think?!

im telling you

might be in the config.php

its querying nothing!
theres nothing to query

do you know that from echo’ing the $sql?
if not, do it.

fine, fine
moment please

nein!

well if it shouldn’t be doing a sql query, you shouldn’t be asking it to see if there are rows

EoN that would be proper if she used “moment mal”
not “moment please”

lol

SELECT id,username,email,realname FROM developer_user WHERE username=’test’ AND password=’38200006543e005b94ab316e87f72921′ LIMIT 1

j2daosh you sure you don’t have js disabled somewhere or something?

there you have it

im pretty sure i haven’t, but i may have some whacked extension that is messed up

brb, need some drink

but if it was me, it would work for you
and i thought you said it wasn’t working for you

j2daosh it isn’t.
it is working on their non-js version though

Just out of curiousity for you Linux developers, what text editor do most of you use for php development?

ok good, as long as it aint just me
textpad

disable js in noscript for both your site and recaptcha.net and refresh and try

i did

ed is leet.
j2daosh works on that version?

Not windows

i even cleared the cache, then closed the browser and retied
oh, *nix, vim
vi Improved

most people will say that emacs wins. it really fails.

I was hoping for a GTK app. I’m using gedit which gets the job done pretty well.

complete with color syntax and all

try ed. even though it’s straight out of 1979 (literally, the first editor), i’m a lot faster at it than a regular editor

back

Aptana works nicely on *nix too

ok, DARKGirl, i’m kind of a newb when it comes to php, but what are you trying to do with line 19?
it looks like this is your login page…
or your login.php rather

why yes it is
you are exactly right!
that is my goal

ok, so your doing your $_POST, md5′ing yourpass, and then your attempting to check against the DB to see if its a valid user right?

aye

ok, quick question. Why are you using LIMIT 1? technically, there should only be 1

don’t use limit 1
because then if it matches 2, your system isn’t secure ; )

boo

whats the diff between mysql_query() and mysql_unbuffered_query()?
who in here uses Firefox?
not you KairuMaiku, we already know it dont work with you
and also, doesn’t session_start() destory and recreate a session?

dunno
me
aww
no
!+enter

run the query in your query runner
and see the results.
ERGH hergggh erggh

dangit!!! what the hell am i doing wrong with this thing
s/hell/heck
sorry

change mysql_unbuffered_query( to mysql_query(

whats the diff between the 2 EoN?

i’ve never seen the unbuffered one in my life
hahaha

lol
eon, you use firefox?

momento

are you itailian or something?
s/itailian/italian
not enough coffee tonight
speaking of qhich… ill be back in a sec

yep, i sure do
there is no other browser

ok the query did nothing

Salut Flash_ouille
hi!

hello
hi uu uu
is anybody in there

Ive tried to run the query but I receive the same error

Is there a php function for auto-generating a random password?

md5(time())
Trim length as you see fit.

Salut Demon`
Salut JTyler
Salut BankHead

what is the error?

I need to list the result of a “while” do you how to do it? i need to get my result $result= list of the while’s result
Salut speed|89

thanks.. interesting idea..

Salut Nexcet

lots of errors

no.
copy the sql query, paste it in your MYSQL (phpmyadmin, or hwatever you use)
run the query direectly on mysql. and tell me the result.
it wont give you an error about session destroy etc, because mysql knows nothing about the session.. you’re only pasting the SELECT statement, not the php

i dont understand ‘query’ do you mean my entire php code?

SELECT id,username,email,realname FROM developer_user WHERE username=’test’ AND password=’38200006543e005b94ab316e87f72921′ LIMIT 1
php passes that exact text to MySQL. mysql says ‘ok thank you, heres the result’. what we’re doing here, is removing php from the picture, and seeing if your mysql query is correct. so go into navicat, and paste that query (somewhere will allow you to write/paste sql)
lawl!

ah ha!
Table ‘developer.developer_user’ doesn’t exist
did you see the error?

sure did
there we have it
it would appear the table developer.developer_user doesnt exist

but thats what i said in the beginning!

lawl!
you did not say that at all
Warning: mysql_fetch_assoc(): supplied argument is not a valid MySQL result resource in …/logon_user.php on line 21
i dont see anything about a table there.

but i did

at least we’ve esetablished one thing, you’re DEFINITELY a girl.

I said that the install did not work at all
i was told to create an empty database with no tables

ahhh ok. do that.

ok, what did i miss?

ahahaha!

the install was supposed to create the tables
but it did not

what was the messages it gave when you ran it?
failing that, i suggest you buy some ‘my little ponys’
and dress them in various outfits

the error you saw

lol
YaY!

you there dark?

i am

ok, i have been gone for the last 20-30 minutes so i need to get up to speed. have you tried putting an echo in front of your $sql and then copying that line of output off the page and running it thru a query analyzer
?

yes

and did it give you anything?

and i get ‘Table ‘developer.developer_user’ doesn’t exist’

ok, then you need to create the table
and/or the database

developer_user is the table, yes?

what are you using for mysql? phpadmin or mysql administrator tools
yes

im using navicat

DB = developer | DB_Table = developer_user

also most mass hosting admin panels apply a ‘prefix’ to mysql databases
usually the username
so if the hosting machine is running something like ensim or directadmin

that is also correct

keep that in mind
usually it’s the username of the database owner

i get hosted thru siteground and they make me prepend my db’s with my username

i have a half dozen different colos that are all configured differently

so when i make db requests, i have to refer to it as $database=”username_website”;

ill add it

lol i started prefixing my databases with my username way back during like plesk 3

“website” being what i call my DB
well i used wamp up till i bought my hosted domain. it was much easier in those days
but i didn’t have the bandwidth for it

ok i’m off to bed all
ciao

later man

Adding prefixes to your database tables is stupid. Any host will give you your own DB- you won’t have to share it.

MarkR42 we’re talking about db names.

Oh I see

and usernames

DB names could be anything of course
I had a project where THere were 100 sites each with their own DB, the DBs obviously had to be named sensibly

MarkR42 but the chance of multiple users wanting the db name “forums” on a server hosting 2,000 people is high

Right, so the sysadmin needs to choose the db names
Or at least, not allow sharing

MarkR42 hard to do when the users are the ones who create the databases, no?

salut a vous!

well if the site owner is the DB admin and he aint friends with the 200 + other admins on the hosted server, obviously there will be DB conflicts

cPanel, Plesk, DirectAdmin are all set up so that the users create their own databases, but they’re all prefixed by their username

I see. I hate those things.

MarkR42 good thing I didn’t mention Interworx then! You can’t hate that

brb, lightning just stuck my building

I’ve no idea what that is

Control panel. Very nice.

But if it’s a control-panel-to-manage Apache-and-mysql, I can guarantee it sucks

Not quite as nice as my company’s custom solution, but nice indeed.
MarkR42 you can manage Apache and MySQL, but to an extent. It more or less is a total system management bit for hosting companies to set up users with.

exactly

The problem I find is that a) They have awful security holes b) They bastardise every package they touch so that it can’t now be configured in the conventional way any more

my hosting company uses cpanel

but like i said. i still prefer my company’s custom solution. we can manage all of that and billing from a shell

Also, c) Idiot users think that screen-scraping them with e.g. curl is a really good idea

MarkR42 obviously you haven’t used interworx

Maybe it’s very good

It’s good enough to the point where someone running centos can install the latest version of php via yum
Or likewise with freebsd and ports
or compile it on their own, if they’re so inclined
it has its own packages for all the internal stuff so you can’t screw with it unless you remove those

cpanel held back the php5 roll out by years.
single handedly

how so?

they didn’t release it in /scripts

i was using perl when all that happened i guess

meaning that whatever unbelievably high % of sites on the net using cPanel couldn’t use php 5

because cpanel only worked on php4, so lazy/fkwit web hosts didnt want to go to php5 which they couldnt run cpanel with

EoN has to do with migration stuff

but in the past month or so, cpanel has become php5 compatible.. probably because the php4 End Of Life has been announced.
cpanel need their backs whipped with a hammer.
true. its a bit of both i reckon

hmmm.

no host wants to migrate a server
let alone hundreds
been there done that.

my cpanel has 5.0.5….

does anyone know of a good (or at least moderately acceptable) video conferencing solution that could be used for an online classroom environment ?

and has for atleast 3 monthes…

j2daosh was in beta for a while

i know that was seriously off topic.

i think php 5 came w/ cP 11 beta
don’t remember. like i said, i don’t use it often
ok i’m really off to bed now
gnite

i’ve determined that 90%+ of the apps sourceforge are awful.
that or their search system is highly inefficient.

lol
i always thought the interface is not crash hot

man
I just wrote a long ass question, and as I was writing it, I found the solutin
it was like, 4 lines

does php send any identifying user agents or something?
with fopen wrappers

TehSausage, why don’t you find out?
Or shall I?

then again, it could be my server blocking it XD
because it doesnt give a certain header

You can ask it to send one, but it doesn’t send a UA string by default

fopen(“http://somehost:80/somepage“); right?

yeah, i think i know why
i’ll put myself on the proxy allow list

im using cpanel x
yeah i know, that was like so 20 minutes ago but what can i say. I love stormy weather.
lol and he aint even here anymore
or well

connect to [127.0.0.1] from localhost [127.0.0.1] 49536
GET /somepage HTTP/1.0
localhost:1337
That’s all it sends.
GET and Host

okay, looks like the slow backup will work as long as my ip doesnt change half way XD

hi. whats the easiest way to have php send email on a local testing server? server is ubuntu with lamp stack. do i need to setup a mail server?

Hi, how do I save an image from a url resource? (url resource containing GD code to create image)

install a mail server such as exim and a mail user agent such as kmail. Configure the exim to deliver mail for some fake domain locally. Send mail to that fake domain from PHP, and pick itup in a local user spool with your mail user agent (e.g. kmail, thunderbird)
If it’s simply an image that already exists, you could load it into a local file with copy(). No GD is necessary. This of course assumes fopen wrappers are enabeld.
e.g. copy(‘http://some.server/something.jpg‘, ‘/local/writable/path/something.jpg’);

oki, im taking a zzz and ill finish (hopefully) the code tomorrow

MarkR42 I might have a syntax problem, do you mind if I show you a quick pastebin?

well, later today

MarkR42, thanks. And how do I tell php about the mail server? The only ini settings i know are for sendmail and SMTP, the latter of which is Win32 only i think?

You don’t need to configure PHP for this. It will use the installed MTA. You need to configure your local MTA to accept messages for a fake domain (e.g. test.fake)
Then you can mail youruser@test.fake

MTA?

and it should drop into your local mailbox where you can pick it up with the MUA of your choice

MarkR42 I create the image using php – header is set to : header(“Content-Type:image/png”)

mail transfer agent?

MTA = message transfer agent, MUA = mail user agent (e.g. interactive email program)

MarkR42 in a seperate file called imageCreate.php

cool – i will give it a whirl. thanks mr.

If you want a hint, the Ubunutu standard MTA is exim. You can apt-get install exim4, then dpkg-reconfigure exim4-config

well, the one time it seems to be working, flush() stops working XD

nice! thanks

Use any mail program you like as long as it can read from a local user mailbox
I know kmail and thunderbird works

thunderbird will do fine of evolution doesnt work

I think evolution has the option of reading a local mailbox too

Warning: fopen(imageCreate.php?color=blue) [function.fopen]: failed to open stream: No such file or directory in /var/www/public_html/test/mail_mime2.php on line 6 http://pastebin.com/m275a41be
Is it possible to dynamically create an image in one php file, then in another file, save it to disk, and then mail it as an attachment?

You don’t need to use two .PHP files for what you’re trying to do
Put the relevant parts into functions and call them from a single page

MarkR42, but I want to re-use the imageCreate.php file to display the dynamic image in parts of a site, then mail the image when needed

kekeke

Put the relevant functionality into a function then, include the file with the function and call it.

in parts of the site, then when I need, make the image into an attachment (might not have to save the file to HDD, just think it would be easier)

Making HTTP calls to your own pages is pointless

did you see my pastebin?
http://pastebin.com/m275a41be

Yes. However, the standard mechanism of code reuse is: 1. Put the stuff into a function 2. Put the function into an include 3. Call the function as necessary

Does any1 know what the name of the background of required text fields in CSS is called?

I should think it’s the same as non-required ones

actually
i think its just e-mail
e-mail and aim
do u know the names of those in CSS?

There is no specific CSS selector which selects based on the function of a text field.
CSS does not include those kinds of semantics (yet)

can i show u my code?

Yes but only if you stop calling me “u”

MarkR42 is it possible to save an image from imageCreate.php ?

Type in English please

sorry
lol

And don’t “lol” either
It makes you smell like an AOL user

lol

I hate AOL

Of course, but don’t do it. Instead put it into a function.

There’s nothing in our channel guidelines that prohibits someone from saying “lol”

I didn’t say it was in the channel guidelines, did I?

Then judge their speech elsewhere.

Alright go here http://23663.darkstargamers.com/phpBB2/profile.php?mode=register&agreed=true

Though…
php-bot: tell duckhunt about typing english
That’s for the “u”

thats what the highlighted text box
is i want to change
and here is the CSS
http://pastebin.com/d7230a099

free shared hosting 3 sending my entire phpbb database uncompressed over http

Examine phpbb’s code to determine what classes / styles it uses and/or add the appropriate ones as necessary. We don’t support phpbb anyway
Black on black is not normally considered good for usability

no thats fine but the bright yellow with white color font
You cant see what you type

I imagine you are getting problems from a browser extension changing the colours of your page.
Disable *ALL* browser extensions and try again
And if it’s a specific browser extension (e.g. Google toolbar, yahoo), read that extension’s documentation
This is severely off-topic for ##php anyway

MarkR42 Can you suggest a way to turn the output of the imageCreate.php file into an imagefile?
MarkR42 without using functions?

Yes, rewrite it as a function and call the function
No, rewrite it as a function and call the function
THe only sensible, logical way of reusing PHP code within an app is by putting it in a function in an incldue file, including it and calling the function

MarkR42 how do I put a function call in a tag like img src=”php function_1(); ?” /

On no account should you be making HTTP calls to your own pages, it’s not only pointless, but complicated and more resource intensive.

thanks it ended up being Mozilla

Make another page which you can reference from an image tag which calls the same function

MarkR42, i went through the exim4 install. what kind of account should i be adding in thunderbird? standard email / unix mailspool? pop3?

I doubt it. I expect it’s a browser extension
unix mailspool
Disable ALL browser extensions, and try again
I’m sure you’ll find it is one of those

MarkR42 ok, I will be back then soon

is there a way i can check the presence of an ftp server on a certain domain?

FTP servers don’t run on domains, they run on hosts.

or hosts, rather
sorry

On a given *host* you can check for the presence of one, yes

can you show me how?

I expect there’s a PEAR module you can use to connect and retrieve the banner. Or you could just connect with fsockopen and read the banner yourself if you know a little about the FTP protocol. If you want to actually connect, you’ll need a username and password, then you can use the fopen
wrappers.
And no, I can’t show you how, but there are some ideas.

dont want to connect, just want to know to see if my ftp server hasnt fallen over

Oh, so it’s a system monitoring application. Consider using something like nagios instead.

yes
okay ill have a look, thanks for your help

MarkR42, works great – thanks mate.

So you can ssend yourself messages from PHP and read them locally? Cool. I think that’s a good dev environment

MarkR42, havent done it via php yet. i need to make the validation in my script a bit more relaxed as it wont accept user@localhost

does firefox automatically try to parse xml files with rss version=”2.0″? my rss generator generates good xml according to w3, but when i replace the root node of “foo” with “rss version=\”2.0\”" i no longer get output and my rss feed aggregator doesn’t show any listings.

MarkR42, just for testing obvioulsy

maybe should ask in #firefox

You can configure the local mail name as whatever you want, and accept messages for other domains too.

it’s a tangent question to… does $this-xml-createElement(“rss version=\”2.0\”"); break the createElement function in any way? i don’t get any output.. no errors, nothing.

Do you have all errors and warnings enabled?
That sounds like an invalid name for an element

$this-xml-createElement(‘foo’) works fine.. everything’s there.
yeah, it’s set to display E_ALL

Last time I checked, the DOM libraries actually did let you create invalid names of elements etc

I just tried calling createElement with a similar invalid name, got a Uncaught exception ‘DOMException’ with message ‘Invalid Character Error’

hm.

Is $this-xml definitely a DOMDocument ?
Or is it some other third party class?

yeah. i’m using xampp for the first time due to an issue with another package needing a version of mysql i don’t have installed.. maybe xampp’s being goofy.
$this-xml = new DomDocument(‘1.0′);

Calling createElement with something which isn’t a valid element name should throw an exception. It does on mine (5.2.1 CLI)

hmm.
i’ll try the cli

can someone please look over a script I’ve done and give me a few tips on how I Should do things better? it would really be appricated…

huh. ran it through the cli (5.2.3) and i got no output.

No exceptions?

nope

$xml = new DomDocument(); $xml-appendChild($xml-createElement(“not valid=innit”));
And it threw an exception

changing “rss version=\”2.0\”" to “rss” gave me the correct output.
no exception on that..
wtf.
hm

Are you absolutely sure that you were running with all error logging, reporting etc

i’ll double check.
maybe this xampp thing is wonky

Try just doing throw new Exception(“test”);
Perhaps the exception is being thrown but not output

ini_set(‘error_reporting’, E_ALL); ini_set(‘display_errors’, TRUE) or 1 or ON.. still nada.
hm.

kuja, hello

nothing in the apache logs, either.

Is there some flag which makes an unhandled exception silently exit?
Have you tried just throwing an exception from the php CLI?

i’ll try that now.

Incidentally, I just tried php -r ‘throw new Exception(“hi”);’ and it threw an exception then crashed PHP

hah

hi all

Figured it out. $foo = $this-xml-createElement(“rss”); $foo-setAttribute(“version”, “2.0″);
still can’t figure out why no exception was thrown.

what library is “recommended” adodb or the one from pear?

doesnt crash in 5.2.0

Hey folks
I’m a php hobbyist and I’m having trouble with a crap CRM app that I’m editing
Basically, I can’t find their API docs (LiveArticle) and I’m just trying to edit a class
I am at the right place in the code, but it keeps saying that my class doesn’t have a custom metho
*method
is there some new way of making methods for a class, besides inserting “function blah() {}” inside the class definition???

can anyone tell me what is the best option for going to use a database library for php? adodb? pear? creole? propel?

thx – I’ll recheck my code for starters….

BUT! classes can inherit methods
bemson, check the classes definition

understood – but this one isn’t being extended from another class

for extend or similar
^^

ok
I didn’t create the class, just trying to modify it

well then remember you candefine a function inside a function
like include() a file that holds function definitions

I’ll try that too – isn’t the same-named function the one invoked at instantiation?

either that one or __construct

And isn’t there another name for that function now…
exactly

^^
also __destruct respectively

I’ll make a construct function then, and see where that get sme

thanks much noimnot

not supported in PHP4 tho
yer welcome

laterz!

If I use LGPL library and my project use GPL licence, should I include LGPL licence too?

if you care talk to a lawyer or read the license
it’s all so fucked in the meantime….
i use the NPL (Nihil public license)
“nihilo ex nihil”

“nihil novi”?

meh?

i use the PMMF license

thx again noimnot – I’m good (nuff) now
later folks!

it used to be known as the “Pay me or I’ll send some Romanian thugs to cut your ears off” license, but that was too long.

hey guys is it possble to make a form with a get request return a address like /this+is+my+search ?!
with mod_rewrite

yes, also without mod_rewrite

uhh sweett how ?

you either use the 404 forward handler, split the request uri and set a $_GET variable yourself, then incldue your site
or just read the man of mod_rewrite

hmm i know
but it is a form that is submitted
and it auto prdouces ?query=

oh
tried having no name for the fields?

know all the other stuff
its stil making the ?=

then no
read the rfc

okay

also that has nothing to do with mod_rewrite
it is client side

okay :/
solution could be make a default handler
that is redirecting with the parameter S

use mod_rewrite to rewrite the url
+

ye iknow

remove the get parameter names could work

time wise could i use = to compare 2 times in the format “11:42″?

kevingruscoe, no

How would i go about comparing 2 times like that then?

but str_replace(“:”,”",$time1)=str_replace(“:”,”",$time2);
or explode(“:”,$time); and compare 2 arrays
your choice

Hello how can I select a row from a MySQL-table like this? $sql_get_userdata = mysql_query(“SELECT * FROM users WHERE MD5(user_id) = ‘” . $_SESSION['adtrader_se']['auth'] . “‘”); (note: MD5(user_id)) Hope u understand what I mean…

copyninja, #mysql

noimnot, it’s kinda dead in there but okay thanks

there is a reason for this
to many people with ridiculous requests i take it

noimnot, say what?

to many people with ridiculous requests i take it

I don’t understand what you mean.

people that are just to lazy to think asking stuff they could answer themselfs

Generally speaking, people will store the MD5 hash and then compare against that, rather than what you appear to be doing which is storing the unhashed value and hashing every time for comparison

noimnot, If you assume someone could answer a pretty easy question (like mine) which would take about 30 seconds of their time, why not just simply ask before you start digging the net for answers? But anyways thank you for your time.

help!
I am attempting to use POST vars with a curl script

Yeah, It’s usually done that way (ie with passwords). But I assume there must also be a way to do the viceversa.

copyninja, there is no reason to do it “viceversa”

but some of the data I am submitting comes from a select that allows multiple forms, how do fix the postvars to accommodate the multiple part
multiple selects*

It’s not normal for a database to implement the MD5 algorithm.

FYI, please be more specific

php.net/faq.html

what is “a select that allows mutliple forms”?

I assume not.

sorry – hardly any sleep… I will try to be more coherent

Thank you. I will find a different way to solve it then.

select mulitple

I am using cURL, and need to post to a page. I am acting as if I am the form. The form I am pretending to submit from has a multiple select… I need to know how to fill out the post vars for cURL to submit

oh, so you mean multiple values, k
you are TEH FORM man!

php.net/faq.html

apparently var_name=single_value doesn’t work

on oracle how can i insert a blob to longraw field?
with php

TML – that doesn’t help, thanks though

FYI, find out how the browser is supposed to send multiple post values..

A doubt about SSL. I am going to prepare a website in SSL. There I am going to show a page with FORM. There the use will have to enter confidential information and the form will submit to the server. My server is Apache with PHP. I do not know how to make the FORM page to be shown and get
form datas securely? Somebody help please….

How can I make this work? function a($x) { function inner() {print $x} }; a(“test”);

there are no websites in SSL
SSL is not a markup language

correct.
How can I make this work? function a($x) { function inner() {print $x}; inner(); }; a(“test”);

wam, why are you repeating yourself?

function inner($x)

I corrected myself.

i was about to say something, not i feel obligated not to; so you learn something

That’s not possible.

Then you can’t do it

I’m creating PDO emulator (because MySQL driver is disabled on ByetHost) based on MySQLi. However, is PDOStatement a child of PDO?

Note that there’s really not such a thing as an “inner function” in PHP anyways.

wam, function inner() { global $x; echo $x; }

I wanted to avoid making it global. But when that’s the only way :/ Thanks.

you don’t need that funciton in function construct
wam, meh

That won’t work unless it’s inserted into the global scope by a()

wam, use classes

sure? I have a sorter for usort.

function a() { function b() { var_dump(“B has been created”); } } a(); b();
Note how calling the a() function creates the b() function in the global scope…that’s because there’s really only one scope in PHP

I need a sorter for usort and I find it difficult to call the sorter function when it’s inside a class. It’s just more convenient so.

Use a callback

with usort?

Yes
php.net/callback will show you how to create a callback that calls a method on an object
!+func usort

Sort an array by values using a user-defined comparison function

“bool usort ( array &$array, callback $cmp_function )”
See how the second argument is of type “callback”?

I see…

php.net/callback will show you how to create and use callbacks

I’m in a class, which is only used static.

php.net/callback
Read it
It will answer your question
As an aside, this is a great example of why you should ask your real question instead of asking how to do something that you assume is the answer to your real question.

So what does this help? I could use a method in my static class. But that wasn’t my question. My question was, how could I pass another variable into the sorter method. Because php obviously doesn’t support namespaces, this is non-trivial.
and as noimnot said, $global looks as the only way out.

somebody please answer to my question too.

It’s not clear to me how namespaces would help you here, but *shrug*

If function inner() was inside the namespace of function a(), it would also share the parameters given to a(). Like in nearly all advanced languages.

hello, guys! is there a function for the complete separation line to the array?

Uh…no, in neither C++ nor C# does sharing a namespace mean they share internal variables.
The what?

my question on php
i need parse string to array

Do you really think, c++ or c# are advanced?

I most certainly do
Sounds like you need a basic tutorial in how SSL works. Maybe plug “how ssl works” into Google and start reading.
php.net/explode or php.net/preg-split

they build on deprecated libraries with deprecated access methods and lean on defined behaviour because they have to to be compatible with the old crap.

I think you’re confusing C# for something else.

Really? How did microsoft implement c#? 80% of the libraries are the old ones from the ’90s. When no real OOO was implemented.

As for calling things “deprecated”, “old” and “deprecated” are different words. Don’t use one when you mean the other. C++ is not deprecated.
You’re mistaken.

i want split without delimeters, for example string “abcd” to array {‘a’,'b’,'c’,'d’}

You would agree, if you knew ruby or python or so.

php.net/string-split
I know both Ruby and Python

then you know how easy it can be.

The majority of C#’s libraries were completely built from scratch for the whole .net runtime

I ready already much info, but now I am going to startup. There I get this doubt how to render the form (in SSL/non-SSL).
?

What do you mean “how to render the form”?
If you’re asking whether the form needs to be sent over SSL, the answer is “only if the form contains something you want encrypted over-the-wire”
If it’s just HTML with no sensitive information, there’s no (technical) reason to send it over SSL

hello
do you know some good courses on the iinternet for learning php

php.net/manual
!tell pomp about getting started

thank you for that web site

Permalink Comments off

im trying to graspt the subtle difference between a segfault and an invalid page fault am i correct in saying

October 31, 2007 at 6:17 pm · Filed under Tomcat

round() ?

ok, thanks

maybe it’s in C99 or something

err it’s C99

i see

no std

Permalink Comments off

im trying to graspt the subtle difference between a segfault and an invalid page fault am i correct in saying

October 31, 2007 at 4:29 pm · Filed under Tomcat

$ gcc -ansi -pedantic-errors -Wall -W foo.c
foo.c: In function ‘main’:
foo.c:5: warning: too few arguments for format
i.e., your use of -Wall should have produced the warning you desire.

nice.. -pedantic breaks with a long list of issues from a system #include

good.

mhm
“long long not supported by c99″ nice
C90*

if you are using other than standard c facilities you must provide the feature test macro (or other options) needed to make them available.

what is a “feature-test macro?” You mean a -D ?

a macro which can be tested to detect the presense or absense of a feature

ahhh
ok, that’s what I thought

or, in this case, to activate a feature — the testing is done by the compilation environment.

right.. like… gcc … -DFEATURE
and the “test” is, (of course), the #ifdef out in the source

to request susv3 semantics (on a system that conforms) you’d use -D_XOPEN_SOURCE=600 (or before any includes where it may matter #define _XOPEN_SOURCE 600).
the system tests your setting of the macro. naturally you can test it too, in your code.

damn
it’s impossible to use all the warnings

right. ok – I got worried when I saw “macro” and was reading it as some form of macro that was running a test,

it breaks soo many times in external headers that is a no-go

not having looked at the page in a while, i probably use most if not all of them.

(missing internal #defines etc..)
“Warning GATEWAY not defined in foobar.h”
(foobar.h is fictive

my usual gcc CFLAGS contains 30 options, 16 of which are -W’s.

I fail to see your issue – if the warnings are not about your files, then they are not your problem.

ditto, i have about 20 -W flags.

heh – youse guys go over the top on them-there warnin’s

but others are c89 conformance related, such as -fno-common and on x86’s -ffloat-store.

shall i contact mac os developers?

why? Do you think it remotely possible they care? ;-

why not have the compiler do as much work for you as possible.. -W flags bring’em on!!

http://rafb.net/p/R8eyHZ50.html
just an example of an error i know shit about

some are simply not applicable on all platforms, etc

ls
oops wrong window.

dude, I told you – they ain’t _your_ headers or problems.

can someone look at the paste and tell me which warning flag triggers that?

because someone related to your silly kernel-build stuff will #define it with a value, somewhere

-Wunder -Werror
err, -Wundef -Werror

i wonder how much stuff you coded with that philosophy

quite possibly on the comline

azi_ failure to define GATEWAY has caused the error.

-Wundef is not the case, let’s see -Werror

all of it – it ain’t my code, I can’t fix it.

yes but this is not my header file.. i even don’t know what GATEWAY represents

i often use -Wundef but breaks conformance with the standard.

hmm, I wonder if semaphores are actually better than pipes …

they count nicely.

i’m trying to graspt the subtle difference between a segfault and an invalid page fault, am i correct in saying that when people say their program segfaulted (ring 3 code) that they usually mean just a regular invalid page fault and the OS shut it down?

well, if its required to have a value, ignoring it can’t be a good idea.

well, so do pipes

but possibly much more slowly.

and pipes feed into the whole io scheduling business.

“failed inlining function foo. foo body not found” , what does it mean in practice? (foo is an extern function defined in some other file)

once again.. His macro is out in a systems file and must mean something to system-building. I’m sure that it’s known when and where GATEWAY gets defined, but _he_ doesn’t know.

you should consider reading the gcc documentation.

well, I guess that pipes can only count one token at a time

ok
btw,thanks twkm

cool.

brb

I thought a struct has always to be alloced?

before its fields can be used/filled with proper data
or I am jsut confused right now

BlindMan, I don’t think so..!

the struct must be defined, then used to define an object, though the member names might still be used without a valid obejct.

allocated?

azi_, I think he means using malloc!

because in some piece of code, the fields are filled with proper values, etc. but I only see a struct delcaration but nowhere a malloc call or whatever
I was generally refering to the process not to some specific function

struct foo { int i; } x; /* no malloc needed */

Yes BlindMan Malloc call is not needed until you need to allocate dynamic memory!

malloc is not required to use struct.

struct foo *p = &x; /* still no malloc */

you don’t have to use malloc

i = 42; /* yay! */

static struct foo x;

ic ic

thus now x.i == 42

ic ic – the noise that chickens make when put into a putty filled jar and rolled into the ocean.

err?

to make it short… you define a variable of the struct type, and it follows the same semantics as for other, primitive data types
int x = 3 ; /* where is malloc() ? */

you have tried very weird things

I also know how to make putty.

figured that now
thanks

download the source, and compile?

hmm, do I really want to know?

(how to make putty)

linseed oil and chalk

putty?

sounds like some kind of goo women put on their face.

that’s foundation.

that’s one of the things.

Can anybody tell me how a C program can interact with the Java GUI …

your operating system might be able to

hi
some Brazilian in the canal?

umeshc, which OS you are using?

linux …

umeshc, Check JNI
Ask in #java..

ok
I am not able to join that channel …
I think that channel is not public …

umeshc, http://java.sys-con.com/read/45840.htm

try ##java

umeshc, You needs to be registered..!

#java redirect to ##java

so, try it directly — your client might be broken

umeshc, Check link i gave you!

Zhivago, I guess he may not has registered his username..!

also check CNI

hi
is here anybody who is familiar with RTP (RFC 1889) ?

zmitya, Ask your question if any body would be they will answer..!

I think there’s an update to this RFC

the current rtp is defined by 3550.

yes you are rigth
so, my generated RTP is choppy/ragged within the LAN too
I can provide wireshark capture if sy would take a look at it

Anyone know of a trie package for text? I’m going blind searching.

How do I make it so that fgets doesn’t hang up if there is nothing in stdin?

huh?
what are you feeding it? the keyboard?

It’s a CGI program, and I don’t want it to hang if there is no input when I’m debugging it

stdin is buffered, by default.
feed it a fake file, when it hits the end – you’ll know.

How could I check to see if there is nothing wating in stdin, and skip fgets?

why not obey the web hosting with cgi support spec?

Becuase I’m too n00k
n00b

I suspect you meant you are a b00b

I don’t know what part of the CGI spec you’re thinking of
yes, of course (.)-(.)

there should be like.. at least an \r\n in stdin for a cgi program shouldn’t there?

what does ‘hang up’ mean?

heh

Take the phone handpiece and hang it on the cradle
Thereby terminating the phone call
Commonly happens to me when I ring girls… :’(

heh

stops in fgets and waits forever … .The example program uses fread, i’ll try that

that’s at least 2 cases needing context

you are not a n00b, and you don’t know how to stop fgets from waiting forever !

it won’t help.

Ah, but your imagination will add a far more entertaining context then reality wwould provide.

FILE *’s don’t support non-blocking io

fread will block as well

well, when I saw “ring girls” – I was wondering why you were ear-tagging them
…which would be a “hang-up”

so, cgi is coming up a for telephones? .. hmmm

ssh, don’t tell them…

read a good book the other day, where the enforce is told to “hurt” some gambler refusing to pay his losses.. Guy sneaks in, ties up the wife nicely, and ear-takes the clown – both ears, awake – with calf-tags
enforcer
takes/tags
I suspect such would “make an impression” – several ways

if that were comprehensible that might be a great story

Sounds painful… You would have the hope the guy wasn’t into S&M though, or he might get off on it.

well, iirc calf-tags are a peg about 3/16 in diameter.. Any way you “slice it”, it’s gonna’ impress someone.

so the person got some ear piercings? probably not that painful if done fast

what does atoi do with NULL?

!google man atoi

atoi(3): convert string to integer – Linux man page: [http://linux.die.net/man/3/atoi]

plus the tags, of course.

but the man page doesn’t say

so some fancy earrings, is that it?

except that atoi() does not detect errors.
crashes.

heh – something like that

try it and see.

i see

I’ll use ? :

I would guess returns 0, but that is a wild guess…

really probably not as painful as the author would have the squeemish believe

The fact someone can defeat yer home-security and do all the above sounds like it might impress the homeowner, neh?

heh

It certainly seemed much more sensible than starting off by smashing his knees with a sledge/bat

Segmemtation Fault

amazing

wow
what an interesting conversation to wake up to
i see something about a dismembered dog and a bat and S&M

it’s been a week of cop-books.

probably a bad idea to do that then

I think so

what is the recommended way to covert a string to a int? I seem to remember atoi is not very robust

hmm, I would have thought that a live dog would have been more interesting.

sscanf?

depends on how you want to convert it …

it was, then it wasn’t

sscanf might do what you want

matt__, zid` thanks

but sscanf is not exactly an improvement over atoi

strtol perhaps

you need to think about what you want to do — there are an unlimited number of ways to convert strings to integers …
for example, you might convert “hello” to 42

well, there are also unlimited strings

not relevant

Zhivago, come on I want to convert “900000″ to 900000

“hello”? or mime, etc

from argv actually

what if the string contains “hello”, what do you want to do then?

strtol maybe

sprintf

strtol, or sscanf, etc – check the manpages and try them

pizza_, tell the user he’s sick
pejo_, how could sprintf can help me?!

what about 10.5

s/can//

it helps me sometimes
so i should help u too i think

sprintf does the reverse

it converts things to strings

alesan, a routine as simple as atoi can be written in no time.

sscanf is better then?

divide by 10 and push, repeat until ax = 0

strtod strtol etc

pop and print

wobster, why should I re-re-re-write it…

oh wait, that’s int to string

at this point, I’d say folks are trying to get you to learn to think AND code

alesan, because you obviously can’t. Like wasting your time here would’ve sufficed to write that.

wobster, converting strings to numbers is a bit tricky I thought

just use strtod

AntiLiberal, for decimal + ascii it’s trivial.

for(index = 0; istrlen(str); i++){num+=pow(10,index)*str[index]; }
there ya go
erm, -48 that
‘0′

zid`, no. ‘0′ is not 0
better

*=10 ; +=

PoppaVic, I just asked for e more robust call instead of atoi().

properly detecting and handling all possible errors is less trivial

zid`, strlen is superfluous.

strtol

use strtol or strtof or strtod

yes, yes it is

strtol?

pow tpp
oo

wobster, you may think so I don’t care too much

it was a one liner

hey
strtol is ok

that wasn’t the implementation that was the algorithm

and you heard of some. Did you check any of those manpages that folks suggested? or even try any of them?

zid`, even in a one-liner strlen and pow is superfluous

I have already written the code and it works. thanks

Does anyone here have any reference to the rationale behind the openat(), fchownat(), fchmodat() and similar POSIX calls..?

zid`: for a start that converts the number in reverse…

I’m trying to understand why they’re necessary

neat

never heard of them

I don’t write for accuracy over irc ;P

zid`: as in 123 becomes 1 + 20 + 300

PoppaVic, sure, as I said I already written the code with strtol and it works. thank you everybody

So even your algorithim is a bet suspect.

chmod wants a filename, fchmod wants a filehandle. fchmodat wants a (directory handle, name) pair; where it will affect the file named relative to the given directory handle.

do strlen-index then geez

And analogous for other cases

what’s the quick way to to a pow, 8+2 shifted?
pow10
nah that wouldn’t work at all

just do a num = num*10 + str[index]

sounds weird

my manpage lists ‘allows application to avoid race conditions hat could occur in open()’ and ‘allows implementation of per-thread “current working directory”‘ as rationale

yea that works

ahh

also don’t call strlen every time through the loop …

it’s time for bed in a minute
I KNOW
can people stop calling me on performance issues on a one liner?
over irc?

zid`: I wasn’t calling on performance issues.

pow

premature optimisation?

Ahhh… Yes; I had consdiered threads, but I didn’t think there were any races with what they’d be able to do anyway..

unless you call complaining about the wrong result a performance issue.

no.

the race condition issue and thread issue were not related, not sure if that was clear

Anyway.. If you wanted a real CWD per thread, you can just pass the relevant flag to clone() surely..?

Is it bad practive to use capitol letters in variable names? like int PostSize; ?

Thoughts vary

I think so.

GLib works to the case of variable_name TypeName CONSTANT

I’m happy to use CamelCase for constants, though.

What’s a TypeName?

GTokenType token_type = G_TOKEN_STRING;

I use a mixture of linux style and windows style

In that case, it’s an enum typedef

so something like foo_t x = FooEmpty;

MakeMe_a_Sandwich

Can someone explain how I am meant to cast/convert from a char* to a u_char* ~~~correctly~~~, thanks

uh, why do you want to do that?

zid`: for( ptr = str ; ptr[0] != ‘0′ ; ptr++){num = num * 10 + ptr[0] – ‘0′; }

why 0

because my keyboard is stupid

why always element 0
oh

yes it would seem so (not passing CLONE_FS), but this may not have existed at the time, or on the platforms in mind when the *at() functions were proposed; i’m not familiar with it at all

s/0/\o
s/0/

it’s just *ptr with a crappy style

the name didn’t give it away?

that’s fucking awful tbh

zid`: would it work correctly?

style
is
melting brain
*ptr plz

I’m using net-snmp library, and they define this type to hold the ‘cummunity string’

if(!str) return -ENOCHEESE; for(ptr = str; *ptr != ”; ptr++){ num *= 10 + (*ptr-’0′); }
uif(!str) return -ENOCHEESE; for(ptr = str; *ptr != ”; ptr++){ num *= 10 + (*ptr-’0′); }/u

should PinTail_b use u_char * = (char *) u_char*; or what

shaves and haircuts.. 2 bits?

doesn’t + have higher precedence than *=
?
arithmetic +, I mean

yes

So num *= 10 + 1 != num = num * 10 + 1 ?
as the first evaluates the 10+1 and then the num *= () and the second evaluates the num * 10, then the +1, then the assignment…
Or do I have it wrong?

is there a way to list the symbols in a .a file?

objdump?

nm, also

correct, not equivalent

this is why I usualy go mad with ()

num *= (10 + 1) vs. num = (num * 10) + 1

num += (*ptr – ‘0′) + (num * 10);

strerrer(errno); is correct right?

zid`: swap the += for an =

bedtime soon like I said :p

zid`: that is why I tend to avoid the compound assignment operators.

Hi, anyone has used semaphores?

-o/

\o/

sysv or posix?

_o\ \o/ /o_

please do not do that.

Any

it’s a semphore!

please ask an intelligent question.

sem_ vs semop/semget, etc

I just want to know if they are useful any more…

ah, I thought that it was a man with leaping frenchman syndrome.
sometimes …

afraid they’ve gone out of style?

God damn it. I failed my qualification exam because of it. I want to argue with my department head about that

well, tell him that you’re a moron, but you deserve to pass because this is america.
if that doesn’t work, offer sexual favours.

bzzt – unfair (again)

I think sems are too tricky to use for any non-trivial case

it’s about using the right tool for the job, we’d need to know the context to know if they’re ‘useful’ for that situation

if that doesn’t work, offer sexual favours —– Is this still America?

i see you have experience with the education system

well, not that one.
but I can imagine how it must work.

Hey could you have a look at this homework? I don’t mean to solve it. Just look at it… I think it’s unreasonable to use sems on such a question

i wasn’t implying anything incriminating

threaten to shoot him

could you just ask an intelligent goddamned question?

Hey I think this is intelligent enough. If you think the same with me, you are more intelligent than my department head
Some monkeys are trying to cross a ravine. A single rope traverses the ravine, and monkeys can
cross hand-over-hand. Up to five monkeys can be hanging on the rope at any one time. if there are
more than five, then the rope will break and they will all die. Also, if eastward-moving monkeys
encounter westward moving monkeys, all will fall off and die.

homework is just homework. If they want you to chisel in stone or use an abacus, then that is what you do.

AAAAAAH
PAST RAPE

do not shout.
well, that’s the kind of thing that a semaphore is useful for.

that’s a perfectly reasonable semaphore question

I’m sorry, I’m over a flaky SSH connection that misses characters and I can’t tell if caps lock is on sometimes.

i’m afraid you’re just going to have to start training as a plumber

monkeys are known to use tools such as branches, stones and semaphores

or possibly as a sanitation engineer.

I can do the frogs where you have to get too groups traveling in opposite directions across the stepping stones game

plumbers have to know too much

have you managed to work out how to do that with a counting semaphore?

Is it possible to statically initialise an array of strings in C99?

ssh doesn’t miss characters
tcp doesn’t miss characters

keyboards do

The correct answer will use 4 semaphores for this. I think it’s tricky enough

nonsense — one counting semaphore is sufficient

zid`: latency is extremely high and it’s difficult to tell if the connection has gone down and is missing all characters, or is just being slow

you set the counting semaphore to 5

that’s the incorrect way obviously!

Zhivago, try it. You will get deadlocks

then each monkey takes it down one.

Zhivago, No. That doesn’t work

what about the east/west constraint

no, I won’t.
well, you just use another semaphore to control direction.
that will give you two semaphores.

Try it

umm, what is there to try?

what problem do you see?

what is the function of each of your 4 semaphores?

you use a counting semaphore to control the rope population.
use another to guard the direction of the flow.

semaphore mutex[2] = {1,1};

heh.. the semaphores are a rope-property

semaphore max_on_rope = 5;

bzzzt, semaphores aren’t mutexes

semaphore rope = 1;

no — you start with the rope_capacity semaphore at 5

capacity and _direction_

a monkey entering the rope reduces the rope_capacity by one.

And direction capacity at 1

when it gets off the other side, it increases it by one

Is this okay? static const char * config_parameters[0] = {“bar”,”foo”,0};

nah, you just need the flow to be left or right, and to wait for it to be what you want.
what’s the problem?

well, increment, decrement – either way. But it’s consistent

That’s supposed to be config_parameters[3] not config_parameters[0]

the [0] isn’t… try []

Thanks, just making sure

where’s the deadlock?

Zhivago, There is another requirement…
No monkey should wait unnecessarily

that is just fairness

Zhivago, You have only one monkey on the rope at a time

not even that, it is just progress

no, he has upto 5 monkeys at a time, all going in one direction.

no, you’re supposed to have up to 5 monkeys on the rope.

The rope can support 5 monkeys, so you should enable this

read the goddamned problem description.
no, my solution supports up to 5 monkeys — re-read what I wrote.

if the rope capacity is 0 another monkey may enter the rope, and reduce the capacity as it does so.

your comprehension of Zhivago’s explanation is lacking

Zhivago, Do you have monkeys going in both directions?

no.

if the rope capacity is 5, no monkey may change the flow of the rope.

there is a second semaphore which controls flow.
a monkey that wants to move against the flow must wait for the flow to change.

wait(count); wait(dir); pass(); signal(dir); signal(count); ?

“will the last person to leave detroit turn off the lights, please”

no …
we run each monkey has a process.
s/has/as

no — you start with the rope_capacity semaphore at 5
a monkey entering the rope reduces the rope_capacity by one.

when it gets off the other side, it increases it by one

Is this your whole solution?

You guys are wasting your time

Hey… No this a serious problem of education!

that’s the basis of it — the flow is also simple — just reverse it, and wait for the capacity to be 5

or the rope to empty

same deal.

Zhivago, can you write the complete solution with wait() as reduce semaphore, signal() increase it?

reassess why you failed your qualifier

plumbing is quite well-paid
at least in london, it’s hard to find a good plumber

Raiford, Because I don’t think this problem with worth doing with semaphores

yes, and you can wear your pants as low as you want

I have a better solution

The real challenge is coming up with an efficient flow control if you have a LOT of monkeys queueing.

you can’t test for the rope being empty if you’re limited to those.
if you have those limitations, then you can set up two opposing semaphores — one being rope capacity the other being rope load.
such that they sum to 5

monkies_on_left = 5; monkies_on_right = 5; //pretend to move monkies return 0;

My better solution is write a request server for the monkeys.

build 2 ropes, one for each direction; it will alleviate the counterdirectional monkeyflow traffic jams
what advantages does that provide?

but if we find we only have monkeys going one wa, 50% of our rope capacity is wasted.

The moneys send out request to pass the river, and the apache geronimo server hosting decide which monkey to go

so what? it’s paid for by the government and will create local jobs.

that just wraps the problem up in an oracle.

It’s an opposition safe seat, so the government don’t care…

were these old world or new world monkeys ?

they are american undergraduates

just getting on the rope costs $25,000

Do you still think this problem is trivial?

do you think your solution is better than Zhivago’s?

yeah, it’s pretty trivial.

I haven’t seen any complete solutions yet

it depends on how retarded your semaphores are, but it’s not intrinsically difficult.

Write down your complete solution and I can find a deadlock somewhere

in a normal counting semaphore system you will be able to do something like semaphore_lower(5)
do you even know what a deadlock is?

That’s what semaphores are designed for. The reader/writer problem is its maximum ability
Of course

define a deadlock

you have just betrayed you haven’t a clue about the power of semaphores.

If it’s too tricky to get right, I woundn’t use it
It’s too tricky for anyone to get right
So all solutions are dirty tricks and they might fail unexpectedly

sounds like a solution from someone who doesn’t understand the problem

Why use semaphores? I know its implementation in the linux based hosting services kernel. I know how to disable interrupts and put the processes on the queues. That’s the right way to solve problems. Why wrap it up?

have you ever used a machine with more than one processor?

I just write a scheduler

why use semahpores? Because they are conceptually easy. KISS.

Are semaphores able to handle more than one processor?

erm
can a sink have more than one tap?

Does love conquer all?
you just need to *believe*

can you guys think of a typical C question that could be asked in a job interview

I really don’t know what kind of synchronization is used in a distributed computing

something with pointer arithmetic

hmm..pointer arithmetic ?

depends on how and what you distribute.

No they will ask you Java

For a C job?

for a programmer’s job, they might ask C question htough

using a pair of semaphores, it’s fairly straight-forward.

you have a variable int[10]; &int[0] == 0xdeadbeef. What is &int[1]?

west and east. if you want to change direction, reduce the current direction by 5, the increase the other direction by 5.

no I don’t think you can beat the professor’s solution with four semaphores

i’ve been asked to write some things related to the fibbonacci sequence, string manipulation

pointer arithmetic , *((x + 5) + 1) reffering to x[5][1] ?

well, that’s because you don’t think.

uhh…. that’s not x[5][1]

no, x[5][1] is *(*(x+5)+1)

it isn’t ?

Sorry. I should look at it carefully

oops..

if you’re doing multidimensional arrays, there has to be multiplication somewhere

think about it — reducing west by 5 waits for west to be empty, increasing east by 5 then transfers the capacity to the other direction.
no deadlock is possible.

The point is, there is no operation for “change direction”

I just told you what it is …
current -= 5, other += 5

So every monkey will try to reduce the capacity of the opposite direction by 5?

no, the direction changer will.
your problem has nothing about when the direction should change.

Who is the direction changer?

Couldn’t you use a rope capacity semphore, and a direction semaphore?

doesn’t matter — the simplest approach is to simply change direction when there are no monkeys left on that side.
you could, but it becomes more complicated, since you need to guard the direction changing.

it’s been a long time, I have done all that pointer stuff

so add in a tollgate which switches direction when one bank is exhausted.

whats the answer for deadbeef ?

seems fairly straight-forward to me.

Same condition as your solution, only permissable when capaccity equals 5

Here is your problem. How do you test the rope’s emptiness?

depends on the size of int )

deadbeef + sizeof(int) ?

assuming 32-bit ints, it’s (usually) 0xdeadbeef+4

isn’t it ?

you don’t need to test the rope’s emptiness.

yeah

you just wait until you can acquire all of its capacity …

though ints likely won’t be at 0xdeadbeef because of alignment issues anyway

at which point you know that no-one else is using it.

I meant what will be the actual address after adding 4, I kina suck at hex addition
the carry-overs and all ?

bef3

deadbef3 yeah.

Then what about monkeys in the same direction. They can’t get onto the rope

yeah…
thankx
bbl
got to eat

Humm, my lego robot has semaphores, wonder if they’re hard to use

that’s … because … you’re … changing … direction.

time to build lego monkeys

Oh you mentioned oracle?

you reduce the current direction by 5 capacity, then increase the _other_ direction capacity by 5.

do you mean the Oracle database?

no, I mentioned ‘an oracle’
it’s a CS term.

A black box that knows all

Semaphores must not be too complex, if they can be implemented on an 8bit uC with kilobytes of ram

So do you think wrapping this up in an oracle is ok?

no he doesn’t, he was describing your solution as an oracle

wy is either trolling or hopeless.

I suspect that latter.

No. I’m not. I’m just discussing a question friendly

too much fluoride?

After you decrease the other direction by five. There is a monkey from the other direction. Because the program is the same one, it will try to increase his direction by 5. Then it will get onto the rope.

no.
after I decrease one direction by five, _both_ directions are zero.
think it through.
only one monkey can reduce the direction by five, because there isn’t another five left to reduce it by …

Hi
alright I just want to create a program that can randomize 10 numbers, I don’t know the syntax for it

every monkey will try to do the same thing

good luck with that

Any help?

but they wont’ get on the rope if they fail — they’ll wait to succeed

It’s rand() right?

got any code?

Yeah, not on me.

So the second monkey in the direction will try to increase its direction by 5, then reduce the opposite by 5. It’s stuck there because the opposite direction is already 0

what do you mean by “randomize 10 numbers”? where are the numbers coming from? do you mean the order or the values?

do the monkies have to do with semaphores

yes

That’s a new invention by my professors

semaphores are to keep multiple threads of programs out of the way of each other right?

Is there anything wrong with using oracles?

simians have been using simple tools like semaphores for thousands of years

xsdg, the numbers can come from being inputted

Liquinn, out.

It’s safe to use if things don’t go too fast.

Airer-girl: are you Liquinn? (or was (s)he asking on your behalf?)
I’m sort of confused now )

he was asking on my behalf…

this is why monkeys don’t change the direction.
as I said before, you have a third party control the flow.

Airer-girl: ok; what do you want to randomize?

oracles are a kind of magic.

say number 1-10

Do you think a third party is a good solution?

sure, why not?

Airer-girl: let me put it this way, what’s a sample input and a sample desired output?

think about traffic lights.

Then that’s it. Enough for me to send my department head an argument to let me pass
I guess there is also some decentralized ways to do this in a network

don’t hold your breath on it…

use consensus.
semaphores aren’t very useful across networks.

the best solution is the simplest correct one.

xsdg, say I put in 1-10 and I expect it to come out with 5 numbers (I think we’re getting there)

Thanks a lot! So actually you support me

Airer-girl: what order?

Airer-girl: it migh be easier if you could write the input code and pastebin it.

Airer-girl: there are parts of the problem you’re not specifying, which is why I asked for a sample input _and output_

It’s too hard to get correct with semaphores except trivial ones

hm I don’t have a code unfortunately

Airer-girl: what RobbieAB said would also work

if you could prove that your solution is simpler than your professor’s or Zhivago’s then maybe someone would believe you

no, i think that your teacher should beat you with a stick.

well can anyone write a simple code for me? -:o

and you didn’t mention anything about networks.

Airer-girl: we can _help_ you
Airer-girl: we won’t do it _for_ you

lol
alright :p

Airer-girl: at the moment, though, you haven’t actually described what you want to happen, so start with that

So your solution is increase(direction[0], 5); decrease(direction[1],5); pass(); increase(direction[1], 5); decrease(direction[0],5); ?

Airer-girl: again, a sample input _and a sample output_ would be good

I could write a server on local machine

no, you have it backward.
hmm, I think that he should probably fail you twice.

brb

Airer-girl: http://rafb.net/p/y2zK2128.html
Is that how you plan to get your numbers?

Hey! Please support me

lol RobbieAB _ lol yes hold on

Remember me in #lisp?

hmm, no.

I had something similar to that too but was too embarassed to paste it because it was too short

will %d try to overflow &some_int?

correct me please

but anyone who can’t handle semaphores is necessarily too stupid to handle lisp.

your proposed “solutions” all add complexity without specifying the algorithm.

that was intended as a concept, it won’t even compile as pasted

change_direction(from, to) from -= 5; to += 5
so simple

Airer-girl: and you want to print the numbers out again in a random order?

You may fail the exam too

what’s wrong with it?

lack of headers for one.

random order yes and 10 number input 5 number output

no, I don’t think so

Airer-girl: please
Airer-girl: sample input, sample output

Airer-girl: is the same number allowed be printed 5 timess?

Airer-girl: right now, your problem is underconstrained

or more than once?

What’s change_direction(from, to)? Will every monkey run this function?

I’m trying! _ RobbieAB nope only one number per time

no, monkeys don’t change directions (second repetition).

xsdg, I said hold on _

Then where is change_direction(from, to) caled?
called

just call it when from is empty of monkeys.

Airer-girl: I mean is input[1], input[5], input[3], input[8], input[1] acceptable output?

No. We have only monkeys right?

you have a rope.

or is input[1] only allowed be selected once?

the rope can decide

Are you introducing a thirdparty?
You cheat!

I thought that the population was unconstrained.

There were infinite number of monkeys on both sides

RobbieAB, nope not acceptable

no, there can’t be an infinite number — there may be an indefinite number, though

infinite monkeys?

DOn’t you know you’ve got to shock( the_monkey[0] )

if there is an infinite number, then the solution becomes very simple
the most efficient solution is to just use one semaphore and never change direction.

You think too hard

yes, only allow one-sided travel

very simple, if you have an infinite queue, stop worrying about changing direction, as you get maximum flow by not breaking for a direction change.

you don’t think enough.

you shock that monkey and i’ll call the aspca

exactly

hello, do someone know how to write in the framebuffer without using GDI functions, with a windows driver?

Talk to Mr. Gabriel.

lolz

no monkey-spanking in ##C

1 more thing and I’m happy with this stage of my project yay!

Your solution is not acceptable. Your rope is alive
Your rope is an oracle

how is that different than your solution? your oracle is your rope.

There is also a race condition in your rope code

I’m catching the output of fortune in a pipe, making a string, and dumping that out. I’m getting some character in my output that just looks like a rectangle though

if( me.state == run ) cover( me); if( me.state == fire ) cover( me); if( me.state == knocked( out-the-trees ) me.state = on-my-knees; cover( me, darling.please() );

You have to use a mutex for “detecting there is no monkey on rope” and “change direction”. Moreover, you switch direction everytime when there is no monkey?

How would I find out what that char is so that I can nix them

that’s an invalid state, unless on my and knees are variables or macros

enums!
hee hee

Anyway you don’t have a trivial solution if you don’t have an oracle.

absolute nonsense.
you just aren’t using your brain.
if you don’t want the rope to decide, then you simply hold an election to decide which monkey gets to change direction next.

The monkeys can get onto the rope after the rope thinks there is no monkey on it

wrong.
the monkeys can get onto the rope when there is capacity.

that usually means you’re doing something wrong
where is the character in the output?

at the point that the rope thinks that there is no monkey on it, the capacity in both directions is zero.
try to follow the basic logic …

xsdg, end of each line

(rectangle generally means “unprintable character”)
are you on windows?

xsdg, no, ubuntu

loop thru the chars, and print each char as %02x format string to get the hex value.

how are you printing (it’s probably a newline)

sec, i’ll pastebin the code

RobbieAB, how do I paste a code?

Don’t say I’m not using my brain. I’m not calling you that way

Airer-girl: it is in the /topic for this channel, but you go to http://rafb.net/paste/ and paste it, and then paste the link here.

well, then _use it_.
don’t make claims which are obviously false.

strInput = (char *) malloc (intNumBytes + 1); intInputBytes = getline (&strInput, &intNumBytes, pipInput);

better would be “%c[0x%02X]“

sure.

oh doh thanks

So if intNumBytes is a static hosting number and longer than the string I get that char?

you probably want fgets instead of getline

wtf is getline()?

Gnu extension

unless you’re doing something that actually needs to worry about newlines

It could be easier if your rope moves your monkey directly.

oh, _that_ one

xsdg, not to my knowledge

yeah, use fgets then

xsdg, usage example?

the memory has random shite in it from strInput[intInputBytes] on.

more portable, and will probably do what you want (make sure to print with printf)
man fgets

xsdg, I’m actually writing the output of this with cairo

if you have trouble understanding how to use it after reading the prototype, I’ll help. It’s good to learn how to read them, though

what is cairo?

in a GUI?

how does the rope move the monkey?

xsdg, i had no idea that man “c command x” was an option

The rope is a commanding authority.

no manual entry for fgets
xsdg, no GUI

is there a manpages-dev package in ubuntu?

The rope pwns _you_.

It can put monkeys onto itself

if so, install it

xsdg, I know I don’t have room for something like that right now

!man fgets

stdio.h – fgets(char * restrict str, int size, FILE * restrict stream); – fgets () function reads at most one less than the number of characters specified by size from the given stream and stores them in the string str – http://www.iso-9899.info/man?fgets

well, google is your friend then

peapicker, thank you

and so is candide, apparently )

Any way you proved I’m right

candide rocks.

Thanks and see you!

alright, is there an easy way to get processor info (make, model, speed) in C?

you, sir, are an idiot. quit wasting our time.

so, hwo would you use it?
what OS?

solaris, for now, but I’m trying to come up with a general solution, if possible

looks to be fgets(strInput,intNumBytes,pipInput);

yup

wy, I must admire your determination. you finally deserve that promotion or access to the class

how are you outputting?

or whatever it is

xsdg, what’s the best hosting way to determine the intNumBytes to be read?

Do the problem before you can say anything against me. I bet you can get it right

the only I prob with that feature is that idjits will use it HERE, instead of at /home/

use a loop with any buffer size

what idiocy do you think that you proved right?

fgets will return NULL when you hit EOF

well, after you hit EOF, anyhow

if they are smart, they will /msg to candide and get the answer privately.

you should read http://www.iso-9899.info/man?fgets

are you obsessing about “how big is a string input”?
you mean “polite”, yeah

smart and polite are usually linked together in my mind.

Some incompatible pointer warnings but it compiles, so far so good

investigate the warnings
warnings can sometimes be ignored, but usually they shouldn’t

It works, the unprintable character is still there though

get you read th manpage?
it is the last character, yes?

Not always the last
It shows up following a colon as well

ok, that’s probably a tab after the colon

does it do the same if you print to stdout instead of your ‘cario’ thingy
cairo.

So then if I strip out tab and newline then I’m good ja?
peapicker, it does not

ok. then itereate thru the string replacing ‘\t’ ‘\r’ and ‘\n’ with ‘ ‘

does it happen when you do `fortune | cat`?

xsdg, no but that definitely proves those are tabs
peapicker, http://www.cairographics.org/

for(i=0;istrlen(str);i++) (str[i]20)?str[i]=’ ‘:;

I’ve bad vibes about his unprintable
somehow, it smacks of utf-8 or some other wchar_t stuff

or… for(i=0;istrlen(str);i++) ( !isprint[i] )?str[i]=’ ‘:;

I suppose I could dig through the cairo side of life and find out how to change character encoding…

er, make that for(i=0;istrlen(str);i++) ( !isprint(str[i]) )?str[i]=’ ‘:;
I wouldn’t really use ?: tho

are you sure you get unprintable characters when you print using printf?

xsdg, I’ll check

I just whipped up a short program which echoes stdin and it works fine

xsdg, yeah when I dump it out to the console it works normally

is there a way to make cairo interpret them correctly? what is cairo outputting to?

png lol
Wanna see the output?

sure

http://123pichosting.com/images/9386cairoFortune.png

seriously. loop thru it, use if( !isprint str[loopIndex] ) str[loopIndex]=’ ‘;

peapicker, I do kind of want to keep the tabs though…

for(i=0,limit=strlen(str); ilimit; i++)…

seem cairo doesn’t want to print them.

cairo should have some way to do this

if it prints on the _terminal_, what was the other dest? a file?

Is it possible to make it so that gdb will step through the program as it is running and automatically halt if something overflows like a string overflow?

so you can manually do extra formatting / position to set/align tab stops, ot you can ignore them.

I’m thinking I could replace the tabs with 5 spaces… sloppy but workable…

also, I notice that lines are broken in the right spot. are you sure you’re not getting CRLFs somehow?

eww

no guarantee of vertical tab stop alignment that way, so you defeat the purpose of tabs.

peapicker, the glory of monospace fonts

libc has some environment variable to do that, in recent-enough versions, I think

might as well just treat all the \escapes as spaces

that is, glibc

brb – recycle

twotab)word word
tab)word word

xsdg, how does glibc know if char string[10]; string[155] = ‘d’; happens?

I’ll soon find out

the word after the colon should be vertically aligned.

because it allocated the memory for you

one with 4 spaces after the colon, one with 2.

i see… damnit

How does glibc know what my program writes to?
I don’t thing glibc can do that
glibc can’t examine every time my program reads and writs to memory – only a debugger can do that

well, I’m having trouble finding whatever variable it was I found last time

argv and argc is magic

I believe it exsts, though I indeed might be wrong

I’ve heard it said the sufficiently advanced technology would seem like magic to someone who doesn’t understand it.

yepeprs

lol, give a caveman a working computer with starcraft to play with

bulibuta pasted “simple test” at http://rafb.net/p/H4MGcj45.html (22 lines)

but our world seem like magic, even if we can understaand how it works, we just don’t know what it is

feck?

ignore that, just a test

okay
what bot is this?

“Ogg get club.. Ogg smash bot. Die bot! Die!”

PoppaVic, confirmed in the cairo docs, utf8… what does that mean?

AntiLiberal, use valgrind and have it drop into debugger when it finds one of your bugs.

well, it means that it’s not really ascii – it encompasses ascii but can use “wide” or “escaped” multi-char “tokens”

hmm, which bot are we smashing?

the heck with feck

4/quit

Curs0r, if your text is clean, 7-bit ASCII, it is also legit UTF-8.

Auris-: problem is, none of that is gonna’ matter when something drawing html doesn’t understand tabs and whatnot.

hello everybody

html is supposed to not care about whitespace too much.

unless you eat the oddballs, you assume responsibility for formatting the intended results.

At any rate, this looks like a cairo problem, not a C problem.

Auris-: he’s getting flakey glyphs where there are tabs, etc

I’ll chat them up about it, if it’s not on their todo list then hopefully they’ll think about it.

I’m pretty sure that you’ll not get anything from them. Officially, those are usually “.netdef” glyphs in a truetype font, and the default output for a “.notdef” glyph is the rectangle.
The thing is, most software interprets (one of PoppaVic’s favorite words) the data to be rendered and then decides what things like tab etc mean to it for how they decide to render.
and this is done before the output layer aip.
s/sip/api
s/.netdef/.notdef

peapicker, I follow you but what is your point?

I’ve had to parse ttf to get glyph metric for custom encoding to PDF libraries before to make sure things rendered properly there. The PDF libraries I use don’t handle tab etc either.

remind me never to go back to c++
that kaeden loser is a real dickwad

when you see a tab, you have to decide how to break the tex, move to a new coordinate, and then output.

_DrFeelGood, don’t get worked up over some loser hiding behind the keyboard

hes one of those types thats always telling everyone to take their conversations into channel w/e

peapicker, that may prove to be fairly simple using cairo

which mean, you never actually output the tab.
I’m sure it won’t be too bad.

well, the easy solution is to sub ’space’; otherwise, manage your own lines, columns and tabstops

http://rafb.net/p/vBf8sU92.html as you can see here it already uses coordinates

a tab, rendered in a ttf, SHOULD give you a rectangle if you are using any standard ttf font I’ve ever used.

peapicker, what would be a non-ttf font that I could test that with?

Yep, those apis are not dissimilar to the Adobe libraries I license to generate PDF .

i was reading somewhere that two instances of an app can share their code segment pages, is this true and how does the os determine (or remember for that matter) that the newly launched app has already been loaded, im assuming a byte by byte comparison or is it just a hash, this just doesnt sound right to me…

guys, can the select() cause ~3ms delay in a program ?

use ttf/otf, it is the dominant font technology. Just learnt o deal with tabs as i suggested.
plus, then you won’t necessaryily be locked into monospace fonts.

when I set 0 as timeout, my program sends the packets out in the right time

my other idea for a stupid toy was better lol
oh well, I started this one

peapicker, are you one of those people that code in variable width fonts

eww

if I set 500us as timeout the packets lates constantly about 3ms

zmitya, you realize that sleeps are “x or more”.

this is on linux, the nice value is the normal one (15)

if you need real-time, use a real-time os/kernel.

It’s related to mmap() usually.

hmm, I have not measured it yet

“I see”

The code segment is mapped readonly and executable. When the new program wants to start, it would mmap() it too. But kernel knows it’s already mapped once, so just adjusts the mappings, shares the pages, increments refcounts, etc…

yep.. It’s “application underwear”

ok, so this happens automatically, without the programmer doing that

it sure better
it’s the underlying idea behind elf and .so (shared object and plugins)

It’s a feature provided by the kernel

No, I don’t code in variable font widths… that would make me Dr Evil.

heh
ok what about windows PE format, does it do anything similar?

hey, now – Me & mr bigglesworth would get upset

Oh.. Windows… No idea. I was talking about POSIX-based systems.

yea i know

Hopefully they have something similar there too though
Probably a function almost exactly like mmap only called MapMemoryPagesFromFileRegion or something equally silly

yea, i’m beginning to hate winapi names

iirc, PE is some sorta’ datafile-format – it has no bearing on the sharing

Yes, that’s also true

Auris-: yes, it sleeps more than the timeout value

at least it isn’t HNZsdfjzsfjLJFKPage anymore.

I hated making API calls under VB6, made me want to blow my brains out

especially the capitalization of each of 10 words in the function name, that gets annoying
then you have to put in 10 arguments
half of which you’ll never even really need
and repeat for a half dozen java api hosting calls, just do do some simple task like read a timestamp
heh

values

_DrFeelGood, many many apis are guilty of providing too much info on a lot of things

no
what he is describing is an overly-generic call

On Windows, I believe the equivalent is something like OpenFileMapping() and family.

define overly-generic

too much info is when people talk on IRC about how they just went to the washroom.

whenever you have a func that takes too damned many args – all the time – and you must remember them and their order

get wrapping paper?
put it in a box

by your above comment, I’d just write a macro that stuffed in mickey-mouse I could ignore.

how do I make gdb stop when a certain variable reaches a certian value?

break if cond I think….

thx

Lively bunch in the cairo room

watchpoint

i miss the old yahoo room days…
good place to go to watch people argue

Oh well, for now I guess my next step would be parsing command line options to set the colors and sizing

section 5.1.1 of the docs, break … if cond

/hel;p
/help
wtf

I got some real good reading done last night on fscanf and the like, should be a piece of the cake ja?

depends how you slice the cake, Curs0r. (grin)

assuming that you like your cakes to be garnished with broken glass and rusty nails …

for line-based input i’d use fgets and sscanf instead of fscanf

amen

i like my cakes with that skunky black butter, and powdered sugar that makes your tongue tingle

ty pizza_

fgets is for FILE* and lines/strings; sscanf is r/o on the results. [f]scanf screw around with the damned file at the same stupid time.

pizza_, those will allow for the whole %i, %s, expression matching yes?

yeppers

excellent

read closely on the field-widths/limits, too
also examine the %n fmt

why do people use the term “lexically scoped”
surely it is more correct to say “syntactically scoped”

This could get long, cairo accepts 4 variables for each font object and I have 3 font objects, 2 dimensions, and 2 coordinates, as well as 1 input and 1 output

I would presume the issue is lexicon/lexigraph/context
wtf is this “cairo” bastard supposed to be?

a terrorist?

it’s a drawing library http://www.cairographics.org/

what is a lexigraph?

it would be akin to a vocabulary/wordlist: where the elements only apply w/i the current context

that’s what a lexicon is

what about numbers?

still words
contextually scoped would be ok too in my book
but lexically scoped suggests something more like “blah”
or /* */

the way folks reuse and misuse terms, I honestly can’t tell you better.
it’s like every person on the planet speaks C++ overloads of their own
so, you are “drawing” text?

PoppaVic, more or less

sounds like a pain in the anus

PoppaVic, actually it was ging great until this text thing
the tabs that is

ging?

oooh a typo crucify me lol

drawing is nice. I wish I had a decent drawing-tool again. Forcing foreign-format text into it like this sounds painful. They sounds like they are missing an interface or three.

stupid foreigners and their wacky writing systems

hmm?

Getting you what you paid for.

all that ching chong chang crap, i can’t understand you!
white power!
dave chapelle moment… don’t mind me

it’s not like white people always speak comprehensible languages

yea, sometimes they speak blackanese
like that duct-tape bandit

very often they choose to speak bizarre tongues nobody knows, like Spanish

rofl

ever been to Kentucky? They speak Kentuckian

i live in VA
if that answers your question

well, you can blame over 50% of the planet simply by claiming it was written by women.

sitting next to KY and WV, im surprised its not worse than it is here

You poor bastard

well, hopefully they’ll learn mexican sometime

ho~ hello Zhivago

actually, the cario interface for rendering text in a TTF looks like most libraries that do the same thing. The interface isn’t the probelm as much as your desire that it work differently.

why does it feel like 6pm already (EST)
i seriously checked to see if my clock was right

hello libra

installed Linux.. ^^;

and the weather is depressing.

where are you located?

a bit off-topic, but does anyone know why my autoconf-generated configure script insists on checking for a fortran compiler?

heh

a system without fortran is not worth compiling anything for.

lament, but my program doesn’t use nor need fortran

just add fortran to taste

lol

can I just add some pepper?

sem_init(&foo, 0, 5); /* 5 should be the initial value, but sem_getvalue() returns 0 right after the init, and the program’s behavior suggests only a single thread has access at any one time */

lo, draichis – are you on the yo-yo today too?

No
I’m not feeling very well.

prolly the weather

Perhaps; would explain it..
I’ll be back in a bit — I’m going to lie down

hmmm
anyone familiar with openssl
i’m trying to do something where i send over the public key in plain text and then the server and client can start communicating encryptedly
so ideally, i want to have the key, and encrypt a blob, send over the blob
rather than doing ssl_write
the client side is php, and I dont see any way to handle established ssl connections
other than taking a blob, encrypting it with the key, and then sending it over… so i would have to do that on the daemon side as well

i thought the point of ssl certificates hosting was to provide a relatively transparent encrypted network layer over which applications could send their normal data

one of the provided header files gives me defines like:
#define KEikRgbRed {red=0×80;green=0×00;blue=0×00}
how can I use these defines in my code?

just like any other define

very. carefully.

(sorry, it is blue=0×00;}
There is a class called TRgb, but I don’t know how to convert theses defines to this TRgb instances really
could you give me an example?

class ? what language are you talking abut ?
about*

looks like symbian stuff.

yes it is. I just can’t find an example to put these defines to use

C doesn’t have classes.

symbian is degenerate C++, we deal with C.

that is all fine, but maybe you can think of a ’struct’ example or anything where to use such define?

nope.
only thing that might work with is an enum.

not with ; in it

Auris-: not really, since green and blue are both 0

yeah.

and the ;
yes
I’ll ask our C++ brothers and sisters, see if they see a use for it.
thanks

i’m not quite sure how that would be used clearly

holy shit.. my order got here.. Damned near got a hernia carrying it into the basement.

experimenting with mcdonalds free delivery?

hello

no, some ammo-boxes, a tumbler, etc

I have a quesion, if within a structure I have two unsigned shorts, is it ok to wrap them within an int?

everything was on sale, so… I got excited

why bother?

probably, check if sizeof(struct) == sizeof(unsigned short)*2

well, it could save space no?

use a union.
sorry, a smaller struct

likely you’d save the space of an int, at a cost of slower execution

well, he can unionize and the space is nice, and the access is there.

it’s 12 and 4 matt__
well I don’t know, perhaps it could also have portability issues, there’s what I asked
in any case to mask out a int is that slower?

A hernia would be a bit extreme. I’ve managed to.. fracture? one of my back muscles which I can assure you is bad enough.

it’s probably not worth the bother unless you’re severely space-constrained

Er, PoppaVic

dude, one box was bigger than me folded

I’m on the point where I got it working and would have to improve, however I’m using this within a cache, and it takes a megabyte on memory so if I could take a bit out that would be nice

Mm. s/fracture/strain/. Very annoying, since the emergency-mode programming has all the rest of the muscles cramping at the least provocation.

that’s why I ask

Still, water/weathertight ammo-cans are always useful

Approximately like the way I was carrying around tables, then

I’ve been cramping since yesterday.. Random ouchies.
I think the coughing and sneezing is doing it

it can be done, sure

you think I should do it pizza_?

I was going to mention external structs.. See libpack
screwing with packing can save space, but it costs 2x time.

Ice really helps. I wonder /why/, but I’m not complaining.

http://rafb.net/p/CeMswG32.html

it seems to null the nerves and chill the cramps/pain – I do the same with foot-cramps: slap the foot on cold concrete

This much, I understand, but it apparently also helps the healing process

hmm.. restricted bloodflow?

although prosikito does have a point about caches; if packing his data makes it fit inside L2 cache where it didn’t before, he could improve speed

I know that hotpads and such are trying to increase the flow

Possibly there’s a tradeoff between healing and letting you /use/ the muscles, and ice twists the knob

well, worrying about that sorta’ shit is “down the road” – he just needs “make it run”, first

are you specifically trying to reduce memory use or improve speed?

either way, it’s strange

I’m on the point where I got it working and would have to improve…

gotcha

I’m mostly concerned on memory saving, a megabyte of cache seems a lot to me, for what I’m doing

I’ll file in the “turn off scarring, and you heal much better” folder. No legal drugs do that, but there are some plants with a similar effect..

well, _I_ never think about “cache” – that shit is well beneath C
alo-vera, etc

yeah that’s the point of ssl, heh

on a system with virtual memory, as long as an app *fits* in memory it matters much less how much memory it thinks it’s using
yeah, so why are you talking about doing SSL’s job for it manually over SSL?

so what you think I should do?

Yes, although that mostly reduces inflammation and restricts immune response locally. (Which is good, for allergic reactions)

because php doesn’t seem to support a regular ssl stream
i’ll have to look more into it, maybe there is a module somewhere

At most you’ll win 10-20% by this sort of fine-tuning. Is it worth the work?

also healing-rate/scaring

nothing wrong with playing around; btw, are you familiar with alignment?

not too much either, I’ve seen how it’s done, but no more

Consider getting a 20% faster CPU instead. Alternately, using a different /algorithm/ can win you far, far more than that.

if it’s going to be a 20% yes Baughn, I’d say
I can’t get it any faster

because, for example, on a 16-bit machine your structure may benefit from having the ‘ch’ member to the end of the struct

a, well I can put it there, didn’t know that

Instead of blindly screwing with the struct, look at the algorithms.

A C compile host isn’t allowed to reorder the elements of a struct, so it’ll insert padding instead to get word-alignment. You can usually reorder it yourself and get rid of the padding without losing speed; sorting them in decreasing order of size is a good start.

A lot of times a struct is overblown, and the funcs using it are sloppy.

a ok thanks, didn’t know that trick

Well, you do need to understand why it works, but.. it might help.

yes I guess so, I’m already looking for it

(FYI, it works because alignment requirements tend to be orders of two)
s/orders/exponents/

yeah, and usually predicated on the size of an int, and then worse as required.

and analyze your algorithms and make sure you replace O(N^2) with O(N*log(N)) where possible, skip work or cache it where sensible, etc.
so what’s the app?
oh, but first, of course, you should make sure that it’s slow enough to warrant spending the time doing this :/

good luck
even profiling is only relative

it’s for doing driving tests within a console (showing images and animations within xterms and like), and I’m using an nfa for driving keyboard input, some entries on the cache, parser’s symbol table,etc
well not only driving theory tests

Better to simply look at the working app and see if there is a misconception somwhere.

not it isn’t any slower actually
but it might be at startup time if there a considerable ammount of documents to cache

startup time doesn’t matter much unless your app gets restarted or crashes often :/

cgi and system, popen, etc

Permalink Comments off

Im trying to set up my gentoo system to use qingy instead of agetty I changed my -etc-inittab to refer to qingy

October 31, 2007 at 2:17 pm · Filed under Tomcat

If I had a raid on previous install, how can I use mdadm to re-ensemble that raid? Havn’t touched any discs or anything!?

you’ll have to limit it though

mdadm –assemble /dev/md0 /dev/hda1 /dev/hdb1

As that would math 1160 for exemple
*match

also, man mdadm

thanks a bunch!

No module named libxml2″ as an error… libxml2 is emerged with the python use flag as stated in the bugs…

How come I get cpu check errors when having one of my discs connected on my installment, but they all work fine with livecd?
What to do?

that is an expat problem most likely, check the topic

i dont have the old expat.. i just made a fresh install and everything worked fine… i know the expat problem, but this?
revdep-rebuild says everything is fine

ok now i’m getting really annoyed
/usr/lib64/libexpat.la: No such file or directory
at least it’s a new error now

*g*

what xorgconf-thingie generates the best config? any recommendations?

X -configure

that one generates something very short… it doesn’t even mention my monitor etc

I always use xorgconfig and edit as necesarry

ah, okay

any one want this pos dell sitting next to me
still giving me network problems lol

hmmm this expat command installs thousends of 32bit libs on my 64bit system…

i have a double block http://nopaste.org/p/aVZsv07i4 strange thing is i dont know how i could remove that. only one of the packages is installed!

send it to colorado, if it runs ill be able to use it

ok, more tips about this expat problem?

got internet to work on it ealyer didnt change any settings now i just rebooted it and now i cant get network to work again

unmerge both

all week without working kde is a bit too much
in fact it’s insane

only one is installed…!

is your card even configured?

have no prob with that…

ifconfig -a

then unmerge this one

magari, yea for some reason it wont load the 3c59x module

how do we know?

live CD or genkernel it works perfect

hmmm ok and after that reemerge ?

did you enable it in your kernel?

yup
tried it as module and built in

i said unmerge..

does dmesg show it being responsive?

yes but i want that one application

ah more tips

its under lspci forgot about dmesg

veah.. unmerge the blocking package, and then try it again

dmesg

good to know my ideas about gtk are still confirmed
“a mess”

| grep eth0 or w/e interface it is

hey all when i try to type in my xterm a utf8 char (or an accented char) it doesn’t show anything, in console it works, and the WM accepts it also, just not the apps in the WM (fex rxvt-unicode)

re compiling the kernel right now

/etc/conf.d/consolefont

why?

it works in the console

dmesg | grep eth0 turns up nothing

ok
does ifconfig -a show an eth0 ?

putting the 3c59x module back as built in

xterm is a bit strange depending keyboard layouts…

it works in gvim, it works in mozilla

lo and sit0

i know you can set something, but i cant recall where

all my apps but not xterm nor rxvt-unicode

lol ok, yeah compile your eth0 into the kernel

ok
thanks

but it shows up under lspci :/

right
its a device
but without a module

ccache rocks when you just change a useflag and reemerge world

hello
which package does libdb2 come with?

lspci will hopefully always show you what devices are connected on the pci bus

ahh okay

just set the db2 useflag… the package will be pulled in automatically if needed

hi people

Segmentation fault

what?

hm, i wanted to install package via asdf-install, but i’ve found out it’s in portage already.

i think if this compile dont work, i am gonna format the drive and start from scratch again lol
i think if this http://www.oxxus.net/java-hosting.htm“>compile dont work, i am gonna format the drive and start from scratch again lol

it should

CLSQL does require it – at least version installed by asdf-install

that sucks dude, sounds like you’ve screwed up things badly

Have you changed /etc/make.profile or /etc/make.conf? If so, what values did you change, what were they and what are they now?

i’ve tried again
and now its comming ok
o.0
computers are temperamental

mine are rarely

snagnever, Some bad memory?

thanks anyway, bye.

only when under intense neglect

so, i’ve just installed gentoo here

snagnever, overclocking?

That’s odd. An intermittant error like that could indicate a hardware fault

bit` probaly bad memory
no overclocking

I reccomend memtest86+

anyone here uses OSS 4?

soon magari

this system runs perfect…. its just this old dell next to me that always has problems heh

so, please, i’ve just isntalled gentoo here… what about now? some suggested emerge –system
is that right?

snagnever, I had some bad memory once. Only two bad bits… they were always on. Took a while to find out. heh

bit` hehehe
i’m experiecing some probl on my nx6600gt agp board

if you can… emerge memtest86+ and boot it

it gets on when it wants
sometimes appears a lot of artifacts

on the monitor?

yes

strange

the pc turns on
but the agp doesnt send image to the monitor

bad gpu or graphics ram

the mainboard says ‘agp failure’
turn-on/turn-off it gets ok

then throw it away…

not too ok
yes i have to buy a new one
but i’d like a pciexpress one
so i’d throw the mainboar away too
the procss too
and so on
– buy a new pc

err pci-e

take a nvidia one theyre much easier to handle with in linux

i could send you a 4MB Pci card, it never gives up running
i could send you a 4MB Pci card, it never gives up running

heh what card moepman?

hey, i’m going to go buy a video card, and was looking at the NVIDIA GeForce 8800GTS, is this a good card?? will it work well in gentoo?

MicMac hehehe

dunno bout working in gentoo but it is a very nice card i say

i am useing a 8800GTX
works great

Can anyone help me with getting acces to my digital cam? probably a usb mass storage media

it said PCI-eXpress 16 (nVidia GeForce 7600GT)

my friend says its an awsome card, and i figure since its nvidia, should work well for me in linux

another thing
are my flags ok? :

all nvidia card work perfect under linux…

breyl is awsom on geforce 8 cards

have you compiled the usb-mass-storage in the kernel?
geforce7 rocks, too

good, thats what i thoight… i made that mistake last time, got an ATI, and now im running windows, cause it worked like shit/im to stupid to make it work

i think i have

language.

ATI cards blow =p

don’t flame, stick to gentoo support mmkay?

…/../include/qgdict.h:60: internal compiler error: in grokdeclarator, at cp/decl.c:6949

bah, it works as root lol

nice!!

what is shown when you type “ls -l /dev/sd*” (send me this as pm)

“-march=pentium4 -pipe -O2 -fomit-frame-pointer”
is that ok?

i got i working, needed to be root, i hate this way gentoo makes user rights

and -j2 for OPTS

you have the strangest errors ive ever seen
add your user to the group plugdev and log out/log in then it should work

disk group too perhaps

look at the safe cflags page at the gentoo-wiki.com
usb mass storage is plugdev

plugdev

it’s disk here

really? all howtos and docs say plugdev and i have a brand new gentoo that also uses plugdev

shall i give a emerge –system now and go sleep?

*shrug* maybe something to do with how I set things up
only udev rules I have that mention plugdev are 99-libgphoto2.rules

is disk needed for generally mount? plugdev only isd used for hotplug storage devices
if you are in the plugdev group the storage like cd and sticks are mountet automatically out of gnome…

finally done compiling and now is rebooting
got failed to load 3c95x module
after that says network interface eth0 does not exist, pleave cerify hardware or kernel module
modprobe 3c95x gives me …. module doesnt exist
Module 3c95x not found.

What does lspci (from pciutils) say your card is?

I am having some issues, though I am not sure if it is related to expat. Mostly the problem appears to be “image corruption”, especially when drawing things cairo-related.

when i try to start x, it says “no devices detected”… is there a way to get a more informative error?

01.0c.0 Ethernet controller: 3Com Corporation 3c905C-TX/TX-M [Tornado] (rev 78)

See /var/log/Xorg.0.log – look for lines beginning with (EE) and (WW) – if you can’t figure out what’s wrong, pastebin it and give us the URL and we’ll take a look

thank you. i tried to replace Driver “nv” with “vesa” and then it works. i guess i am missing nvidia-support in the kernel or something… i thought i had it there

dont use the nvidia support in the kernel
use the nvidia-drivers package for best results
its a kernel module
works great

Anybody else had this problem of /dev/hdaX “not found” at boot?

Check the values of VIDEO_CARDS in /etc/make.conf and “emerge -p xorg-server”

AllenJB, 01.0c.0 Ethernet controller: 3Com Corporation 3c905C-TX/TX-M [Tornado] (rev 78)

sometimes

I updated baselayout, and now the thing is complaining that I don’t have UDEV support
even though udev is installed.
how did you fix it?

I forget where the udev guide is

I’m going through it now…

Ensure you run a config file management tool such as “dispatch-conf” or “etc-update” and accept changes to the udev configuration

The module “openvpn” does not exist even after i have the tun module loaded and stuff configured?

okay

I upgraded to udev a long time ago, sorry rhombus

what does dispatch-conf do?

etc-update, just better.

is there anything in particular i must do before i can use “Driver “nv”" in xorg.conf? i have an nvidia card… 8800 GTS

For a detailed description, see “emerge –help –config”

more automated? or better?

both

There are also some good articles of gentoo-wiki.com (an unofficial, user-run site) describing how to use both tool

do I use both or just dispatch-conf?

Different. Personally I find dispatch-conf easier to use

there is no module “openvpn”

Just pick one. They basically do the same thing

that’s the error gentoo is showing me.

gotchya

i was hoping you folks could shed some light

most importantly, dispatch-conf logs all changes so you can revert if something goes wrong

weird

Is dispatch-conf native to gentoo, or do I have to install it?

dispatch is reletivly new?
convenient

Native. It’s part of the portage package

forums; google isn’t turning up anything

fabulous.

how do i put a rootsystem /bin etc /home etc on a mem stick and boot from it like a usb hdd ?
not like livecd or liveusb

I believe it’s newer the etc-update, but it’s been around for several years atleast.

as it makes a tmpfs

ok ive messaged mbot about libxpat and im typing in what it says to do and i get “all ebuilds that could satisfy ‘portage-2.1.2.12′ have been masked

ok thats what I thought

anyone has an idea?

It looks like you’re missing an = sign there

you may have to partition it

humm, DRM permission denied on my users … what group should they be added to to be able to use the opengl stuff?

video

have it in user and have restared X .. still get permission denied

There’s a summary ofthe important groups on the “Finalizing” page of the Installation section of the Gentoo Handbook

yea i saw that hehe my bad

i did already partition and ext2 formatted it

hmmm, don’t know then, sorry#

then install grub

at the moment it doesnt boot but has a livecd sguash.image on it witch is loaded when i load the livecd
but then home isnt writable for some reason and gnome fails thus
its a silly mem stick hack
i need to improve
k thkx chithead

heh I love hacked Gentoo boxes like that.. got a couple myself

anyone can help?

care to tell me more
what did u do /hack ur way into ?

?
just weird configurations that one uses with boxes that shouldn’t be

well i changed the 3com support back to a module
the kernel is compiling now, i guess i will see what happens in a few mins

sorry Cirindius, you said something about openvpn?

yea, for some reason, i’m getting an error that says, The “openvpn” module does not exist

when you run it? what’s the exact error?

That’s the exact error

did compiling it in fix it?

magari, now sure yet still compiling
magari, it wasnt working as buildin so i changed it to a module and seeing if that will work

i can’t start my net.eth0 because the “openvpn” module doesn’t exist..

ok
I would search forums.gentoo.org for you card and see if anyone else had any issues

I’ve used openvn for a while and I’ve never seen it, sorry; but from what I can gather from google, it seems to at least at some point have been related to a kernel issue

Seems like it.
I’ve tried the forums, no help either

what are the variables involved from when it stopped working?

um, i missed out the tun module prior to configuring the openvpn keys and what nots

so i went to recompile the kernel with tun as a module
and that’s what i got

i reinstalled my gentoo to switch form 64 bit to 32 bit and now having problems to get my audigy 2 run. I compiled into the kernel the snd_emu10k1, loaded the module, but amorok / kmixer still cant find a card. alsamixer seems to work

how does it act when tunneling is compiled into the kernel?

same thing.

is “Gentoo Installer LiveCD” have gnome or i should download gnome after install Gentoo?

the latter, you’ll need to ‘download’ gnome after you install portage
that is to say use emerge, a part of portage, to install gnome for you

ok, mzbot isnt responding to the reemerge-deep thing

any way to download it before install Gentoo becuase my internet is slow

well, really you should get a minimal install, then you can use emerge -f to download all of the source for the software you specify (such as gnome). That way you can compile it later.
You need to understand that Gentoo is designed to be as bare as possible initially, but is totally flexible in what you can install.

so I have to restart named if I update the zone files?

i messaged mzbot about faq expat and the suggested command failed on gtk like it said i8t might but the suggested /msg mzbot reemerge-deep gets no response

ok, will Gentoo work softly with my 233MB of RAM and P4 2GHZ?
i will use GNOME

Softly? I’ll put it this way, my main machine until a month ago was a p3 800Mhz
Again, the beauty of Gentoo. I run fluxbox, he *points* uses KDE, you use GNOME: it doesn’t matter.

i am installing gentoo on my p3 1gig 256meg ram comp right now

hmm
Dabbill, what is your Desktop Environment?

SmartUnix, will probly go with KDE if i can get the base system stable
in the past on my big system i have run gnome and KDE
seem to work about the same

XFCE with 2Ghz CPU and 1GB ram… mmm

just depends on the one you like the feel of better

loopback mount. Can I mount more then one FS?

no i mean will GNOME work speed, speed is important for me

fluxbox is faster, heh

ratpoison!

wilymage, yeah i know it well

running gnome on this system right now (4600+ dual core, 2gig ram) with beryl
KDE / Gnome seem to run about the same speed wise from my experance

ok i will try Gentoo , thanks guys

never played with fluxbox

fluxbox i used it for 3 weeks or more

hey there guys, i’ve got a problem that i’ve started a thread about on linuxforums.org that i need some help with
http://www.linuxforums.org/forum/gentoo-linux-help/101617-notorious-xfree86-dri-missing-problem.html

the latest RC version, but i don’t know why it was shutdown X when i was working with Pidgin, so i back to Xfce4.4

if anyone would care to take a look i’d be grateful

anyone has an idea what on earth this openvpn module is?

Cirindius what about it?

openvpn.ko ?

for some reason, i can’t start net.eth0, openvpn as well
all of it is reporting that some openvpn module does not exist

what audio codecs do i need under gentoo ( for mp3 ogg, aac

google; forums turn up nothing i can work on

did you try to install openvpn?

I did

any errors?

no.

system is rebooting

can someone can someone help me get my libxpat problems solved. the mzbot help doesnt work

just that, net.eth0, openvpn isn’t working.

Okay, so my problem seems to be with displaying PNGs. The rendering of PNG images results in a “corrupt” image.

try unmerging openvpn Cirindius

got failed to load 3c95x again

and rebooting?

This happened after deep emerging some packages. Any ideas?

rebooting wont be neccerly _
what packages kambei ?

ah.

In particular, I deep emerged the new banshee. expat has definitely been upgraded.

can’t find service ‘openvpn’ needed by ‘net.br0′;

blerg! my junior dev just pulled my power cord out my computer!

any one got any ideas of what i can do to get my eth0 working

When trying to start gentoo with one hd I get cpu check errors and kernel panic. But when I try with livecd it works just fine. What can I do to make it works on my gentoo system=

hii

compile a good kernel.

can someone help me with libexpat? the mzbot help is broken

how can i manually install distfiles? i downloaded a distfile to /usr/portage/distfiles and emerge still tries to download it itself in the net and fails.

it doesnt work. the suggestion to /mst mzbot remerge-deep gets ZERO response from mzbot

br0 is for the bridge right Cirindius ?

broken?

yes broken
it doesnt work

Skunky, I have a good kernel. It worked for 19 straight days until I got that kernel panic

teach I440r_ expat

expat-2.0.1 is not ABI compatible with earlier versions. Run the following in bash to fix: “emerge -n1 ‘=portage-2.1.2.12′ portage-utils && emerge -qep world | grep –color=never -oFf (scanelf -plRBF ‘%F %n’ | awk ‘/libexpat.so.0/{print $1}’ | xargs qfile -qC | sort -u; echo -e “libtool\ncurl”) | xargs emerge -1″. If a specific package fails, “/msg mzbot remerge-deep” for advice (gtk+/pango are common offenders).

yes

do you really need that bridge now?

“/msg mzbot remerge-deep” for advice (gtk+/pango are common offenders).\
that does NOT work
that gives ZERO response

is that not something from openvpn?

faq remerge-deep

/msg mzbot faq remerge-deep

If, while trying to fix the expat mess, a package won’t build then run the following (substituting “gtk+” with the name of whichever package that failed): “emerge -n portage-utils && grep -vxFf (emerge -qep system) (emerge -qep gtk+) | sed -rne ‘/^\[ebuild/{s/(^| )\[[^]]*\]( |$)//gp}’ | xargs qatom | awk ‘{print $1″/”$2}’ | xargs emerge -1″. Afterwards, run “emerge –resume” to continue.

someone should fix the topic

astorm thats not what it says to do lol. thanks

it’s just missing a command to mzbot.

Ah hahah, ok, so i had a problem with kdelibs not installing , something wrong with qt.. then turns out to be expat.. Well the main thing i had to do was emerge -1 a few things.. libXft was the one that i had to look into a fwe pages of 1 topic on the fourms.. should be on the first post too..

How do I get a list of mounted partitions by UUID?

this expat thing is a pita

doesn’t happen all that often fortunately
and you’ll know how to deal with similar “updates”

meh, took me a few hours to find what to do

non-fluid distros don’t show these updates, because they only do them at major releases

Running revdep-rebuild right now.. sofar so good

hm,.

teach I440r_ about expat

expat-2.0.1 is not ABI compatible with earlier versions. Run the following in bash to fix: “emerge -n1 ‘=portage-2.1.2.12′ portage-utils && emerge -qep world | grep –color=never -oFf (scanelf -plRBF ‘%F %n’ | awk ‘/libexpat.so.0/{print $1}’ | xargs qfile -qC | sort -u; echo -e “libtool\ncurl”) | xargs emerge -1″. If a specific package fails, “/msg mzbot faq remerge-deep” for advice (gtk+/pango are common).

does noone has a suggestion?

thanks for pointing it out

you probably didn’t download all the files
or the checksum didn’t match

about what?

try emerge -pf thepackage

2 am, stay awake or go to

and grab the urls (each line is a separate file, multiple urls in the line are mirrors)

lol yw

how does mirrorselect work? it fails when i run it. (it gives 0 results, after spending forever doing it’s “thing”)

ah, where can i edit this file?

what file?

or the mirrors…

just d/l one of these

they are all broken!

and put in /usr/portage/distfiles
impossible, what package are you trying to merge?

rt73 from overlay sunrise

any idea on the issue i’m having mate?

net-wireless

ah, sunrise
you’re probably out of luck then
maybe the file got removed

you’ll have better results talking to the whole channel

find it, the exact version

yea, because a newer driver version is out, but no new ebuild

possibly.

and check with md5sum that it matches the proper files/digest-…

i found the exact version in archive.org

i think i’ll just clear up the vpn mess
and start from scratch

and downloaded it to distfiles — still not working

Well, so bump the ebuild?

ah ok.

how can i get some diagnostics about my resolution/refresh rate etc in X?

vali, xrandr -q will give you a list of supported ones
Although it seems to be wrong for me
:/

hmm

what means bump?

how can i check what packages i have installed? i don’t mean packages that have been installed and then deleted, but packages that are currently installed

it means cp thatebuild-version.ebuild thatebuild-version+1.ebuild

i need a fricken motocycle

When i try to emerge Storable or some other packages i always get this Error: gcc-config error: Could not run/locate “i486-pc-linux-gnu-gcc”

the md5sum is different and filesize.

so it’s not the same file

yes , openvpn fucked everything up
I’d suggest you remove it

language.

cheers.
i found out why

then try a older version or something
sorry

any one got any ideas of getting my network card to work…. on boot it fails to load 3c95x module

ah, ok

so what was it Cirindius ?

ok andother version from archive.org makes it — thx!

sure it’s not 3c59x Dabbill ?

i had some script for openvpn, followed it from the wiki in /etc/conf.d/net

My nfsd says “mountd[6663]: getfh failed: Function not implemented” in /var/log/messages when trying to mount from a client. What to do?

and apparently, the machine was trying to find the openvpn module specified in it
uand apparently, the machine was trying to find the openvpn module specified in it/u

hehe

i spent like 3 hours figuring that out

i’m so up to my neck here
that i can’t think straight

Sometimes that happends

Pretty much the case, i was contemplating a wipe and i was .. nah

Does anyone in here have hands-on experience with SNMP ? Im wondering if its enough, to receive SNMP messages, to just set up a python app listening on port 161 UDP to receive SNMP messages, or if its more complex than that?
All i want to do is customly receive the raw SNMP messages i might add, i dont need the parsing etc to know what part means what

Appiah, your right …

anyone there, who may help me, please?

\o\

Appiah, on boot it says …. Loading module 3c59x next line says Faild to load 3c59x

http://catb.org/~esr/faqs/smart-questions.html

did you compile it as a module?

Appiah, yes

do you have mountd, portmap, nfsd running on the client side?

I’ve seen people build it into the kernel , then try to load it as a module

Appiah, i have tried it both ways no luck either way

what’s it say in dmesg when you modprobe the module?

showmount -e server shows the exports

=/
you sure your card uses 3c59x ?

shouldn’t my gentoo-sources and linux-headers be the same version?

that’s a start jorvis_

Error inserting 3c59x (/lib/modules/2.6.22-gentoo-r5/kernel/drivers/net/3c59x.ko): Incalid module format

On client site??

no.
yes, on client side.
you didn’t compile the kernel module with the same version of gcc, you used a different tree to build the kernel than the module, or some other bizarreness.

I was trying to compile host something that failed because it couldn’t find config.h. I did emerge linux-headers –pretend and the version was older than my kernel

SkunkyFone, how do i fix that?

I don’t understand why on *client* side. May you explain?

that’s normal.
recompile your kernel

make sure your working directories are clean

how can i change a source file in /var/portage/…/program/work/ and use emerge to install that!?

because, that’s how nfs works.

then compile

normal, but do I need to get the current version to have config.h?

Appiah, what do you mean make the directories clean

no.

makehttp://www.gentoo.org/doc/en/articles/linux-kernel-compiling.xml Dabbill
ops

hm.. every KDE package fails building with the following error : error: #error No known way to implement funopen anyone clued why is that?

http://www.gentoo.org/doc/en/articles/linux-kernel-compiling.xml

thanks

I emerged linux-headers and the compile still fails because config.h couldn’t be found

config.h has been removed in linux 2.6.19

jorvis_ what are you compiling? linux/config.h was renamed to linux/autoconf.h

cisco’s vpn client

for path” and then ” mountd[6663]: getfh failed: Function not implemented”

so it seems to be made for kernel 2.6.19

you rebooted

say what?

i say, you didn’
i say, you didn’t search google at all.

hi all. i got a via rhine onboard grafik board and a via soundkad (using alsa) and i cannot use those drivers directly build in the kernel. only moduls will work

Seems to me, that there is not any nfsd-module on the server. It have the genkernel. Isn’t it build-in the?

an i dont know what i am doing wrong

I’m trying to set up my gentoo system to use qingy instead of agetty; I changed my /etc/inittab to refer to qingy for my tertminals, but when I boot up qingy spits out a bunch of text about its usage and then gives me a countdown until it restarts, and I can’t seem to kill it. Do I need to give qingy some arguments?

1) you have to speak english here. 2) you didn’t google. the first 10 entries when you google for “getfh failed” it tells you several things you might have screwed up.

Sorry wasting your time. Thank you.

exit

Permalink Comments off

Im trying to test some fonts Ive installed xfs xlsfonts terminus-font and Ive updated the xfs config and started

October 31, 2007 at 1:29 pm · Filed under Tomcat

_AxS_, ok. thanks.

quickpkg it too if you want, then you have a really quick undo if something messes up

_AxS_, will do. And all this simply because the ebuild was moved?

i wish i would have writtten ebuilds ..

before you were saying about the gnome-window-decorator that it might be in svn right? http://svn.gnome.org/viewcvs/ — cant find it

note, though — i am _sure_ that if you run ‘fixpackages’ everything will be fine
no, i said in an overlay

ohh _AxS_ I thought they are the same
so how could I look for it?

google, mostly.. or ask in #gentoo-desktop or #gentoo-xeffects, etc..

_AxS_, ok. trying fixpackages first.

/etc/conf.d/apache2 as well

‘fixpackages’ did it for me. i didn’t have to re-emerge apache at all

thx _AxS_

any luck?

1st hit just was searching for the wrong thing Im gonna install the overlay now and take alook if its in it

in what package does net-setup and so on reside?

http://overlays.gentoo.org/proj/gnome @ _AxS_ here I am

sweet
it doesnt. net-setup is for the installer. you want to configure your network in /etc/conf.d/net
see /etc/conf.d/net.example for all the options

ow, oki

what type of config do you have? standard dhcp?

nvm, I’ll config it myself; )

k
gl

I’m over pppoe btw

ah. ok there’s instructions in there for setting that up too

hello all, me noob, i just wanted to know, how do i update portage?
bhello all, me noob, i just wanted to know, how do i update portage?/b

emerge –oneshot portage

emerge –sync for the tree

update the tree or update the software?
(tree = database of packages)

or –sync for updating the the package tree, as Santanabaer said.

ci vediamo piu’ tardi gente!

kerframil, well, then i shall tell you the entire problem, wanna listen?

go for it

Serbia!!!!!!!!

here i go…

take your time man

Hey, when I get dhcp it assigns me a hostname even though I have one defined in /etc/conf.d/net and /etc/hosts what else do I need to do?

you want to not use the one dhcpcd gives you?

_AxS_, i just wanted to intsall the latest testing version of enlightenment 0.17, which is masked, i unmasked all the thing using google help, and then when i emerge, it still is trying to d/ 0.16.8.5 version and not 0.17

did you set one in /etc/conf.d/hostname?

Serbia!!!!!!!!

?
how did you unmask it?

Serbia!!!!!!!!

tnx

stop that please

_AxS_, i copied all the things, to that file package.keywords, from internet

heh

Serbia

lol nicht ending message
nice*

always these provinces with their desire for independence

….ok so i don’t see an e17 in portage.. are you using an overlay?

Serbia

whay???????????

teach Milosh about topic

The main topic of #gentoo is _support_ for Gentoo Linux. If your question isn’t Gentoo-related, or if it’s Gentoo-related, but not support-related, ops may ask you to move the conversation to #gentoo-chat or another more appropriate channel.

_AxS_, helloooooo

…ok so i don’t see an e17 in portage.. are you using an overlay?

what is gentoo

Tanja & Milosh this is not a kiddie chat

whay???????????

I like kids

time to go

yep

_AxS_, what overlay? be clear me noob

care to do the honors?

all yours

heh

lol

Serbia!!!!!!!!

ok. in order for you to emerge e17, you need to have ebuilds for it. there arent any ebuilds for it in the portage tree.

_AxS_, i just run the command emerge-webrsync, what exactly does it do?

Serbia

hehe kojiro

heh

so if you want to use it, you need to get an overlay.

_AxS_, ohh gentoo didnt allow me to d/l lol

why is the word $erbia kinda blacklisted lol?

that updates the portage tree, but that wont help you get e17

ow, now I see lol

because it is spamming and offtopic

Hi I am emerging xorg-x11 on a new machine and it is failing at package bdftopcf. Anyone encounter that before? The error message does not indicate any packages were compiled without any important use flags.

http://packages.gentoo.org/search/?sstring=enlightenment — that shows the versions that are available in portage right now

_AxS_, whats overlay? you mean i have to install e17 through some otherway instead portage??

i think this is a guide
http://gentoo-wiki.com/HOWTO_emerge_e17

Any word, if repeated nonsensically, and in spite of warnings, will result in such a ban.

besides, they were clearly idiots.

yes. follow what feuerdrache posted, that will help

feuerdrache, lol thanks

anytime

_AxS_, thanks man,

np, gl

kxdocker is the shit with gentoo and beryl :-p

Language, please.

but it is

all dockers sucks

is there a way to strip all the gcc messages that come up during an ebuild so one is only left with warnings?

kicker will rule them all :X

kxdocker is awesome

still language

anyone got a PC line Sc5100 soundcard?

it actually uses compositing window managers

yes, that can be done

how so? would i need to grep each log file?

so it doesn’t use the CPU at all, and it supports real transparency

_AxS_, hey plz temme, we have to update tree regularly or what?

Sounds great. So how can we help you today?

KungFuJesus, sticking with kicker is the best decision u can make

feuerdrache, I have MAKEOPTS=”-s” in make.conf

it is usually a good idea yes. once a week or so i would recommend.

thanks

I’m just testing out an IRC client, Ksirc

unless you do not want to update any of your software. if that is the case, then do not update the tree

how does gentoo run on alphas?

#gentoo-chat might be more appropriate for that

quite well

_?AxS_, i just want to ask one more thing, if didnt update it some day “x” on which some new thing is added to portage, then if i try to emerge that thing i will not be able to d/l

the usual way i would assume

whom we’ve to bug regarding a simple package version-bump? actually I’m talking about #186075

in terms of performance

#gentoo-alpha might be a good channel to ask in

yeah, but with the obscure architectures like that, those channels tend to be abandoned
god knows ppc is empty

yes. you need to update the tree before you get the new ebuilds, as it is these ebuilds that ‘emerge’ uses to install stuff

_AxS_, i just want to ask one more thing, if didnt update it some day “x” on which some new thing is added to portage, then if i try to emerge that thing i will not be able to d/l
_AxS_, thanks
_AxS_, sorry that i repeated

hmm, I’ll try it

x) and you will then be able to emerge what you want

_AxS_, lol i got it

does anyone have a clue how can I encode the 2 audio channels of a wav file to 2 separate mp3 files? I used sox till now but it segfaults big time

revdep-rebuild might fix sox

Orveldv, just fix sox will save u some time

as per your original request, unsure. i haven’t played with audio much

Orveldv, I sincerely suspect ecasound can do what you wish

mencoder can do audio right?

_AxS_, hey one more thing, whats an overlay??

_AxS_, i think so

thanks guys but rebuilding sox doesn’t help; actually it segfaults with a preemptible kernel
weird thing

a mini portage tree that sits on top of the official portage tree.. its a way of adding or extending the official tree

hey all
my machine crashed when gcc finish compiling… it definatly didnt install

thanks I’ll have a look at ecasound

hello

using the installer?

using portage

_AxS_, ohh thanks

how can i just install it now without telling portage to compile again?

u lost me

–resume

i did emerge gcc
and the machine crashed right after it said source compiled

ok. if that didn’t complete then you need to do it again

it completed

why did it crash?

overheated
=/

is there a ldap aware password auditing application that anyone knows of ?

by complete, i mean, copied the source files to their proper installation locations

you could probably do something through pam couldn’t u?

provided that the system is using ldap for auth yes…

I make a emerge -eav world and I have 922 packages. At 491 of 922 I have a package that it give me an error at compile. How I can make recompile emerge -eav world from 492 package?

you know of a pam based audit ?

_AxS_, I am looking through my kernel config (make menuconfig) and I cannot find TKIP encryption for wireless.

not directly no, but im sure there is at least one

art4art, don’t use -e

‘/’ and then type TKIP

this is so beat

dmead, it;s for gcc upgrade

you can’t cd to the work directory and do make install?

Which package failed?

its under the wifi networking section, im sure. you’ll have to enable certain other parts before you’ll get that option

howdo you have 900+ packages to reinstall just for a gcc upgrade?

okay.
_AxS_, thank you

if he’s going from gcc-3 to gcc-4 or something, yeah…

Hi, I have a very obscure problem. Maybe related to memory corruption. Makefiles get completely borked with ’strange’ characters.

Wow, didn’t know menuconfig was searchable. Thanks for the tip.

memtest86?

np its very handy

Skinkie, i’m having fs problems too

rej, vmware-modules, but it;s not important, I removed now and I want to make from 491 – 922

does yours get strings like ??????
“???????”

No kidding… I’ve been manually telling people what menus to go through… searching for a config option puts it right there on the screen for them.

sure an option, i did a badblock on the fs
also

i get that too

could anyone tell me why my system is being built for i486 when it should be i686? I’m using the 2007 x86 stage3 and a 2005/2006 boot image

emerge –resume –skipfirst, then remove vmware-modules afterward.

did you get the answer you wanted? emerge –resume will start on the package that failed, and emerge –resume –skipfirst will do the next one (skipping the failed package)

Skinkie, is there anything odd in your dmesg?

nothing

you should have used an i686 stageball to begin with.

that’s why — the x86 stage3 is for i486. you want the i686 stage3

that said, you can use a x86 stageball just fine but changing CHOST after the fact is hihgly problematic.

i’m running 2.6.22-gentoo-r1 on a Celeron

bah ya I see that now … tis what I get for doing crap at 4am :/
thx

gentoo XaeroX chost upgrade

http://www.gentoo.org/doc/en/change-chost.xml

can can you be sure this is FS or MEM?

it doesn’t stop you using whatever CFLAGS you want though. either accept it, or start again with the right stageball.

Skinkie, well i get FS and mem errors

if you’ve invested too much time in the installation process, it might be worth it

and memtest doesn’t show anything wrong

I have compiled a lot of programs without issues…

but generally it’d be easier to start over

Skinkie, so have i

I’m downloading the i686 tarball now and will go back to that point

i think i’ll run memtest86 and see if anything shows up

in your case, you might have fried something… i’d recommend memtest86 for a couple of days.

Skinkie, go fer it… but it takes a while
a couple of days?
that long?

yeah. well, over 24h

wow

rej, _AxS_ I must make an emerge -eav world

Hi there, I’m trying to re-install Gentoo on my server. I’ve downloaded the MinimumCD but when trying to boot from it, I get an “Disk error 80, AX = 4280. I’ve validated the MD5 hash of the file I’ve downloaded and everything seems to be ok…

I noticed it when initially setting up make.conf but didn’t pay attention to it (it was 4am) and I got to the point of the kernel and started to build the utils I would need and noticed it was sayiong i486 ….

yea this machine is over 4 years old now
probably just gonna get a new one
the ram is :
XaeroX, i think you have to start over

i havent seen that before…

hey can someone help me please.. when i want to upgrade system i get this error http://pastebin.com/m62358c68

I pretty much am going to

oh, ure not too far off then. copy the kernel config to /boot (and umount), reformat / and do the stage3..

_AxS_ neither have I. Haven’t had any problems before of this kind. I’ve googled it and found a few other people than me having the problem, but the only suggestion to the problem I’ve seen is checking the integrity of the downloaded file.

I didn’t build the kernel *yet*

_AxS_, gcc upgrade from needs emerge -eav world, and I have 922 packages. From now It compiled 491 but it’s stop it for that error. Now I want to make an emerge -e world from the 491 package to 922, not from 1 to 922

right. so don’t run emerge -eav world, run emrege –resume –skipfirst

i only have 64mb mem… so really for 24h

yeah.. well, you have to let it run for a sufficient enough time to catch errors. they can happen at random. i mean, if it takes only 10 min before you see errors,then ibviously u dont need to keep running it
wgetpaste -i for us?

here is the error code http://www.pastebin.ca/639478 I don’t know what wgetpaste -i was supposed to do.

it sends ‘emerge –info’ to rafb.net/paste and spits out the url

hello all

some know how resolv this problem
* Updating /etc/modprobe.conf by hand … [ !! ]ning: could not generate /etc/modprobe.conf!

paste the url
thats the point of that command

huh?

it said command couldn’t be found. I type wgetpaste -i at the command line
I am probably typing something wrong

oh, u need to emerge wgetpaste first. btw, for that error, could we get a few more lines?

emerge wgetpaste furst
first*

the good stuff is above ‘make[1]: *** [bdftopcf] Error 1′

lol why does everyone react to my “hi” by going offline?

is there an option to check the swap space for errors… or is that far less likely?

‘badblocks’

hehe ok here it is http://www.pastebin.ca/639480

that ‘automatically’ adds the badblocks too?

but if youre getting errors there then you’ll see a bunch of IO errors spat out from the hdd

mmm…
true

Anyone have install tinyerp 4.1.1 on gentoo?

depends on the options. i dont think by default it marks anything

I’ve some issue on the tinyerp init.d

wgetpaste is done compiling and I typed the command but it is still processing

http://cpp.sourceforge.net/?show=38909

thanks…I have not used wgetpaste before

k. well, the error seems to be with ‘libXfont’ lmk that other url and we’ll see why

this is the code i am using for getting the network card name to network interface mapping

wgetpaste is still processing data because i t hasn’t returned the cursor

wow really? i thought it would finish in like, 10sec.
your net connection is still up, right?

really…I’ll try it again
I started it again and am waiting for results

cruxeternus:you there

if its easier (ie, this takes too long), pastebin.ca’ing ‘emerge –info’ will do what i want too

is it supposed to open a web browser window because right now I will mention that I am ssh’ed into the new machine and it just has the bare minimum on it. I don’t have X compiled yet.

no, it’ll just spit out a url

here is emerge –info http://www.pastebin.ca/639487

hello
erreur de configuration – �l�ment ��GETPASS_ASTERISKS�� inconnu (avertissez l’administrateur)

LOL

_AxS_, whats layman?

huh. ok everything there looks good. try ‘emerge -1 libXfont’ and then ‘emerge –resume’
a tool that lets you manage overlays easily

okay giving it a shot

hah, easily
i tried it for kde4/cvs

_AxS_,

easier than doing it all by hand, yeah

anybody has a better idea for this, the network interface name to network card mapping
how can i do that
http://cpp.sourceforge.net/?show=38909

you have to do this in code?
or u just want to know what they are?

what’s a good set of USE flags for a laptop I intend to be doing some coding on and will be using XFCE?
never can figure out which flags to set

i am makin a map of interrupts of lspci and ifconfig output
through code

standard ones, probably. don’t make ‘doc’ global, is all i recommend

_AxS_, rej thanks, it works

cool

try an emerge –verbose, and pick and choose

I am rebuilding libXfont right now

i first use lspci -v and then ifconfig -a

why shouldn’t one use doc globally?

Did you see our comments above? DrChandra suggested the excellent tool ‘lshw’ (`emerge lshw`) which maps ethX to PCI slots.

it makes circular dependencies that are a royal PITA to get around.

anyone can help me with start-stop-daemon in the init.d?

coollinux, look at procinfo :p

_AxS_, lol, plz, how do i search for file at command line? plz

http://rafb.net/p/cIcIPY43.html

$ man find

….what are you searching for?

I can get them work on CLI but script don’t work..

thanks i never knew that. I wanted the python documentation, and thought i’d be smart to have all my other progs’ docs. I’ll remove the flag

ust any file?

any help?

best to use on a per-package basis

yeah, go to #gentoo-dev-help, i’ll help u there

I’ll remember that

what to look in procinfo, btw

thanks _AxS_ and feuerdrache

that also maps interrupts/hardware
already
:p

How does the sa-learn work? I did: sa-learn –sync –ham [messagefile] … then I sent that same message and it was again detected as spam … what am I doing wrong?

_AxS_, Thank, I’ll be there

i gave, find gaim, it says no such file or directory

$ find / -name “gaim”

how can i see which USE flags were used to compile a package? thanks

garette, try “which”

dcoats, it works I am using report_safe 0
dcoats, How does the sa-learn work? I did: sa-learn –sync –ham [messagefile] … then I sent that same message and it was again detected as private and secure spam free hosting services … what am I doing wrong?

or if you have slocate setup “locate”

thanks everyone

or go nuts with $ du / | grep “gaim”

o.O

is there a somewhat comprehensive list of USE flags anywhere?

/wc

www.gentoo-portage.com

a few yes.

thanks

can i have specific code for this

/usr/portage/profiles/use.*

lshw

gentoo-portage.com, gentoo-wiki…

okay same error again

bootstrap_dude, rofl

ok, apparently I hit an RTFM spot

you didn’t ask for a specific

erg.

but a comprehensive list

and view the output (2) cat /var/db/category/package-version/USE

http://www.gentoo.org/dyn/use-index.xml
i keep it bookmarked in firefox

I’ll figure something else out then. Thanks for the help!

thanks kerframil

sorry i’m not of more help. check bugs.gentoo.org — this has probably been posted there

bear in mind that, if you use the first approach, any flags that have an asterisk next to them indicate that they have changed (so the currently effective state of the flag will be the opposite of what is shown for those)

Just `emerge lshw` and you have it.

hi, what about the portage bug that it’s blocking the KDE stabilization?

thanks feuerdrache that’s what I was looking for. I remember having seen this before, I’ll mail it to myself and bookmark it later

cool

it’s being worked on, but there are a few other things slowing it down anyway

yah i have it
but i am writing an application
which should show network card listing with interfaces of its own

ok, I hope they solves soon!

be patient, young grasshopper

I’ll be, thanks you

you can look through the source of lshw and see how they do it.. ?

hmm, ok
i wil try with that
if you guys have any other idea, do let me know
this thing is really bothering me for long, now

m0t3jl:http://www.faisal.com/docs/salearn.xhtml

During install it says I should upgrade portage it says so, which it does. Once done do I need to sync again?

juc0, no

ok, cheers

you should only sync once a day anyway

basically get about 3000 spam messages , dump them in a folder and train spamassassin with them

What does lshw fail to do, that you need to write an application for?

you will be wasting your time doing them one or even a few at a time

i think he wants to do this in his own application, for the purposes of his application

http://spamassassin.apache.org/full/3.0.x/dist/doc/sa-learn.html

dcoats, so it is impossible to tell it to not mark as spam one message that will always look the same?

I thought he was just trying to solve a one-time IRQ collision problem. Not worth writing a program for.

hm.. i have just emerged X and i have noticed my keyboard layout is not correct here . where could i set this

whats the lastest version of kde in gentoo? I emerge –sync and all im seeing is kde 3.5.5

Are you trying to a an IRQ listing for the network interfaces?

but i only need the url, what should my sed rule be like?

yes
i first did an lspci -v

http://packages.gentoo.org, searc for ‘kde-meta’ — it will show you the list of versions availab.e

made a list of network cards with attributes base address, memory and i/o ports

and give you an indication of how you may need to keyword / unmask to install them

then i did ifconfig -a and matched the three thing to ifconfig output

evening

1.0. There will be an “irq” file there, with the IRQ number in

whereever there is a match, i consider that i have a network card name and its network interface

do i have to keep the messages which I already trained or I just have to train them and then I can delete them?

if its one message that will always look the same then whitelist it

Keep in mind that this is not a programming channel.

yes train and delete

yah i kmnow
but gentoo is for programmers

what is the end purpose for this ?

this is a utility i am making

The topic is a bit more specific than that. It’s for people who are having a Gentoo-specific problem.

for configuring networks
yah i know but i am trying to do this on gentoo

dcoats, I can whitelist a message?

I tried the userspace ‘memtester’ now, but it doesn’t find any problems… so i wonder if memtest86+ will

YOu can also find info in /sys/class/net/interface_name

since what you’re doing isn’t gentoo specific (any linux this will work on), you should probably find a linux development channel

user-space memtest isn’t that much of a test, the OS gets in the way

ok…

can I create one sa-learn storage for every user on my system?

then i need to ‘bypass’ the firewall

can you suggest me a good tool like you see on some screenshots to monitor system state (cpu,load,etc..)

anyone know any specific problems with fglrx and Xorg7.2 and which kernel would work best?

yes they will all use the global unless you specify differently in user_prefs

im still losing to fglrx

you’re using the ~arch fglrx right?

so when a user types sa-learn –ham [messagefile] than he us actually using some global file automaticaly?

I run ~arch on a fglrx machine, and I update kernels as fast as Gentoo releases them. They all work.

and yes you can whitelist by address,spf or subject

i think he’s using stable ati-drivers.
my guess, anyhow

yeah, i’ve used most of the fglrx releases.. with most of the kernels
and it still hangs my box after about 10sec

dcoats, ok … this sounds nice…

i tried the latest from atis site too, no diff
and im really at a loss here, dont know what to try next, and i hate hardbooting my computer

i built a new kernel recently,and whenever i try to start up NetworkManager w/ the new kernel
the kernel panics, and the computer locks up hard
have you any suggestions?

and when i start X and run fglrxinfo, it shows all ok, and then the system freezes after a few secs

oh, its hanging your system? i thought it was spitting out an error… are you x86 or amd64?

x86
yeah, i hangs completely

no idea then…
you’re sure its fglrx and not something else that’s causing the issue?

it has to be some xorg.conf setting, since it used to work fine like a few months back

I will create folder Spam-NotSpam where users would put their messages, marked as spam, but not beign spam and there would be a cron job which would sa-learn those messages and move them into users inboxes, so the users would know that the messages are now sa-learned

like, something else in xorg-7.2 ?

well, i dont know, it all started when i accidentaly did deep world emerge with ~x86 enabled on all packages
well, i’ve tried xorg 7.1 and 7.0 too

umm, you didn’t try and undo that did you?

well, i did a new emerge with –newuse and whatnot
and now im using stable sources
now im compiling the kernel once again and hope for the best.. but i doubt that it helps

i’m thinking more about gcc and glibc and the rest of system

_AxS_, layman -f gives me error http://rafb.net/p/i6eEXL70.html

well, glibc and gcc should be stable

not if you emerge -uDN’d with ~x86

gcc 4.1.2
yeah, but i undoed it

can anybody point me to some good channel for linux development

_AxS_, i dont even understand what those errors are abt lol plz help

looks like your networking is down.

_AxS_, i use a proxys server
_AxS_, should i have to do special settings for layman or what?

ahh. man layman, there is probably a way to specify your proxy

and xorg.log shows no errors …so its a bit of a problem

_AxS_, lemme check

nor does dmesg

yeah that’s the problem with a freeze. cause, well, its frozen.

i know
what i mean that i got no errors before, nothing from agp or fglrx..

/etc/layman/layman.cfg see proxy

the screen just freezes. with a few artifacts

also, you will want to adjust your /etc/make.conf and specify your proxy server in there too.
yeah that’s expected, unfortunately. its a hardare issue, software won’t catch anything before it just dies.

so anyone knows the name of the tool in X which display CPU usage, fan speed etc?

i think ill go and get an nvidia card

conky, perhaps?

kgrellm?
or gkrellm

what is better?

dont know, i dont use em

conky

i wish i could get the driver workin so i could play NWN

they both do the same thing, its a preference.. conky writes directly into your desktop, gkrellm is a window you can move around

from behind router, I can access a handful of web pages (all located in the same country) and IRC but not (for example) google.com. From router I can access all the web. Any ideas what’s missing in my firewall?

boot, wish me luck

no idea.. but this would be why i avoid shorewall and other firewall scripts.

thats difficult without knowing what rules you already have, the question may not be “whats missing?” it may be “what needs removing or modifying?”
shorewall was always kinda crazy for me too , i found simple iptables to be great on linux based systems

dcoats, currently all the rules I have are the ones in /usr/share/doc/shorewall/examples/two-interfaces/rules

add log to all your deny rules and watch the log

how is your firewall actually connected to the Internet?
by what medium?

add rules for valid traffic you see being blocked

kerframil, ppp connection

PPPoE?

kerframil, yes

ah. do you know about the MTU clamping issue?

hey, does anyone know which USE flags i have to compile mplayer with to make mplayer resize movies?

1492 or w/e ?
…what do you mean by resize? it depends on your output method (xv or whatever)

what is a good graphical ftp client for linux besides gftp ?

kerframil, don’t know anything about that

i like kasablanca

i thought that just made things slow, it didn’t cut access all together ..?

if your ruleset seems to be correct and you get sporadic results with different sites from clients behind the firewall, that might be the problem. try inserting this rule now: iptables -I FORWARD 1 -p tcp –tcp-flags SYN,RST SYN -j TCPMSS –clamp-mss-to-pmtu
I can tell you from personal experience that it can cause havoc

noice. good to know

then test a client again

_AxS_, well, i know i have to choose my output method, but the problem is i don’t have opengl output method which i used so far on ubuntu

kerframil, that worked!

zsz:

add ‘opengl’ to its use flags and then u will

hmm.. ld : cannot find -lGL what causes this?

kerframil, care to explain what the rule does?

on another note, shorewall is kind of sucky in my view. I use firehol (which actually has a directive for clamping)

anyone use escputil ? I am after the ink level

sure, see: http://lartc.org/howto/lartc.cookbook.mtu-mss.html

AxS_afk, i have opengl there.

oh, well u should be able to use the opengl output method then..

hey AxS_afk

i’ve even reemerged
and opengl is working, i have direct rendering.

and if that’s specified then u should be able to resize whichever way ubuntu does
how goes?

agh, ok i gotta go.

sup?

fun

not sure

it’s OK from your firewall because the firewall is the one to which the relevant interface is attached (and it knows the MTU).

ok how you doin’? /joey
I’m having a problem with my firefox if you are interested

in firehol, you can specify “tcpmss auto” at the beginning of your ruleset. I’m not sure how to handle it in shorewall.

and $ mplayer -vo gl2 foo.avi doesn’t work?

bootstrap_dude, let me check it

ah, specify CLAMPMSS=Yes in shorewall.conf

kerframil, ok, thx

bootstrap_dude, it doesn’t, movie window doesn’t show up.

hmm whatever I do, cant get compiz-fusion to work correctly, when I have the composite activated in the xorg.conf with the needed options it starts with xinerama failing, then … as soon I activate the fusion-icon … screen freezes completly, pc needs to be turned off .. (not to mention several apps of X dont get executed anymore xfwm4, xfdesktop,) what could I do?

BBd przy otwieraniu/inicjalizacji wybranego urzdzenia video (-vo).
it means “Error at opening/initialization chosen video device”

does “$ mplayer -vo help” print the gl methods?

it doesn’t.

what does “$ eselect opengl list” say? try adding yourself to the “video” group

bootstrap_dude, eselect says i’m using xorg-x11 implementation, i should switch it to nvidia, right?

mine ‘eselect opengl list’ was saying xorg-x11 only, but I could set it to ati

Yes.

and then reemerge?

Not sure if you need to, try before you do.
Huh?

yey .. the gentoo box finally got into production !

what’s it doing?

dbus-daemon is running (as startup service) but hald fails to start because /var/run/dbus/system_bus_socket is missing. do i need to start my blackbox session with dbus-launch? or will i have to use gnome, as in the manpage examples?

mplayer -vo help doesn’t show opengl method, i’ll reemerge it.

well .. routing .. mysql db .. file server host .. lots of stuff
eating about 6( of 8 )G of ram ..

give ram :E

Frigolit, will do .. for now .. it’s in testing ..

weird, Checking for OpenGL … no

;D i meant, give me RAM

lol ! .. no no no .. the ram comes packed up with 2 dual core xeon procs .. so .. no can do !

:
i need more RAM for my workstation, it only has 1 GB
should buy, along with more disk

lewl .. my workstation has 4

sweet

the gentoo box has 2 Tb of space .. :”

i hope you run 64 bits if you have 4

so 2 disks?

Drakonen, ofcourse !

mine has 1.3

with regards to RC_NET_STRICT_CHECKING, is there a way that I can make it depend only on one specific interface? IE, I don’t want net services to start if net.eth0 didn’t start, but wlan0 did

DrEeevil, no .. 4×500G
will soon move to 8×500 .. but they’ll be in raid .. so I’ll still have 2 Tb ..

what, no redundancy?

gonna replace my IDE disks with 3×500GB SATA drives :q

if ‘rc-update add xdm default’ add xdm to the default runlevel what is the command for removing it?

I got this lil prob with removing /boot dir, says device or resource busy, tho nothing’s mounted there, it’s jut a plain dir Oo

DrEeevil, well .. not on this gentoo box .. yet !

rc-update del xdm default

vertigoacid, thanks :/

yep

anyone?

explody, wha’cha seein’ at df -h -T
or ‘mount’

as I said, nothing’s mounted there, my root partition is mounted on /

exyan*

that’s all

exyan, check if you have cd’ed into the directory or a child on any running shell, or have it open in a file viewer or any other process is running with a file opened in there.

well .. still try ‘mount | grep boot’

…
I just triple checked

exyan, can’t you check with lsof?

not realy :/
what does lsof do ?

quit makin my client blink at me

exyan, thy lsof | grep /boot
try

explody, lol ! )

hmm distrowatch marks gentoo a dying distribution?

netcraft confirms, gentoo is dying

it shows open file handles. (correct me!)

and netcraft confirms it’s dying

*blink*

damn, too slow :p

what is netcraft?

lol ? who’s dyin ?

gentoo

getting knocked by ubuntu I’d imagine

I’m on minimal cd chrooted, so..
that doesn’t work

nah, distrowatch never had good statistics

that’s not about statistics

are you trying to remove the boot you think you are?

http://distrowatch.com/dwres.php?resource=major

just like everyone else, and just like gentoo did to debian a while back… it will all even out

read their article
scroll down to gentoo linux

I just wanna delete the damn empty dir

I don’t want gentoo to die

DJGummikuh, they’re lyin’ !

coz I forgot to mount the boot partition xD

ubuntu sux !

bleh

I don’t like it .. I never did !

so, everyone here interested in getting gentoo back to the top -

it’s like .. one day it didn’t existed .. and next day .. it’s on top 1 !

exit chroot and delete it from there. bet you’re in that directory outside of the chroot

file bugs, write patches, help out

still busy

reboot then

nvm, I figured it out ;]

emerging vlc (and mplayer as well) want to drag in gcc 3.3.6 for some reason, while I am on 4.1.1, is this sane to do or should I go around it (and how?)

emerge libstdc++-v3

teach _someone_ gcc-3

Some packages, usually pre-compiled binary packages such as mozilla-firefox-bin, are written in C++ and are only compatible with the application binary interface provided by gcc-3.x. If this is the case then you can avoid having to install gcc-3.x entirely by installing libstdc++-v3 beforehand: emerge -1 sys-libs/libstdc++-v3.

woopie thanks

might be a proprietary codec

yeah, doing the damn ms codecs

I am already one, that would be a dupe

it would be something that’s Linux native but binary-only.
so I doubt the MS codecs are to blame.

maybe quicktime or real, I added them all as use flags

I believe it’s been more than a month here

I’m trying to make an easy move to “bleeding edge” .. so far .. 50% of the packages in the system are working great !

haven’t had a problem

let me mess with that a bit too, thanks for the help

rvalles, goodie then ! ..

it’s sad that we still don’t have 4.2.1 tho
it’s been out for weeks already

it doesn’t change the solution. I’m just pointing out what might be the cause.

what was the alternative package to nopaste called ?

and it’s supposed to just fix bugs

wgetpaste

I meant .. 50% of the packages are bleeding edge and working fine

forgot it
ahh thx

well… ACCEPT_KEYWORDS=”~amd64″ didn’t work for me in the end glibc got angry with me

Frigolit, lol I have the bleedin’ glibc too

kerframil, understood, thanks for the help

sys-libs/glibc-2.6 / sys-devel/gcc-4.2.0

then when i removed accept_keywords, half of the packages in my system wanted to downgrade

)
ACCEPT_KEYWORDS=”~amd64″ this I should put in my make.conf .. right ?

it is called bleeding edge for a reason I guess..

hi everybody

_someone_, great stuff too ! )

I have a huge problem while upgrading mysql
emerge tries to update apache-2.2.4-r10 and fails

…

lol .. only 50 packages needing upgrades

how does it fail?

here’s the log: http://pastebin.com/d5b5cf412
a href=”http://pastebin.com/d5b5cf412″http://pastebin.com/d5b5cf412/a

… oh goodie !

I always wanted to do that ! just never had the courage !

hmm, that is weird

i know
I followed the mysql-upgrade guide, and did a revdep-rebuild …

I see no error message, so it’s libtool failing as far as I can tell

so I have to do what?

emerge -upND system — any updates there?

anyone remember the tool to use .nzb files (download from nntp) ?

weird

hi all

DrEeevil:

what profile are you?

is it possible to make net services depend on a specific interface being up, rather than just any, lo, or none?

default-linux/amd64/2006.1/server

you can, but i believe you need to edit the init.d script directly

http://pastebin.com/d15bd5ed6

hm.

that is my listing of kde components

this would be the init script for each service, yes?

seems i have two of everything.

2 livers too?

sounds not that good …

yeah. the other thing you can do is make it require ‘all’ devices are up, and then just set the default config on the other devices to ‘null’

50% of the upgrade done .. got fine

So who would recommend emerge –update –deep –newuse world against -03? :-0

I have been installing gentoo ( with stage4 for ps3 64bit userland ) and when everything finished, I reboot and I get kboot prompt, but I can’t go further. I tried with gentoo and others, but can’t… I am restarting with minimal install CD now, any help/clues on getting it to work ? ( as a side note: I user reiserfs, but the kernel in that stage4 has reiserfs support if I am not wrong… )

OH, i know nothing about apache-2.2

thanks

meh. it’s really only netmount/portmap that I’m worrid about, my LAN NFS mounts take forever to timeout when i’m away from home. easier to mess with those two init scripts

*nod*

i’ve actually made my nfs nounts noauto and then have code in local.start which mounts them if my system can see the server’s mac address on the network

does a ois ebuild exist?

niriven, http://www.gentoo.org/doc/en/gcc-optimization.xml

is there any way I can make programs stop compiling locales for other languages except en ?

/etc/locale.gen

http://bugs.gentoo.org/buglist.cgi?bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&emailtype1=substring&emailassigned_to1=1&emailtype2=substring&emailreporter2=1&bugidtype=include&chfieldto=Now&short_desc=ois&short_desc_type=substring&long_desc_type=substring&bug_file_loc_type=substring&status_whiteboard_type=substring&keywords_type=anywords&field0-0-0=noop&type0-0-0=noop&cmdtype=doit&order=Reuse+same+sort

en_US ISO-8859-1

+as+last+time&sourceid=Mozilla-search

en_US.UTF-8 UTF-8

sheroc, I do not know enough about that stage4, but does it have a seperate boot partition on a different file system ? if not reiser4 would not be grubbable

this is all I have put there ..

BlackBishop, to remove already existing unrequired locales `emerge localepurge`

sorry for the link

http://xrl.us

Speen, the link looks pretty )

it doesn’t use grub for booting

Speen, use xrl.us

SORRY!

only 3 lines long ! )

sheroc, ok, again, I know nothing about those leet gadget builds…

hehe

is that link complete?

hang on

thanks anyway

what are u doing again?

not following, I wondered about a problem someone had thats all

try this: http://xrl.us/3mmw

ah ok

any other idea bout my apache problem?

sorry, no.. apache 2.2 is ~arch right?

Speen, http://bugs.gentoo.org/attachment.cgi?id=126004 and what should I do? download that and use emerge?

yes

it could be the ebuild just isnt ready yet

try older version

one more: revdep-rebuild tries to remerge net-www/apache-2.2.4-r7 any idea how to get rid of this?
i’ll give a try

gl!

does RPATH also affect dlopen or only when the loader loads the binary and resolves the libraries?

thx

Speen:

they changed the portage category for apache, didn’t they?

yup
www-server now I think

but why?

+s there

anyone here use escputil ?

no idea

i guess this is my problem …

I guess a sincere efford to clean up the tree

crap, they messed up my system … revdep-rebuild tries to merge net-www/apache ….
ucrap, they messed up my system … revdep-rebuild tries to merge net-www/apache …./u

if it tryes ot

kinda makes sense but I never used apache on gentoo here, so I have no idea how that connects

where are the colors specified for bash ?

it needs to

are you still there?

yes

/etc/DIR_COLORS ?

I’ll think about it

5, sector 135064 does anybody know what is this error means

What did you try to do?

when i ssh to a redhat box the folder and file colors are not the same as my local system

i seem to have two version of kde pkgs installed, how do I remove older pkgs?

try this http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=3&chap=5

ioc0: WARNING – IOCStatus(0×0006): Insufficient Resources
bootstrap_dude the server hangs when it that time

Well ofcourse, It’s another system, copy /etc/DIR_COLORS to ~/.dir_colors on the remote system

Hello

faulty hardware?

can someone tell me why you should recompile your entire system when upgrading to the new gcc?

read the guide on that ?

yeah, if it just started, and wasn’t a problem before (ie faulty drivers?), then I’d definitely say hardware issues. grab your HD manufacturer’s diagnostic utility

gcc 4.*? There’s a Guide you probablt should read on that

http://www.gentoo.org/doc/en/gcc-upgrading.xml

moving from minor gcc versions
not major

minor should not matter, as explained in the guide I think

I would most likely still upgrade my system, not world, but that is just me
(and kernel)

I’m trying to test some fonts. I’ve installed xfs, xlsfonts, terminus-font and I’ve updated the xfs config and started it. For some reason xlsfonts doesn’t list terminus. I looked in the kde font manager and its not there either. I noticed one called Bitstream was there, but it also isn’t listed by xlsfonts. Does anyone know what I might be doing wrong for xlsfonts not to see them?

bootstrap_dude, vertigoacid thx .. but its only happens on specific time period

so, it’s intermittent. still need to make sure you’re drive isn’t dying

oki .. done .. bleedin’ edge ! .. up and working
for about 200 packages or so ..

see you tomorrow

221 to be more exact

Isn’t terminus a console font? Not sure how X handels them, but make sure X11 loads all the font render ext., i.e speedo,type1,xtt

i did emerge gcc

Source compiled.

sort of hanging there =/

top it and check if its active
I bet it is

console and x. I’ll check the x config, thanks.

(top as in top, not stop, had to be clear)

Is there anyway for x to read changes to xorg.conf without a restart?

wtf ?!

i’m running gentoo 2.6.16-hardened and want to setup ntp so that the server clock is correct. where can i find some good docs on this?

dmead, could be that scanelf is running

hmm

Nope

Speen, heh ?

hey, I just installed gentoo.. not sure what i’ve done wrong. Grubs working ok, just wanted to edit something, but /boot is empty

it’s not active, actually

nothin’

weird..

did you run mkfontdir and mkfontscale in the directory where terminus was added?

Speen, just noticed the one letter difference

You probably need to mount /boot

Is it mounted?

how do i issue one command to make all git-repository packages within portage to update themselves and recompile?

dmead, give it time, gcc is never fast, not here anyway, and some compile screens are deceiving too

oh, shouldn’t it have done that automatically?

No I didn’t, but there are also a bunch of the font directories commented out in xorg.conf

Not normally. The handbook suggests making it ‘noauto’ in fstab.

wouldn’t an emerge have done that automatically?

linux1, http://gentoo-wiki.com/NTP maybe?

juc0, depends on your fstab if it is auto or not

probably. is /usr/share/fonts/terminus in your xorg.conf?

it is now ;]

mount /dev/hda2 /boot – still appears to be empty. and mount -l shows: /dev/hda2 on /boot type ext2

having /boot not mounted by default is the norm

oh no, sorted it
cheers guys

then an X restart should do it. xlsfonts *does* list several terminus fonts on mine

Nice work

so do I need to edit my fstab to make this auto?

nah

remove ‘noauto’

I think that’s what the “X” USE-flag on fonts do

no need to mount /boot unless you run grub or something

I prefer to leave it ‘noauto’, so I don’t accidentally overwrite it.

ah, makes sense

hi. i’m trying to emerge kde, and I got dev-php5/pecl-zip blocking. any idea?

you might look in ‘xfontsel’, too. look for terminus under ‘fmly’

And you can just type ‘mount /boot’ next time, and as long as it’s in your fstab, it will mount correctly. So it’s not that hard to do when you need to edit files in /boot.

0 sandbox

juc0, just remember a grub upgrade, kernel recompile, config change in grub.conf will need some manual action

oh right, good to know cruxeternus

hey guys how can i get my domainname set ? dns_domain_lo doenst seem to work

ah yea, makes sense

the kernel recognizes, the node is created under /dev/input, the xorg.conf ServerLayout and the InputDevice seems ok. What else should I check please?

I tried that first, didn’t find it. I’m gonna restart x in a second. I’ve modified the config

/etc/conf.d/domainname

dmead, not looking good there… does top show any mem or cpu activity on the gcc ?

my make.conf has “X” in it’s USE flags, but still when I do emerge -vp it shows I need the X flag, why is that?

rickest, isnt that file deprecated ?

what is the problem?

my make.conf has “X” in it’s USE flags, but still when I do emerge -vp it shows I need the X flag, why is that? is there some override somewhere? (package.use dosen’t have anything related)

omercnet, if you selected a profile, you would not need X as a use flag

a profile ?

_someone_, negative

the pointer doesnt react when i move this mouse. what is strange is that i removed the synaptic pad from the server layout but it still works…

I have run this memtest86 thing, it had 8 passes, but no errors at all… so i’m placing my arrows on something else

like i said, u gotta run it for 24h or more before most ppl would consider it to be ’successful’

ooo now tee is running

_someone_, a profile ?

damnit hostname -d says ive no domain set for my boxes

_someone_, omercnet the desktop profiles automatically adds it

omercnet, its in the handbook, but… uuhm.. I cannot link directly to it
omercnet, basically a profile is a set of default use flags for a certain purpose, desktop, server, that ‘work’

are you sure you have “X” and not “x”(lowercase)?

bootstrap_dude, yes

does xorg really need 145 packages Oo?

USE=”X -ipv6 -fortran unicode”

on a fresh install

omercnet, try the good old book (http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?full=1) and search: ‘Choosing the Right Profile’

what exactly is a ‘meta package’?

Yes, X.Org’s “New World Order” of things, it went modular

omercnet, you need to understand what the profile sets, if you dont, you end up with double, and possibly later on conflicting flags

_someone_, gotcha

they are all small. having them separate means if you don’t want all of xorg, you dont need to get it all

omercnet, the command given in that section allows you to see what is set by default for you… X is one, on a desktop, you would never have to set

I noticed the smallness, so I don’t mind

Yeah, alot of them are “protos” that take 2 seconds to emerge…

please is someone here that can help me with my stupid ut2004 server?
I have the problem that my server won’t use the ports I tell it to use and therefore is not pingable

Hello. I trying resize partition with gparted, but when I unmount it, it’s mounted automatically.. How I can umount it?

DJGummikuh:maybe gentoo-server?

well this is offtopic in any gentoo channel

omercnet, beware that setting a profile may be a huge recompile doing a emerge -uDNp world (if it was not set before), and if you did still run below 686, it may even be a chost change (not advisable)

any suggestion? bootstrap_dude?

I just hope someone uses a ut2004 server as well and I can query with him

DJGummikuh:gentoo-offtopic?

too few ppl.. I just search for someone I could then query with

is reiser4 usable yet?

what was the problem again?

I suggjest u google for unreal admin or whatever (don’t remember the correct address)

therefore I need a channel like this with 900+ ppl
I have… their howtos don’t help me

well if you get someone going ” please stick to topic..” then you know why.

sorry. the pb is: the pointer doesnt react when i move this mouse. what is strange is that i removed the synaptic pad from the server layout but it still works…

since I only requested help and am now only arguing with you about posting a request, it’s more likely I’ll hear that for the OT discussion with you
problem is the server is connectable but the ping says N/A

The Gentoo Forums are a good place for user anectodes

a lot of servers are not pingable ;]

DJGummikuh: topic… ( )

yes but ours used to be pingable

bootstrap_dude, cool

gosh did that get some attention lol

probably your xorg.conf not setup correctly

so ur problem is that people can’t ping ur server ip @ dos?

how can I see what is depending on a package ?

my apache problem is solved!

shall we query?

what mouse device are you using?

nah, this would be ON topic

the server is not pingable but that’s another issue

if it is /dev/input/mice then i believe any mouse device -should- still provide signal (including your synaptics)

the ping in the ut browser uses the server port
so this is something entirely different

also, do you have a fn-key combo that switches between pointer devices? that might be part of the problem

anybdoy ?

What’s the best way to install KDE? I want it quite minimal, but do want kdm, konsole and such. Do I want kdebase-meta or should I start with kdebase-startkde? – What even is a ‘meta package’?

missed it… ??

the server port is open and can be used (ppl can connect) but somehow the ping is NA

so people don’t see their latency in-game ;D?

_AxS_ a simple usb mouse. the node that is created when I plug the mouse is /dev/input/mouse0

no ppl don’t see the servers latency in the browsertab

seems so ..

is there a way to see the effective use flags?

ping works on ICMP, the rest is udp / tcp — icmp may be being blocked.

I don’t really understand what u want, if ur saying that people can’t ping as in ping through DOS, then say so

is there a way to see the effective use flags? (after profiling and all)

and the funniest thing is I tell it to use port 7777 but instead it binds to 7787

BlackBishop, I only know the ‘old ways’ doing an emerge -etp word | less and check indentation on depends… but that must be outdated now

not this ping… this ping gets calculated from a connection package to the gameport

you should probably start with kdebase-startkde and go from there

oh.

world… bah

no idea

there must be a better way !

no I meant the “Ping” field in the game browser of ut

anyway, it’s something with ur firewall

…dpes gpm work with that device? ie, is it just an Xorg thing?

which I got no knowledge of ^^

that it can not be pinged by dos (or linux for that matter which is why I read dos as denial of service rather than disk operating system and that way did not understand what you were trying to tell me) which is because of the firewall

yeah that’s way off-topic here.. sry man.

which is why I wanted to query with exyan …

DJGummikuh try a telnet on the right port ?

good night fellow cows

good idea

moooo

Jell-O-Fishi: good night jell-o-fishies

good night fish

muuuu

I’ve just configured my kernel and made it compile, the last lines from the compiler before it stops is a couple of warnings and when I try to copy the bzImage to /boot it seems that the file is not existent. Is that to be understood as the compiler (or I) failed?

/usr/portage/profiles/default-linux/x86/2007.0/desktop – and still it says I don’t have X flag set

moo/
?

omercnet, goto that file and check…. X is there, I am sure…

_someone_, X is also in /etc/make.conf
_someone_, still not working

emerge –info , you dont see an ‘X’ ?

if you have GRUB installed, the kernel should copy itself with $ make install

yes, that would be a good assumption.

I’m about to install kde, but will eventually want to install the nvidia drivers. Should I plan ahead and emerge KDE with any specific use flags?

does the kernel abort compiling at the same position every time?

_AxS_ I didn’t know there was such combo. Which one should I try? About gpm -m /dev/input/mouse0 -t ps2 doesnt do anything

bootstrap_dude I’m following the Gentoog guide, so I don’t think GRUB is present.

i would recommend installing xorg-x11 first, and then kde

$ echo ‘USE=”$USE X”‘/etc/make.conf and try again?

oh, ok

well, -t ps2 is useless because its a usb mouse

_AxS_ Ok thanks, I guess it’s back to configuring and trying again

you better use genkernel for first use, you can create any custom kernel later once your base system is running

_someone_ USE=”X acl a …” – it’s the first flag. still when I compile gtk+ I’m getting an error no X flag is set
bootstrap_dude, didn’t work

omercnet, did you try what _AxS_ said ?

i would just edit /etc/conf.d/gpm and set MOUSEDEV to /dev/input/mouse0 , then /etc/init.d/gpm start

_someone_, if you mean bootstrap_dude, then yes, and it didn’t work

emerge –info , you dont see an ‘X’ ?

ohhhhhhhhhhhhhhhhhhhhhh
jesus

yes ?

I need to recompile cairo with X flag set

reemerge cairo

yeah

ahhh usual pitfall

emerge -uDN world should do it

yea
_AxS_, I was just about to say, that I should rememerge world with the new profile

cairo is small, fast to recompile

epsilon, thanks I didn’t think of that – perhaps I’m still stuck in the narrow way of thinking that comes from using windows all these years

omercnet, just… never ever ever do a -uDNP, I did once, on world… sniff

hehe.. yeah, no capital P

_AxS_, P should get an automatic a

hmm, it doesnt work. restarting gpm give a “[!!]” without any additionnal infos. Is there a log file anywhere?

Just a quick question. Where can I see what version of the kernel I’m playing with?

uname?

uname -a

/var/log/messages maybe

uname -r

Thanks

or dmesg
im guessing though that the error would have to do with the device

hey again

lets try something raw. ‘cat /etc/input/mouse0′ and then move it around

now i got the fglrx driver to work until i move the mouse

sorry, /dev/…

even fgl_glxgears works fine

cool, that’s some progress….

yeah, and now it didnt crash the whole system. just X and the monitor goes to standby

there’s no chance you have a hardware conflict huh?

so i can ssh and reboot
could be.. ill try with software mouse to see if that helps

oh, that’s very strange. killall gpm and then restarting the service worked

monitor goes to standby on starting X….

hm. ok well thats good. so the mouse works in gpm?

probably bad H|V sync values

tho the system did hang when i tried to kill X via ssh.. but thats no biggie..

bootstrap_dude, yup or bad driver support (vesa does not do high resolutions) or bad xorg.conf entirely

_AxS_ yes. It works on another box too. So the problem is probably with xorg.cong i guess?

yep,seems so

Greetings

i
hi

checking for Qt… configure: error: Qt (= Qt 3.3 and 4.0) (headers and libraries) not found.

balty, I would guess your driver choice and kernel support for it

I vaguely recall that xorg will try to EDID scan for the correct ones if you comment(#) the HorizSync line

i cant get my ps/2 mouse to work

anybody have problems with RPM version 4.4.6

rpm? we don’t do rpm’s here….

_someone_ I think so since the device node is created/removed when i plug/unplug it

why not?

rpm: stack smashing attack in function rpmfiNew()

im trying to configure x
and it wont work

why not?

i have no idea why
ps/2 is enabled in the kernel

vandle, usb mouse ?

no
ps/2 mouse

there are 980+ other people on the channel, if you want support, you talk to all of them. I am not your personal answer-bot

vandle, hmz.. nevermind.. then you need to set xorg right, did you check activity in command line ? cat’ing the device to screen and move mouse, checked dmesg for errors ?

It worked
Thanks

_AxS_ the InputDevice section in my xorg.conf is pretty simple. Is there an alternative driver to “mouse”?

evdev

hmm.. seems to work.. now just to try mouse with 3d stuff

you could try “evdev” driver, $ emerge xf86-input-evdev

balty, its best to find out in command line mode what actual device reacts to mouse movements

is this a gentoo machine, and what is the raid card again?

how do I get pppoe-start to start on boot ?

ok, no more help for you.
can’t answer questions.

imps2 maybe?
there are a few different mouse drivers

_someone_ as recommended _AxS_ I changed /etc/conf.d/gpm to set /dev/input/mouse0, restart gpm and it worked

balty, _AxS_ is our hero

fixed my qt problem via export QTDIR=/usr/qt/3

what package had this problem?

wonder why it didnt find it on its own

it might be good to file a bug about it, if the package needs qt3 but doesn’t specify that

_AxS_ I dont understand… On my other box, this mouse works with the “mouse” driver. The xorg.conf is similar

um kde somethin lemmie check emerge log

dunno.. i’m just guessing right now..

did you forget to add mouse to your INPUT_DEVICES line?

balty, usually just cat /dev/mouse0 or mice will show what reacts, if none do, your mouse is dead (usually usb related)

the device is correct, we covered that part…

kde-base/kdelibs-3.5.5-r10

ah I missed it

so how do I get pppoe-start to start on boot ;x?

if that’s an init script, rc-update add pppoe-start default

it’s not

but I don’t know much about pppoe-start

hm whys my emerge log filled with ebuilds
last i checked everythiungs compiled from source

_someone_ there is some output on the /dev/input/mouse0 node. I think there isnt any pb with the kernel/hardware. Seems like a xorg pb. Maybe the xorg.conf but i don’t know. Maybe should I try evdev

so you can’t just make something start on boot Oo? strange

well, you can add it to local.start

yeah, but that’s not the “right” way

coz I just need 2 commands to get my connectivity up
cba messing with conf.d/net etc

Skunky /etc/make.conf seems ok too. I have the INPUT_DEVICES=”keyboard mouse”

Permalink Comments off

« Previous entries Next Page » Next Page »

Archive for October, 2007

im trying to use an old ibook hard drive for a linux installation qtparted is telling me the partition changes

Lo Im trying to get htaccess files to work Ive set AllowOverdie to All in the default file and corrected the DocumentRoot

hmm im trying to half-ass tidy up this old php4 piece of shit site that i dont really care about too much if i

im trying to graspt the subtle difference between a segfault and an invalid page fault am i correct in saying

im trying to graspt the subtle difference between a segfault and an invalid page fault am i correct in saying

Im trying to set up my gentoo system to use qingy instead of agetty I changed my -etc-inittab to refer to qingy

Im trying to test some fonts Ive installed xfs xlsfonts terminus-font and Ive updated the xfs config and started

Categories

Meta

Blog Archives

Search