[03:24] <usual> Anyone happen to run ubuntu-server on a HP server with a SATA raid card?
[03:31] <usual> maybe if I try 2.4 instead of 2.6
[04:26] <xnull> hi i want to install ubuntu as a server
[04:26] <jonesssss> hey
[04:26] <xnull> what does it have? for a server
[04:26] <jonesssss> have you installed ubuntu yet?
[04:26] <xnull> yes
[04:27] <jonesssss> LAMP, linuc, apache mysql and php
[04:27] <jonesssss> linux*
[04:27] <xnull> i want something hardened
[04:27] <jonesssss> such as?
[04:27] <xnull> lets say kernel+grsecurity php+hardening-Patch etc..
[04:27] <jonesssss> lol
[04:27] <jonesssss> well
[04:28] <xnull> does apt knows about that?:)
[04:28] <jonesssss> im not sure what you mean, im guessing you just want security
[04:28] <xnull> yep.
[04:29] <jonesssss> is the pc going to be a perminat server or one you install on your everyday use pc?
[04:30] <xnull> server apache+mailserver
[04:35] <xnull> brb
[05:29] <infinity> xnull: "hardened PHP" is complete useless wank.
[05:29] <infinity> xnull: As for grsecurity, if you want it, you get to compile your own kernel.  No big deal, I tend to compile my own kernels for all my servers too.  *shrug*
[05:30] <xnull> infinity: me 2, but why hardened php sux?
[05:31] <infinity> It changes the behaviour of PHP enough that the few slight benefits you get from it seem to be outweighed by the "it doesn't work how the manual says it does, and 3rd party PHP apps all break" factor.
[05:31] <xnull> hmm
[05:31] <infinity> It also just seems to be a wonderful way for the Hardened PHP folk to get press when there are PHP vulns discovered (which Ubuntu fixes anyway, so...)
[05:33] <infinity> Anyhow, to each their own, I suppose.  I'll never ship the "hardened" patch in Ubuntu, though.
[05:36] <neuralis> personally, i find most php applications to be amazing exercises in human idiocy from a security and code quality point of view. but that's just me.
[05:39] <infinity> Yes, that they are.
[05:39] <infinity> "Input validation, what's that?"
[05:40] <infinity> But it annoys me, as the PHP maintainer, that PHP "programmers" blame the language for their lousy programming, rather than blaming themselves.
[05:40] <infinity> 99% of our security releases for PHP are programmer error (input validation being a big one).
[05:40] <infinity> We don't do glibc security releases to build in input validation into every function you may abuse there, but apparently, PHP is supposed to hold your hand.
[05:56] <neuralis> well, it's a non-trivial issue. php is a language ghetto, and it lowered the bound to entry for programming to an IQ of 70. then a lot of people with neighboring IQs went and started programming in it, which promptly turned away the vast majority of smart people who wanted to use the language.
[05:58] <neuralis> nowadays, php has very strong community value -- it keeps most of these people away from other languages.

[05:59] <infinity> Keeping idiots away from Perl and Python is a noble goal.
[06:00] <infinity> The real shame is that PHP /is/ useful, despite the thousand of people who make it look like it isn't.
[06:00] <neuralis> aye.
[06:00] <ajmitch> I've seen some good things written with php
[06:00] <ajmitch> & some shockers
[06:01] <infinity> ajmitch: I try not to admit to it often, but I can bang out rapid prototyping in PHP about twice as fast as Perl (and Perl about twice as fast as Python)
[06:01] <infinity> I full admit that the latter is due to a lack of experience in Python.
[06:01] <infinity> But the former is just because PHP really is *that* quick for prototyping.
[06:01] <neuralis> ajmitch: a good hacker will write good code regardless of the language, so that there exists some good stuff in PHP isn't surprising.
[06:02] <neuralis> infinity: for people who know the value of using the right tool for the job, PHP has some excellent uses.
[06:04] <neuralis> oh, and you can definitely get to the point of producing python as quickly as perl. i'm a convert after reaching stage 6 of nat torkington's seven stages of a perl programmer.
[06:07] <infinity> neuralis: I expect that one day I'll be proficient in Python, but currently it's a painful affair compared to Perl.
[06:08] <infinity> With PHP, though, the great this is that half the time when I want to do Radnom Thing X, I realise it's already been implemented in C, and it's literally a single function call to make it go in PHP.
[06:09] <infinity> In Perl or Python, that tends to translate to at least 5-10 lines of code, even if the functionality exists in an extension, and many, many more if it doesn't.
[06:10] <infinity> PHP's "make it all C" approach tends to lead to the unfortunate acceptance by idiots, but it also makes it very, very quick to "get shit done".. A horrible double-edged sword.
[06:10] <neuralis> that it is.
[06:10] <neuralis> i'm trying to see if i still have some of my more amusing perl snippets around
[06:16] <neuralis> oh, here's a fun one: http://pastebin.com/766666 is a full-blown interpreter for an obscure flavor of simple assembly, with the target being brevity. the next shortest solution was about 5 times the length :)
[06:17] <infinity> And that's precisely why most people hate Perl. :)
[06:17] <infinity> Mine tend to aim for readability.
[06:19] <neuralis> infinity: i wrote ~100kloc of relatively pretty, easy-to-maintain perl code for "serious" uses. this wasn't one of those :)
[06:19] <infinity> Obviously not. :)
[06:19] <infinity> For maximum unreadability, write a brainfuck interpreter in Perl.  Even if you TRY to make it readable, it won't be.
[06:19] <neuralis> haha i actually never tried that
[06:20] <infinity> (And if you try to make it unreadable, it'll be hell)
[06:20] <neuralis> didn't a recent ioccc go to a brainfuck interpreter?
[06:21] <xnull> I installed qemu .. how do I start him ?
[06:22] <neuralis> xnull: you need an image to start it with.
[06:22] <xnull> neuralis: i have.. but I don`t have a meniu or something ?
[06:25] <infinity> neuralis: I intended at one point to write a brainfuck variant (my current favourite is "whitespace") calles "escapes", which would use ',",\ etc as the keywords.  Wouldn't make the code any less readable than brainfuck, but it would make interpreters nearly impossible to understand.
[06:25] <neuralis> xnull: no menu. qemu -hda <imgfile>, man qemu for detailed options
[06:26] <neuralis> infinity: http://laptop.org/~krstic/timwi-bf.c (2005 winner by arne heizmann)
[06:26] <infinity> Oh, that's pretty.
[06:27] <neuralis> cracks me up.
[06:28] <neuralis> infinity: you should write an escapes spec for edgy.
[06:29] <infinity> Only if the spec includes rewriting core infrastructure in Escape.
[06:29] <maswan> a "port all init.d scripts to escapes" spec?
[06:29] <infinity> Ed Zachary. :)
[06:29] <maswan> :)
[06:30] <neuralis> infinity: why the hell not? linspire is switching everything to haskell, and what better way to one-up them.
[10:31] <ajmitch> infinity: what interesting -server stuff do you have ideas about?
[10:31] <infinity> I have no ideas, cause I've been flat out on "closing dapper" and "opening edgy".
[10:31] <infinity> \o/
[10:31] <infinity> Haven't had the time to be creative.
[10:31] <infinity> neuralis: How about you?
[10:31] <ajmitch> right :)
[10:32] <ajmitch> currently I've just got the network auth & am hopeful about selinux
[10:33] <ajmitch> I'll try & dream up new ideas by tomorrow, though I won't be there to discuss them in person
[10:34] <infinity> I'd like to see some SEL stuff, but I'm doubting that "good end-to-end SELinux policies" is a realistic goal for edgy.
[10:34] <infinity> For edgy+1, it's totally doable, because the SELinux effort in Debian is awaking from its slumber as well.
[10:35] <ajmitch> I think it's fairly likely to get some in for edgy - comprehensive coverage isn't the initial goal
[10:35] <ajmitch> basic tools & a modular reference policy that covers common cases will help
[10:35] <infinity> I'm all for trying for "something less crap than what we have".
[10:35] <infinity> Bordering on "useable", even.
[10:36] <infinity> Are you in contact much with Russel?
[10:36] <ajmitch> manoj has packaged up the latest reference policy for sid
[10:36] <ajmitch> I talk to him form time to time
[10:36] <infinity> He's left RedHat recently and is back on the Debian/SELinux warpath.
[10:36] <ajmitch> yes, I saw that
[10:36] <infinity> His goal is to do for Debian what he did for Fedora.
[10:36] <ajmitch> which is great news for us
[10:36] <ajmitch> he's keen to see it happening in Ubuntu as well
[10:37] <infinity> If he'd like to work directly with us to push it in faster (especially if what he wants to do is too agressive to make the Etch release), then that's awesome.
[10:37] <infinity> Otherwise, I'm happy to inherit from Debian.
[10:38] <ajmitch> to my great surprise, we can do 0-day NMUs to get selinux functionality in debian
[10:39] <ajmitch> apart from that, I've got some sparc toys to play with for now
[10:39] <ajmitch> & I have to find out what's happening with the hypervisor & virtualisation
[10:39] <ajmitch> though that falls more under the kernel area
[12:24] <trs80> ajmitch: network auth?
[12:24] <ajmitch> yes
[12:25] <trs80> what does that actually mean?
[12:26] <ajmitch> client/server authentication using ldap & kerberos for some sort of single sign-on
[12:26] <trs80> cool. does this mean lib{pam,nss}-ldap will hit main?
[12:27] <ajmitch> using pam_krb5 for this
[12:27] <ajmitch> but possibly :)
[12:29] <trs80> what will you be using for network filesystem? nfs 3, 4 or cifs?
[12:29] <ajmitch> I'm just focusing on the authentication part here
[12:53] <neuralis> infinity: saw the -meeting notice, i'm considering myself poked re: specs (they've been on the todo list for a while, hoping to get around to them late this weekend.)
[02:41] <edneymatias> morning!
[02:42] <lionelp> hi edneymatias
[04:55] <Linuturk> where is the php.ini file?
[05:17] <Linuturk> I need to install the GD Library for my LAMP server. what is the easiest way to do this?
[05:19] <kermitX_> Linuturk, gd for php? try php5-gd
[05:20] <Linuturk> yeah
[05:20] <Linuturk> i installed the default LAMP install
[05:20] <Linuturk> how would I upgrade to the gd?
[05:20] <kermitX_> just install that package.
[05:21] <Linuturk> i'm not seeing it in synaptic
[05:22] <Linuturk> nvm
[05:22] <Linuturk> found it
[05:23] <Linuturk> anything I need to do to configure it?
[05:24] <kermitX_> Linuturk, nope.
[05:25] <Linuturk> i'll need to restart apache before it will take effect?
[05:37] <gpd> [synaptic in the default LAMP install o_O] 
[05:54] <kermitX_> ;)
[08:50] <niallobroin> I'm trying to do a dapper install with RAID1 disk setup, but when it comes time to configure the MD devices, one of the partitions I've setup on one disk isn't offered in the list. All partitions are correctly marked as type fd. Any ideas?
[08:59] <J_P> hi all
[09:01] <J_P> people, I have my ubunu-server and I would like change de locales, so i try do "dpkg-reconfigure locales", but this not works (but in debian works), this command does not open dialog to change locales, onde generate locales up-to-date. Anyone know How I change the locales ?
[09:02] <gpd> you probably need to change the dpkg query level (can't remember the flag right now)
[09:06] <gpd> dpkg-reconfigure -plow
[09:07] <J_P> ok :_0
[09:07] <J_P> :_)
[09:07] <J_P> trying...
[09:07] <J_P> gpd: "dpkg-reconfigure -plow locales"  right ?
[09:08] <J_P> gpd: dpkg-reconfigure -plow locales  result is equal as dpkg-reconfigure  locales :-(
[09:45] <J_P> gpd: decobriu algo como fazer p usar o dpkg-reconfigure locales ?
[10:46] <mystamax> anyone around?
[10:51] <gpd> J_P: install localeconf
[11:04] <J_P> hi all
[11:09] <gpd> J_P: install localeconf
[11:09] <J_P> gpd: ufa!! trying!
[11:18] <J_P> gpd: yes, after dpkg-reconfigure localeconf works.. :-)