[12:08] <sabdfl> Burgwork: yer welcome, though the fix is still in review :-)
[12:08] <Keybuk> bluefoxicy: that won't make any difference
[12:08] <Keybuk> (opening and closing apps)
[12:10] <bluefoxicy> Keybuk:  if you close an application that is taking a lot of ram, and suddenly 400 megs of ram are freed, how much ram was that application using?
[12:10] <bluefoxicy> (for colloquial definition of "using")
[12:11] <Keybuk> bluefoxicy: unknown
[12:11] <Hobbsee> morning all
[12:11] <Keybuk> you certainly can't tell by measuing how much the kernel gave back
[12:12] <Keybuk> for all you know, the kernel chose to flush a bunch of page cache at that moment too
[12:12] <Surak> Hobbsee: night :-)
[12:12] <bluefoxicy> possibly, but not correlative.
[12:12] <bluefoxicy> also page cache only gets flushed under memory pressure
[12:12] <Keybuk> bzzt
[12:13] <bluefoxicy> besides, there's no other good way to measure memory usage
[12:14] <bluefoxicy> RES doesn't count swap; VMEM counts mmap()'d and other COW pages that haven't yet been written to; and shared counts stuff that may be shared by a process and itself exactly.
[12:14] <Keybuk> there is no way to measure memory usage, because memory usage does not make sense
[12:15] <Keybuk> we can only talk about mythical values
[12:15] <doko> mdz: pitti was faster with the glibc upload
[12:15] <bluefoxicy> fair enough
[12:15] <Keybuk> for example, we could say how much memory evolution would use if that was the only binary running on the system
[12:15] <bluefoxicy> but at the same time we can create trends in said values.
[12:15] <bluefoxicy> which have a physical effect.
[12:15] <bluefoxicy> you don't need to quantify things 1:1 when you do analysis like this.
[12:16] <zyga> bluefoxicy: you know...
[12:16] <Keybuk> if you really car
[12:16] <Keybuk> evolution uses roughly the amount of memory needed to hold the headers of every message that need to go in the list
[12:16] <Keybuk> it's high
[12:16] <Keybuk> but it's not 600MB
[12:17] <Keybuk> it's around 80MB
[12:17] <zyga> bluefoxicy: do this, log each allocation via .alloc/free/realloc and check out how memory is actually being used, looking at number of pages allocated really doen't tell the whole story
[12:18] <Keybuk> zyga: that's not necessarily telling either, sadly
[12:18] <Keybuk> free doesn't return memory to the kernel, remember
[12:18] <bluefoxicy> for example, I'm going to take the security+ exam soon; the pretest had 30 questions and I got 83%, mastery was 70%; I finished approximately 47% above passing.  Someone told me the test will have about 80 questions; I'm estimating a 20% variance, so I should finish with 38-56% above passing.
[12:18] <zyga> Keybuk: yes, that's why I recommend that
[12:18] <Keybuk> an application using malloc/free badly can be "holding" far more memory than it should
[12:18] <zyga> it tell's the story of the memory needs of the app
[12:18] <bluefoxicy> the numbers say I'm likely to pass; they also say I could study operational security and probably pass with a higher score
[12:18] <zyga> not the memory flow between userspace and the kernel
[12:18] <bluefoxicy> But as you say, mythical numbers that mean nothing right?  :)
[12:19] <bluefoxicy> Keybuk:  and you're right.
[12:19] <Keybuk> bluefoxicy: uhh, that's completely irrelevant
[12:19] <zyga> Keybuk: I am well aware of all that, it's just a better number than rss
[12:19] <bluefoxicy> evo may have [10M]  .............................................[1 byte]  and hold a bunch of unused memory
[12:19] <Keybuk> it would only be relevant if your score was based on the number of questions you answered differently to anyone else in the room
[12:19] <bluefoxicy> Keybuk:  I was illustrating "mythical numbers" having a real point and being useful.
[12:19] <Keybuk> so if your test was scored on the number of questions you answered uniquely and correctly *then* it would be a useful comparison :p
[12:19] <zyga> bluefoxicy: or maybe run valgrind tru the app you want, find a leak!
[12:20] <bluefoxicy> Keybuk:  "obviously" i can make no assumptions as to what my score will be; I could very well fail horribly.  But I can say what is likely the case.
[12:20] <Keybuk> because then the people that answered all the questions right could still fail, because their unique answers were fewer than anyone else
[12:20] <bluefoxicy> heh
[12:20] <Keybuk> at that point, you get to the complexity of discussing memory usage on Linux
[12:20] <bluefoxicy> that changes the problem.
[12:21] <Keybuk> evolution's primary memory usage problems are
[12:21] <Keybuk> a) a large stack of (shared) dependencies
[12:21] <bluefoxicy> zyga:  valgrind is noisy; it'd be nice if I could actually produce a map of "there are leaks here, this is the percentage that we find to be leaking" and send it to the devs
[12:21] <Keybuk> b) a high portion of memory not returned to the kernel
[12:21] <Keybuk> I doubt it has a single "leak"
[12:21] <zyga> bluefoxicy: valgrind is not noisy but I will not speak of it further
[12:22] <zyga> bluefoxicy: (just a note that not all memory is freed, ever, before the process exits so valgrind is far from perfect)
[12:22] <Keybuk> zyga: valgrind usually doesn't mind provided the app has a pointer on it
[12:22] <Keybuk> it can cope with that
[12:23] <Keybuk> though it doesn't cope with you not bothering to free something because you know the app is exiting, even though the pointer will be lost before the end of main
[12:23] <Keybuk> but then if you're running under valgrind, the actual termination isn't useful
[12:23] <Keybuk> provided you don't lose hold of the pointers during ordinary use, you don't have a leak
[12:23] <zyga> Keybuk: true
[12:24] <Keybuk> you may, of course, be holding onto pointers you don't need
[12:24] <zyga> this reminds me of the multitude of the 'library valgrind behaviour' files valgrind comes with
[12:24] <Keybuk> I read an interesting paper the other day
[12:24] <Keybuk> parsing config files, etc.
[12:24] <zyga> so that stuff that's 'tricky' yet right is not messing the output, I wonder how that stuff is created
[12:25] <Keybuk> the WRONG way to do this is to try and use fgets and dynamic allocation to cope with arbitrary length lines
[12:25] <Keybuk> even using realloc on a static pointer is bad, as you'll end up littering memory
[12:25] <bluefoxicy> zyga:  I used valgrind a long time ago, every time memory leaked it cried about it, instead of being summerical.
[12:25] <Keybuk> the RIGHT way, which for some reason everyone is scared of, is just to mmap() the config file and treat it as a big string
[12:25] <zyga> bluefoxicy: try reading the very few last lines next time
[12:25] <bluefoxicy> I'm the kind of guy that does an strace -c :)
[12:26] <zyga> Keybuk: then we hit mmap issues but I agree mmap is the right way to let the kernel manage your memory needs
[12:26] <bluefoxicy> sigh.  I need help with the shell, what the hell.
[12:26] <zyga> then again unless you are parsing bazillion config files it not that of an issue
[12:26] <bluefoxicy> Does anyone know how to filter characters
[12:27] <Keybuk> mmap issues?
[12:27] <Keybuk> bluefoxicy: -v
[12:27] <bluefoxicy> why am I doing this
[12:27] <Keybuk> | sed -e "s/[A-Z] $//"
[12:27] <Keybuk> or
[12:27] <bluefoxicy> this is stupid
[12:27] <bluefoxicy> thanks
[12:27] <zyga> Keybuk: yes, it's sometimes problematic with regards to porting and not being 'always works' kind of stuff
[12:27] <Keybuk> ${%[A-Z] }
[12:27] <Keybuk> ie FOO=100K
[12:27] <Keybuk> echo ${FOO%[A-Z] }
[12:28] <Keybuk> zyga: *shrug*  all the world is Linux ... it's so much easier to code for
[12:28] <zyga> (you usually need to take care of the part when it fails to be 100% sure)
[12:28] <Keybuk> I think too much time is wasted supporting 20 year old versions of AIX
[12:28] <zyga> Keybuk: nah  - all the world is the stuff I want to code for
[12:28] <zyga> but that's getting more than just linux :)
[12:28] <Keybuk> or, at least, all the world is POSIX
[12:29] <bluefoxicy> bluefox@icebox:~/programming/vuln/printf$ j=0; for i in $(du -s $(find /usr/lib/openoffice/program/ -name \*.so) | awk '{print $1}'); do j=$(( $i + $j )); done; echo $j
[12:29] <bluefoxicy> 101796
[12:29] <zyga> Keybuk: but then you miss windows and that huts in lots of places
[12:29] <bluefoxicy> wow openoffice.org mmap()s 100 megs of libraries.
[12:29] <Keybuk> zyga: I don't care about windows
[12:29] <Keybuk> bluefoxicy: ah, now, if you want to discuss bloat ... I have no problem with you targetting OpenOffice
[12:29] <Keybuk> most of those libraries are private too
[12:29] <zyga> Keybuk: as I said, in some places
[12:29] <Keybuk> so not only do you have a whole bunch of lard mapped in
[12:29] <bluefoxicy> Keybuk:  and there's one real soffice.bin
[12:29] <Keybuk> but then you have all those dirty plts over the top!
[12:29] <TheMuso> c
[12:29] <bluefoxicy> Keybuk:  I know, this is fucking stupid.
[12:30] <Keybuk> "Why OpenOffice.org is slow"
[12:30] <bluefoxicy> the dynamic linker is like AHHH GOD AHHHH MUST RELOCATE MAP RESOLVES SYMBOLS AHHHHHH
[12:30] <bluefoxicy> Keybuk:  OOo is also slow because it uses C++
[12:30] <bluefoxicy> Keybuk:  but, again, it is slow because the linker has 30 times more symbols to deal with from the crap C++ adds to the symbol table :P
[12:30] <bluefoxicy> so if all those libraries were built into the main executable, the problem would again go away.
[12:31] <Keybuk> zyga: "portability" oddly enough
[12:31] <bluefoxicy> hehe
[12:31] <zyga> bluefoxicy: hint, c++ has symbol hiding now
[12:31] <bluefoxicy> zyga:  this doesn't mean developers use it.
[12:31] <zyga> Keybuk: portablility across software platforms often sucks, ack
[12:31] <bluefoxicy> you have to explicitly declare symbols as hidden; or set the default behavior to hidden and explicitly set visible symbols to default.
[12:32] <zyga> bluefoxicy: are you sure...?
[12:32] <bluefoxicy> zyga:  we do have one nice thing here though.
[12:32] <bluefoxicy> Michael Meeks has been trying to take the from-the-side approach
[12:32] <bluefoxicy> and we got some very nice things out of it (even if the RHAT guys are boning him and trying to steal the credit)
[12:32] <Keybuk> *shrug* he's an idiot
[12:32] <Keybuk> if they just built the fucking thing properly in the first place, they wouldn't have this problem
[12:32] <Surak> night all
[12:33] <Keybuk> "Doctor, it hurts when I do this"
[12:33] <bluefoxicy> zyga:  Im sure about the symbol hiding thing.. it's __attribute__ __(("hidden"))__ or some such crazy syntax
[12:33] <bluefoxicy> Keybuk:  who, Meeks?
[12:33] <Keybuk> that's the C syntax, certainly
[12:34] <bluefoxicy> the stuff he's doing would give a gain even with properly hidden symbols; it just makes the linker do less work with what it has to do
[12:34] <zyga> bluefoxicy: AFAIR there was some 'smart' hiding by default so the load on the linker is comparable to that of C
[12:34] <bluefoxicy> zyga:  ah, nice.
[12:34] <bluefoxicy> zyga:  I am more interested in -Bdirect binding to eliminate vague linkage where possible; but drepper despises the very idea
[12:34] <Keybuk> yes, but if OO.o was built properly in the first place, they wouldn't need to be mucking around with the linker
[12:34] <Keybuk> yeah, I agree with Ulrich here, I'm afraid
[12:35] <Keybuk> it's taking a chainsaw to the linker because you happen to have dropped it off a cliff
[12:35] <Keybuk> it's far easier if people just don't break it in the first place
[12:35] <bluefoxicy> Keybuk:  if OOo was built properly in the first place, mucking the linker would reduce the load time from 1 second to 0.1 second instead of from 12 seconds to 5 seconds
[12:35] <zyga> I'm off 
[12:36] <bluefoxicy> also direct binding makes the linker skip searching spaces "we know" it won't find a symbol in
[12:36] <zyga> this kind of talking does nothing positive to oo. :)
[12:36] <bluefoxicy> i.e. it makes it not look in the toilet for a floppy disk :P
[12:36] <zyga> I look forward to edgy and to being here again
[12:36] <zyga> take care and good night
[12:36] <bluefoxicy> zyga: heh, later man.
[12:36] <elmo> where's logmsg() meant to come from in perl?
[12:37] <bluefoxicy> Keybuk:  btw, they did put precomputed hashes and symbol/hash sorting into the mainline binutils, 50% speed boost :)
[12:38] <bluefoxicy> I'm hoping edgy+1 will use that, by backport patch or appropriate upstream version
[12:38] <bluefoxicy> -Bdirect is a slight mess now
[12:38] <bluefoxicy> it'd create some junk that would get deprecated eventually, so I'm not going to advocate that
[12:38] <sladen> bluefoxicy: plugging routers into a switch is normally practice, it allows [amongst other things]  (a) monitoring of ports (b) running the router port trunks == breaking out lots more "ports" (c) detecting failures (d) dynamically "re-routeing" at level 2 when things go wrong [eg. an abstraction layer] 
[12:39] <bluefoxicy> sladen: router -> switch (connected to a subnet full of hosts) -> other router?
[12:41] <bluefoxicy> sladen:  assuming the default gateway leads to a specific point; and the next subnet has a default gateway leading back to that point; as your network grows, the load on any given router link is going to grow to n^x load, where (n) is the number of subnets behind that router port (including its local subnet) and (x) is the number of hosts per subnet
[12:42] <bluefoxicy> no sorry x^n
[12:43] <bluefoxicy> anyway. if you do router<->router, you can lower the value significantly; isolate the default path; and often approach n^1, linear growth instead of polynomial
[12:43] <bluefoxicy> you can hook the routers to eachother in a mesh form and get (c) and (d) (this is what routing protocols are for)
[12:44] <sladen> I don't quite follow, but yes, if the packet goes in, then out of the same port on the router, the link will be running at 2x (really not a problem given that it's going to be a non-blocking link and the packets are going in opposite directions)
[12:44] <bluefoxicy> monitoring of ports.. no comment.  The only thought I can come up with is "plug a passive IDS inbetween" which sounds like "add excess hardware" although it wouldn't be particularly bad
[12:46] <bluefoxicy> sladen:  I'm saying the dedicated link between subnet A and the internet (which is where default routes "usually" take you) will also carry ALL of subnet B's traffic; if there is another subnet C on the other side of the default gateway for subnet A, then we've now loaded bandwidth on our link between A and C
[12:46] <sladen> and how do you detect when the IDS goes tits up?...
[12:46] <sladen> transparent anythings are generally a bad idea
[12:47] <bluefoxicy> if you hook the two routers together, then we've loaded bandwidth between them and to the Internet and whatever; but connection to networks that AREN'T experiencing a network storm are going to be just fine.
[12:47] <bluefoxicy> Typically an IDS is connected to a monitoring station
[12:47] <bluefoxicy> which may or may not be part of your core network
[12:47] <bluefoxicy> (i.e. it can't talk through the IDS and might not be connected to anything else; this is perfectly fine)
[12:47] <bluefoxicy> and it'll say "HOLY CRAP IDS DIED!"
[12:48] <bluefoxicy> you could even have multiple IDSes and a separate network backbone in between them, in theory; but this is blah.
[12:48] <sladen> (no doubt running Windows and connected to a switch port hanging of the other side of the router ...)
[12:48] <bluefoxicy> switch->span port->IDS<->Monitor station
[12:48] <bluefoxicy> is the normal configuration
[12:49] <bluefoxicy> most people stick them after the point of presence and before the routing infrastructure starts to branch
[12:49] <bluefoxicy> or if you have a BIG network you stick multiple ones around on different network segments since one won't handle the massive load
[12:50] <bluefoxicy> when I did it the monitor station was plugged into the IDS directly because we didn't want people hacking into it
[12:50] <bluefoxicy> they hacked the CEO's desktop instead, after stealing our router using SNMP to make configuration changes :/
[12:50] <bluefoxicy> however, WE SAW WHO DID IT.
[12:53] <bluefoxicy> what did that look like... {Internet}<->[Router] <->[PIX] <->[Switch] <->[Patch Panel] <->{network}  with a monitor station dangling off the PIX I believe (we were using the PIX for port redirecting and firewalling)
[12:53] <bluefoxicy> anyway
[12:53] <bluefoxicy> this is not important
[12:56] <Keybuk> PUSH PINEAPPLE, GRIND COFFEE!
[12:56] <Hobbsee> okay?
[12:56] <Keybuk> sorry, it had to be sung
[12:56] <zul> Keybuk: you ok
[12:58] <Keybuk> I could be very evil, and inflict you all with the MP3 that Rhythmbox decided I wanted to listen to
[12:59] <zul> heh...still listening to megadeth so do you worse
[01:00] <Keybuk> http://www.netsplit.com/tmp/03.Black_Lace-Agadoo.ogg
[01:02] <zul> oh...my...god
[01:02] <zul> hey ajmitch 
[01:03] <ajmitch> hey zul 
[01:14] <bddebian> Heya
[03:35] <imbrandon> o/~ .... sing with a hoola meoldy .... o/~
[03:37] <zul> uh...ok
[03:38] <imbrandon> heh , keybuk's song
[03:38] <zul> that was just evil
[03:38] <imbrandon> lol
[05:12] <jeff_> When I install a package for an open source program is the source usually installed as well?
[05:15] <LaserJock> jeff_: not usually
[05:16] <jeff_> I'm looking for devilspie's source code in particular
[05:16] <bddebian> Rarely if you are doing a binary package install
[05:16] <LaserJock> jeff_: normally it takes up too much bandwidth and hard drive space
[05:16] <LaserJock> jeff_: you apt-get source devilspie
[05:16] <bddebian> jeff_: Then you need to apt-get source devilspie
[05:16] <LaserJock> but make sure you have the source repositories enabled
[05:16] <bddebian> Aye
[05:17] <jeff_> which source repositories should I enable?
[05:18] <poningru> if you got the program from apt-get then you can get the source with the same repos source
[05:18] <jeff_> ah, thanks
[05:18] <jeff_> I got it now
[05:19] <jeff_> hrm, is there a standard name for source packages?
[05:19] <jeff_> I installed it using apt-get source devilspie, but I don't see it when I search for devilspie in synaptic package manager
[05:20] <LaserJock> source packages don't show up in synaptic
[05:22] <jeff_> ok, is there an obvious reason for that?
[05:22] <LaserJock> could be
[05:23] <HrdwrBoB> because it would be terribly confusing?
[05:23] <LaserJock> apt-get source is run as a user and grabs the source package and unpacks it in the current directory
[05:24] <jeff_> ok
[05:31] <poningru> needed some help with a bug
[05:31] <dmb> i don't know if this is the right room, but i just wanted to let you know that the mirror http://ftp.wayne.edu is not working
[05:31] <dmb> the ubuntu download page links to http://ftp.wayne.edu/linux_distributions/ubuntu/6.06/ubuntu-6.06-desktop-i386.iso, which isn't working
[05:32] <poningru> bug 55706
[05:32] <Ubugtu> Malone bug 55706 in python-uncertainities "python-uncertainities python2.3/2.4 breakage." [Untriaged,Confirmed]  http://launchpad.net/bugs/55706
[05:32] <poningru> its a simple fix on postinst
[05:32] <LaserJock> poningru: #ubuntu-bugs is the place to go
[05:32] <poningru> no one helped there :(
[05:32] <dmb> oh wait, i think thats because of the change to 6.06.1
[06:30] <jeff_> Anyone here familiar with GObject and signals?
[06:31] <jdub> jeff_: gimpnet is going to be a better place to start for that :)
[06:32] <jeff_> gimpnet?
[06:32] <ajmitch> afternoon jdub 
[06:32] <jdub> jeff_: irc.gnome.org / irc.gimp.org
[06:32] <jdub> yo ajmitch 
[06:34] <jeff_> jdub: thanks
[06:51] <BenC> anyone able to process linux-source-2.6.17 so l-r-m will get built?
[07:00] <infinity> BenC: Oh, I suppose so. :)
[07:00] <BenC> infinity: thank you great archive maintainer :)
[07:03] <infinity> Hrm.
[07:03] <infinity> I only see two arches in the NEW queue...
[07:03] <infinity> Did the others get processed already?
[07:05] <BenC> all 5 built successfully
[07:06] <infinity> Yeah, and amd64/powerpc/i386 got rejected.  Checking why now.
[07:06] <infinity> 20:34:14 INFO    Rejected:
[07:06] <infinity> 20:34:14 INFO    Exception while accepting: 'Description'
[07:06] <infinity> Awesome.
[07:07] <infinity> Broken control file?
[07:07] <BenC> is it one of the new udeb's?
[07:07] <BenC> speakup and nic-firmware were added
[07:07] <infinity> It's not that specific, sadly.
[07:07] <BenC> do udeb's have to have a description?
[07:08] <BenC> I'm pretty sure that's what it is
[07:08] <infinity> Everything should, afaik.
[07:08] <BenC> bah, kernel-wedge didn't complain and neither did dpkg-dev :P
[07:08] <BenC> *sigh*, guess I'll reupload
[07:09] <infinity> It could be soyuz being too anal, but I can't see a valid reason for a package NOT having a description either.
[07:39] <jeff_> I'm getting a glib.h: no such file or directory error even though I have libgtk2.0-dev installed.
[07:51] <BenC> infinity: does initramfs-tools not handle DSDT.aml's anymore?
[07:54] <jdub> jeff_: you need the glib dev package installed to have the glib header file
[07:54] <jdub> jeff_: channels on gimpnet might be more appropriate for hacking on gtk+/gnome stuff than this one
[07:55] <jeff_> jdub: thanks, but it seems that I do have the glib dev packages installed
[07:56] <BenC> jeff_: then likely you aren't using the right CFLAGS to find the headers...this is a question for #ubuntu though
[07:56] <BenC> or what jdub suggested
[07:59] <jeff_> ok
[08:12] <infinity> BenC: If it doesn't, I'm not the one who broke it.
[08:12] <infinity> BenC: I'll look when I get hom in an hour.
[09:16] <bencer> hi all, where can i find a list of changes/updates for 6.06.1 ?
[09:16] <lucas> is there something such as snapshots.debian.net for ubuntu ?
[09:17] <lucas> I'd need the Sources file from every day over a long period of time
[09:38] <hunger_work> Is it possible to get an misdn updated? ISDN is still pretty popular in germany and misdn is version 0.0.0+cvs20041018-5.
[09:39] <infinity> lucas: Since we switched to launchpad (early this year), everything's stored in launchpad, per version.
[09:40] <lucas> yup, but then, t's not possible to get per-day Sources file :( (or I'd have to do a lot of data harvesting)
[09:43] <mvo> hunger_work: that would be a whishlist bug for the kernel team I guess
[09:59] <pitti> Good morning
[10:00] <Nafallo> morning pitti, Hobbsee :-)
[10:00] <Hobbsee> hi Nafallo, pitti 
[10:03] <AlinuxOS> Good morning pitti ;)
[10:03] <pitti> hey Nafallo 
[10:03] <pitti> moin Hobbsee 
[10:04] <pitti> AlinuxOS: hey, long time no see! how are you?
[10:05] <AlinuxOS> pitti, Im fine thanks!Finished exams, now in Germany by my girlfriend :) (re-starting gnome translation)
[10:05] <pitti> cool
[10:05] <AlinuxOS> pitti, and what about you ? :) alles in ordnung :D
[10:06] <pitti> AlinuxOS: ja, prima!
[10:06] <AlinuxOS> pitti, great!
[10:09] <Hobbsee> hi sabdfl 
[10:09] <pitti> moin sabdfl
[10:09] <Fujitsu> Good evening, sabdfl.
[10:09] <Hobbsee> pitti: hey where's the duncecap?  i think i need to wear it for a while.
[10:10] <Hobbsee> hehe
[10:10] <pitti> Hobbsee: heh, what's up?
[10:11] <Hobbsee> pitti: er...cleared crap out of my home dir and backed up, adn used the backup to overwrite the original.  and then found that i'd missed a couple of folders.
[10:11] <Hobbsee> like.... .thunderbird, .gnupg
[10:11] <Fujitsu> Ouch!
[10:11] <Hobbsee> and goodness knows what else i havent noticed that i'm missing yet.
[10:11] <Fujitsu> I hope you had backups of your keys...
[10:11] <hunger_work> Hobbsee: Outch! You do have backups?
[10:12] <Hobbsee> hunger_work: of .gnupg/ yes, anything else, no
[10:12] <Fujitsu> Owy.
[10:12] <Fujitsu> This is why I use IMAP on all my things :P
[10:12] <Hobbsee> which is the @kubuntu.org/@ubuntu.com stuff too.
[10:12] <Hobbsee> Fujitsu: gmail doesnt offer it
[10:12] <pitti> Hobbsee: argh, you lost all your dotfiles/dotdirs?
[10:12] <Fujitsu> Hobbsee, neither does my ISP, but my server takes care of that.
[10:12] <Hobbsee> pitti: got some of them.
[10:12] <pitti> Hobbsee: you don't have a backup of your secret key?
[10:13] <Hobbsee> pitti: yeah, iv'e got backups of all of .gnupg
[10:13] <pitti> *phew* :)
[10:13] <Hobbsee> heh, yeah
[10:13] <Fujitsu> Losing keys is... really irritating.
[10:13] <Fujitsu> But moreso if they're actually used for something important, like yours :P
[10:13] <Hobbsee> that was my first reponse - oh shit, here's my revoke.asc, but i dont remember seeing .gnupg....
[10:13] <Hobbsee> Fujitsu: heh, true
[10:14] <Nafallo> pitti: btw, are you familiar with backuppc's source or was it someone else?
[10:14] <pitti> Nafallo: not really; I still use my ancient home-grown flexible automatic network backup system
[10:14] <Nafallo> ah, oki. I have a small problem with it I should start hacking on...
[10:15] <Hobbsee> but i know it's certainly not recent.
[10:15] <Nafallo> Bug 54918
[10:15] <Ubugtu> Malone bug 54918 in backuppc "DNS-lookup to AAAA (IPv6) doesn't seem to work" [Wishlist,Unconfirmed]  http://launchpad.net/bugs/54918
[10:15] <Hobbsee> aye, my computing work is gone.  good thing i didnt start the assignment yet
[10:16] <Fujitsu> :(
[10:17] <pitti> Hobbsee: I guess that must happen to everyone until they install an automatic backup :/
[10:17] <Hobbsee> pitti: could do that.
[10:19] <doko> pitti: still seeing the dh_strip failures with the new glibc
[10:21] <pitti> doko: hm, works here now on amd64
[10:21] <pitti> doko: objdump -h /usr/lib/*crt*.o|grep debuglink   ?
[10:22] <pitti> doko: you have libc6-dev 2.4-1ubuntu8 now?
[10:23] <Hobbsee> oh holy crap :(
[10:23] <pitti> doko: oh, wait - did you try this with a package which wasn't cleaned before building again?
[10:23] <doko> pitti: the latter ... will rebuild
[10:23] <Hobbsee> hunger_work: i've lost all my saved passwords too :(
[10:23] <pitti> doko: i. e. if you built the package with just dh_clean; debuild -nc, then you'll still get the bug
[10:23] <pitti> doko: since the built executables will have debuglinks
[10:24] <hunger_work> Hobbsee: I can recommend faubackup.
[10:24] <pitti> doko: ok, *phew* :)
[10:24] <Nafallo> ubuntu main and everything :-)
[10:25] <Spads> reverse incrementals > *
[10:29] <siretart> any archive admins around?
[10:42] <nags> seb128, http://people.freedesktop.org/~prashmohan/gedit/ also a README there :)
[10:44] <seb128> nags: 
[10:44] <seb128> mget: Accs chou: 403 Forbidden (gedit-data.xml)
[10:44] <seb128> mget: Accs chou: 403 Forbidden (gedit-main.xml)    
[10:44] <seb128> mget: Accs chou: 403 Forbidden (gedit.py) 
[10:44] <seb128> mget: Accs chou: 403 Forbidden (gedit1.py) 
[10:44] <seb128> mget: Accs chou: 403 Forbidden (gedit2.py) 
[10:44] <seb128> mget: Accs chou: 403 Forbidden (gedit3.py) 
[10:44] <seb128> mget: Accs chou: 403 Forbidden (gedit4-data.xml)
[10:45] <nags> seb128, oops
[10:45] <seb128> nags: could you fix that? :)
[10:45] <nags> sure
[10:45] <seb128> thank you
[10:49] <nags> seb128, can you try downloading the same from here - http://people.freedesktop.org/~nagappan/gedit/
[10:51] <seb128> nags: works fine
[10:51] <nags> seb128, cool :)
[10:52] <pitti> iwj: wrt ffox, what do you think about calling it '2.0~beta1' instead of '1.99+2.0beta1'? Personally I find the former a bit nicer to read
[10:52] <seb128> pitti: is 2.0 > 2.0~beta1?
[10:52] <pitti> iwj: (and we can use ~ since breezy or even earlier)
[10:52] <pitti> seb128: yes
[10:52] <seb128> ah ok
[10:52] <pitti> seb128: that's the point of ~ :)
[10:52] <seb128> yeah, I never used it :p
[10:53] <pitti> you can read it as 'minus epsilon'
[10:53] <pitti> seb128: ~blah is always smaller than the empty string (also handles the dash correctly)
[10:53] <seb128> good to know ;)
[10:54] <Fujitsu> Yeah.
[10:55] <Hobbsee> gnomefreak: how stable is it?
[10:55] <Hobbsee> ie, how many crashes do you get?
[10:58] <gnomefreak> very compared to 3.0 :)
[10:58] <gnomefreak> Hobbsee: none
[10:59] <Hobbsee> gnomefreak: nice!
[11:00] <Hobbsee> gnomefreak: when's it supposed to be released?
[11:00] <gnomefreak> Hobbsee: do you have a package that is easy to merge that i can try to merge. (looking for one not likely to break)
[11:00] <Hobbsee> gnomefreak: merges.ubuntu.com/universe.htm
[11:00] <gnomefreak> spt-oct last i heard
[11:00] <Hobbsee> could be anything on them
[11:00] <Hobbsee> right
[11:00] <gnomefreak> so it doesnt matter whos name is there?
[11:01] <Hobbsee> gnomefreak: nope.  dont touch zope or beagle
[11:01] <Hobbsee> hey cool, this email isnt *too* out of date.
[11:02] <gnomefreak> k thank you
[11:02] <gnomefreak> thanks got amarok beta last night ;)
[11:04] <gnomefreak> does it matter if i grab new or updated?
[11:05] <Hobbsee> gnomefreak: nope
[11:05] <gnomefreak> ok here i go ;)
[11:15] <gnomefreak> should this merge script ask me "are you sure you want to delete everything in /temp/
[11:18] <Hobbsee> gnomefreak: yes
[11:18] <gnomefreak> k got scared there ;)
[11:18] <Hobbsee> gnomefreak: where are you getting the 2.0 packages from?  mozilla site, or from the post on the mailing list?
[11:18] <gnomefreak> mozilla
[11:18] <mjg59> Wow, this stuff /works/
[11:18] <gnomefreak> hold on a sec ill get you the page
[11:19] <Treenaks> mjg59: is it an Acer? :P
[11:19] <mjg59> uswsusp
[11:19] <gnomefreak> Hobbsee: http://www.mozilla.org/projects/bonecho/index-2.0b1.html
[11:20] <gnomefreak> brb while all these things are downloading
[11:24] <Nafallo> mjg59: nice bughistory in debian atleast :-) (was "lucky"link at google ;-))
[11:27] <thom> segfaulting for config problems is very classy
[11:28] <Hobbsee> heh
[11:28] <Hobbsee> indeed
[11:28] <mjg59> I've junked all the config parsing code
[11:28] <mjg59> In fact, I've junked most of the code full sotp
[11:28] <gnomefreak> Hobbsee: can you look at this i get 4-5 lines that start with C is this a conflict or what does it mean http://paste.ubuntu-nl.org/20267
[11:29] <StevenK> gnomefreak: The REPORT mentions them.
[11:29] <Hobbsee> gnomefreak: yeah, it shows the debian and ubuntu changes with the >, <, and lets you sort out what you do and dont want
[11:30] <gnomefreak> StevenK: im reading that now :(
[11:30] <mjg59> Anyway, it's much faster
[11:52] <gnomefreak> before i go nuts is it as easy to ust take the newest versions with patches and make a source.tar.gz and than continue? if not how do i make this work
[11:52] <gnomefreak> s/ust/just
[11:52] <Hobbsee> gnomefreak: cd into the directory MoM created for you, and then do the changes from there?
[11:52] <gnomefreak> it created a dir?
[11:53] <Hobbsee> should have done
[11:53] <Hobbsee> you used grab-merge.sh didnt you?
[11:54] <gnomefreak> yes
[11:54] <gnomefreak> is this the one with help and debian nad pixmaps
[11:54] <gnomefreak> its the only normal looking folder in ~/temp
[11:54] <gnomefreak> the rest are tars or dsc
[11:55] <tepsipakki> Kamion: maybe debian-install rebuild is all that is needed to pick up the new libc-udebs?
[11:55] <Hobbsee> er, which are you getting?
[11:55] <gnomefreak> gnomebaker
[11:55] <Hobbsee> aah, right.
[11:55] <Hobbsee> you're looking in the dir where you ran the script?
[11:55] <gnomefreak> yes
[11:56] <Hobbsee> hmmm
[11:56] <gnomefreak> im in where it left the changes and the report adn so on
[11:57] <gnomefreak> the merge-genchanges and merge-buildpackage
[12:07] <gnomefreak> im wondering if i should request a sync on gnomebaker than when its done re try?
[12:08] <Hobbsee> gnomefreak: bug report, subject: [Edgy MoM]  Please sync foo version from debian sid, get a MOTU to ack it, and subscribe ubuntu-archive
[12:11] <gnomefreak> the version would be the latest version i got in the download right?
[12:12] <Hobbsee> you're wanting the latest debian, shown in the report
[12:15] <gnomefreak> ok how do i get a motu to (ack) it. oh and what is ack?
[12:15] <Hobbsee> ack = approve
[12:15] <thom> acknowledge
[12:16] <gnomefreak> ah
[12:16] <Hobbsee> ask for someone to approve it in -motu, i expect
[12:16] <gnomefreak> k
[12:18] <Hobbsee> gnomefreak: or subscribe ubuntu-universe-contributors to it
[12:18] <gnomefreak> k
[12:33] <madduck> lucas: do you have the scripts for the comparison available somewhere? (re: blog)
[12:34] <lucas> I'll put them somewhere
[12:34] <sivang> hi all
[12:35] <sivang> does anybody have a recommendation for a wirless gateway/router ? by reference to a brandname / mafctr. ?
[12:35] <thom> netgear dg834
[12:35] <sivang> thom: hi, what about the EDIMAXes and friends? ;-)
[12:36] <thom> sivang: what about them? the dg384 works beautifully
[12:37] <sivang> thom: Do you have any expreince with those ? are they good as the netgear ?
[12:37] <ogra> if you look for cheapo stuff, i had no bad experiences with belkin in that price class
[12:38] <ogra> (dont ask for model names ... but they are generally cheap)
[12:38] <sivang> ogra: cool, thanks :)
[12:38] <sivang> ogra: and also much more spread here (as well as the edimaxs) then netgreas..
[12:40] <lucas> madduck: http://www.lucas-nussbaum.net/bazaar/cmpver.rb
[12:42] <iwj> pitti: I was going to ask whether ~ is allowed in Ubuntu version numbers.  Nice to have `yes' as an answer.
[12:42] <Kaleo|work> rodarvus: do you think we can have another UVF for xserver-xorg-video-i810 ?
[12:42] <iwj> pitti: But I won't do ff 2.0~beta unless Debian do, just to avoid aggro.
[12:42] <pitti> iwj: we have used it for backports for quite a while now, and even Debian can use it now :)
[12:42] <rodarvus> Kaleo|work, we had one yesterday :)
[12:42] <rodarvus> version 1.6.4 is on the archives
[12:42] <pitti> iwj: oh, right, if Debian has an orig.tar.gz, we should use that
[12:42] <iwj> pitti: Right.
[12:43] <rodarvus> I'm working on libdrm/mesa update, to support 3D on the new cards supported by this driver
[12:43] <rodarvus> Kaleo|work, unless a 1.6.5 was released in the last 6 hours (time I've been sleeping), there is no need for another UVF of xserver-xorg-video-i810 ;)
[12:46] <Kaleo|work> rodarvus: that's exactly the case
[12:46] <Kaleo|work> ;)
[12:46] <rodarvus> oh
[12:46] <rodarvus> heh
[12:46] <Kaleo|work> see bug ... hmmm
[12:46] <rodarvus> Kaleo|work, yes, we can have UVF exceptions for them
[12:47] <rodarvus> for this new version
[12:48] <rodarvus> Kaleo|work, I was expecting a 1.6.5 soon, but not in 6 hours :)
[12:49] <Kaleo|work> hehe
[12:51] <Kaleo|work> yes rodarvus, bug 55907
[12:51] <Ubugtu> Malone bug 55907 in xorg-server "Slow indirect rendering after update to version 1.6.4" [Unknown,Fix released]  http://launchpad.net/bugs/55907
[12:51] <Kaleo|work> I linked to upstream
[12:51] <Kaleo|work> before I went to bed ;)
[12:53] <rodarvus> yeah, i810 yesterday
[12:53] <rodarvus> but was not my fault this time :)
[12:54] <Hobbsee> rodarvus: crud, how's it broken?
[12:56] <rodarvus> Hobbsee, upstream removed a few compatibility information from version 1.6.4 of the i810 driver
[12:56] <rodarvus> restored on 1.6.5
[12:57] <Hobbsee> rodarvus: right.  so X is gone till the new updates, or what?
[12:57] <rodarvus> no
[12:57] <Hobbsee> oh good
[12:57] <rodarvus> its just slower for people with Intel boards, and only some of them
[12:57] <Hobbsee> right
[12:57] <rodarvus> i830 or newer, I think
[12:58] <Hobbsee> right
[01:00] <Nafallo> rodarvus: I didn't see any effects on i845GE :-)
[01:01] <rodarvus> good
[01:01] <rodarvus> anyhow, 1.6.5-0ubuntu1 was uploaded
[01:02] <rodarvus> should be compiled an published in one or two hours, hopefully] 
[01:02] <Nafallo> nice
[01:02] <ajmitch> hi
[01:03] <AlinuxOS> pitti, ping
[01:03] <Kaleo|work> rodarvus: thanks so much !
[01:03] <rodarvus> Kaleo|work, thank YOU for the bug triaging
[01:03] <rodarvus> bug triaging is always helpful and appreciated :)
[01:03] <pitti> AlinuxOS: pong
[01:06] <webben> pitti: i just noticed your comment on https://launchpad.net/distros/ubuntu/+source/sudo/+bug/50797
[01:06] <Ubugtu> Malone bug 50797 in sudo "sudo built with --with-secure-path is problematic" [Wishlist,Unconfirmed]  
[01:07] <webben> pitti: rather than dropping --secure-path, why not restore the ability to change the secure path in /etc/sudoers ?
[01:07] <webben> (i was actually working on this, i got the secure path working at the expense of causing some other bug, if you'd like to see a diff)
[01:08] <webben> (but i don't think it's hard -- i don't know c++ and i got halfway there)
[01:08] <pitti> webben: it's only C :)
[01:08] <webben> pitti: see, i don't even know which language it is
[01:08] <pitti> webben: sure, if you have a patch for this which doesn't cause regressions and does not require us to change configuration files, that's fine
[01:08] <webben> pitti: and i still managed to hack support back in
[01:09] <webben> pitti: the problem is my patch caused some other annoying bug, and i couldn't figure out why
[01:09] <webben> pitti: and it required one deeply inefficient reuse of existing code
[01:09] <pitti> ^ for reasons like this I'm very cautious about sudo patches :)
[01:09] <webben> pitti: it really does need somebody who knows what they're doing i think
[01:09] <webben> pitti: which isn't me (although i did try and will happily share the diff)
[01:10] <webben> if anyone else is willing to take a look
[01:10] <pitti> webben: maybe you can add what you have to the report, and describe the problem?
[01:10] <pitti> then everyone can comment (including myself, but not now)
[01:10] <webben> pitti: that's a good idea :), i'll have to reinstall my package to find out what the funny bug i caused was, but i'll try and attach it to the bug at some point today :)
[01:12] <pitti> webben: great
[01:12] <pitti> seb128__: does your net connection suck so hard today? :)
[01:12] <seb128__> pitti: yeah, it's not that happy today :/
[01:13] <AlinuxOS> seb128__, wifi? 
[01:13] <seb128__> AlinuxOS: no, dsl line
[01:13] <rodarvus> pitti, so, when bzr 0.9 is going to hit the buildds? :P
[01:14] <pitti> rodarvus: don't ask me; LP turnaround has been ludicrously slow in the last days
[01:14] <StevenK> The LP developers haven't been pedalling hard enough.
[01:14] <rodarvus> that means you're going to upload it soon?
[01:15] <pitti> rodarvus: erm, me?
[01:16] <rodarvus> well, you were the last uploader of bzr :)
[01:16] <pitti> meh, I just requested a sync
[01:16] <rodarvus> haha
[01:16] <pitti> OTOH, if you guys want me to, I'll package it
[01:17] <doko> infinity, cprov: please requeue gcj-4.2 on powerpc and sparc (if the chroot problems are solved)
[01:17] <pitti> traditionally, jbailey has packaed them, but he won't have much time nowadays
[01:18] <nags> seb128__, ping
[01:18] <pitti> rodarvus: ah, Debian already has 0.9rc1
[01:19] <cprov> doko: we need to check with infinity if the chroot are fixed, them I can requeue if necessary
[01:21] <pitti> http://bazaar-vcs.org/Performance/0.9 is quite impressive
[01:28] <rodarvus> very impressiv
[01:28] <rodarvus> impressive
[01:28] <ge_ubuntu> does somebody see me here? I just registered
[01:28] <pitti> ge_ubuntu: ack
[01:28] <Fujitsu> ge_ubuntu, I can't see you.
[01:30] <ge_ubuntu> I am trying to priv somebody, seems like i am not visible
[01:30] <ge_ubuntu> or they dont see my priv emesg
[01:33] <ge_ubuntu> mako: do you see me?
[01:47] <iwj> Did we know that gnome-cups-add segfaults ?
[01:47] <pitti> I saw it crashing sometimes, but never really reproducable
[01:48] <iwj> Ah, well, it's nice and reproduceable for me.
[01:48] <Treenaks> core files!
[01:48] <pitti> apport!
[01:49] <pitti> I'm checking every half an hour, hoping that this 2.6.17-6 kernel finally condescends to make it to archive.u.c
[01:49] <infinity> doko: I can look at it, no worries.
[01:50] <Hobbsee> yay, another kernel update.  and this time, I DONT HAVE TO COMPILE NDISWRAPPER FOR IT!!!!!
[01:52] <sabdfl> Hobbsee: is ndsiwrapper included now?
[01:52] <mjg59> ndiswrapper has been included for ages
[01:52] <ajmitch> more that she has a new wifi card
[01:52] <Hobbsee> sabdfl: it is, but i've always had trouble with it, never being able to make it run
[01:53] <Hobbsee> finally, some good news for the day.  or the week.
[01:53] <doko> infinity: thanks
[01:53] <sabdfl> Hobbsee: happy for you! however, we can't tell all our users "follow Hobbsee's example and go replace your hardware" :-)
[01:54] <Hobbsee> sabdfl: haha....that is true
[01:54] <Treenaks> sabdfl: sure we can :)
[01:54] <thom> pfft, it seems quite reasonable to me
[01:54] <Hobbsee> sabdfl: if you told them that, the next thing you'd be saying is "follow Hobbsee's example, and order everybody else around so that things actually get done"
[01:54] <pitti> Hobbsee: well, it feels a bit like 'please adapt your problem to our solution', but then again, hardware sucks :)
[01:55] <ajmitch> seems reasonable to replace hardware that has only proprietary binary drivers :)
[01:55] <mjg59> sabdfl: There's someone in the London office today, right?
[01:55] <pitti> ajmitch: full ack
[01:55] <Hobbsee> pitti: well, the reason i havent tried looking into fixing it that much, is that it *obviously* works for other people, as there are no bug reports with my particular problem.  and i wasnt sure if it was ndiswrapper upstream, which had problems with marvell cards for a while anyway.
[02:00] <Hobbsee> i figured it was probably user error, rather than application errorr.
[02:00] <Hobbsee> -r
[02:00] <mjg59> Don't we ship a driver for the Marvell chipset /anyway/ ?
[02:01] <infinity> We do.  And it even kinda works.
[02:01] <sabdfl> mjg59: couple of folks, yes
[02:01] <Hobbsee> mjg59: we do?  how on *earth* do you get to it?
[02:01] <mjg59> sabdfl: Ok, cool. I'll drop down and pick up the Toshibas I left behind, then
[02:01] <mjg59> Hobbsee: modprobe mrv8k
[02:01] <Hobbsee> mjg59: ahhh....
[02:02] <mjg59> Hobbsee: But it ought to be autoloaded. What does lspci look like for your card?
[02:02] <Hobbsee> mjg59: i'd have to *find* it.  it never autoloaded.  it's a netgear wg511v2 made in china card, marvell 8335 chipset
[02:03] <mjg59> Hobbsee: Then it's probably a missing PCI ID. 
[02:03] <mjg59> What's the lspci output?
[02:03] <StevenK> Hobbsee: When you say find, do you mean, unearth from the hole you buried it in?
[02:03] <Hobbsee> StevenK: no, i think it's at the bottom of my bag.
[02:03] <mjg59> Hobbsee: Oh, sorry - I thought you meant find the driver!
[02:04] <mjg59> Hobbsee: Sure, no rush
[02:05] <Hobbsee_> mjg59: 02:00.0 Ethernet controller: Marvell Technology Group Ltd. 88w8335 [Libertas]  802.11b/g Wireless (rev 03)
[02:05] <mjg59> Hobbsee_: Can you give lspci -s 02:00.0 -v -n ?
[02:05] <mjg59> Just the first two lines of output
[02:05] <Hobbsee_> 02:00.0 0200: 11ab:1faa (rev 03)
[02:05] <Hobbsee_>         Subsystem: 1385:4e00
[02:05] <Hobbsee_> mjg59: ^
[02:06] <mjg59> Now, that's interesting
[02:06] <mjg59>  {PCI_VENDOR_ID_MARVELL, 0x1faa, PCI_ANY_ID, PCI_ANY_ID, 0x6b00, 0x1385,\
[02:06] <mjg59>  W8335},
[02:06] <mjg59> I'd have expected that to get it to autoload
[02:07] <Hobbsee_> mjg59: well, it certainly doesnt :P
[02:07] <mjg59> I should really remember this stuff
[02:07] <mjg59> Hobbsee_: Anyway, it'd be great if we could work through a debug session at some point
[02:07] <mjg59> I have to head out now, though
[02:07] <Hobbsee_> mjg59: you mean you dont know everything?  shame!
[02:07] <Hobbsee_> mjg59: sure, ping me when you've got some time :)
[02:08] <Hobbsee_> or email, or whatever.
[02:09] <Hobbsee_> sudo modprobe mrv8k still doesnt show flashing lights on the card.
[02:09] <Hobbsee_> oh well.  debug session would be cool :)
[02:12] <Hobbsee> right...am i here now?
[02:13] <ajmitch> Hobbsee: somewhat
[02:14] <ajmitch> kind of you
[02:15] <Hobbsee> yeah :)
[02:15] <Hobbsee> ajmitch: no, extremely kind of me would be to use my whip, surely?
[02:15] <ajmitch> probably
[02:19] <Hobbsee> oh yeah, that's right :)
[02:19] <Hobbsee> that did get done, it didnt just stay as an idea in my head.
[02:20] <Hobbsee> wb zul 
[02:20] <zul> thanks
[02:21] <ajmitch> hi zul 
[02:21] <zul> hey ajmitch 
[02:22] <raphink> hi
[02:23] <raphink> is there a packaged build server ?
[02:23] <raphink> that gathers the scripts to deal with GPG signatures, building the packages for multiple architectures, building the repository
[02:24] <HiddenWolf> handy, even
[02:25] <raphink> what is that?
[02:25] <HiddenWolf> raphink: a joke. :)
[02:25] <raphink> ic
[02:31] <sivang> are the instructions over https://wiki.kubuntu.org/KernelCustomBuild  right for working with latest edgy?
[02:32] <zul> sivang: yep
[02:33] <sivang> zul: cool, thanks.
[02:44] <sivang> zul: specifically, is there a certain way to distinguish between custom built kernel packages and the ones provided by the repository ?
[02:45] <zul> sivang: if you use the follow the instructions in the wiki page it will create a custom kernel for you
[02:46] <sivang> zul: yes, I realized that. Asking in advance, do you know if someone can use this instructions to create a custom kernel that will have pkg naming and versioning as the repo provided ones?
[02:46] <zul> sivang: i believe so
[02:47] <sivang> zul: for https://wiki.ubuntu.com/SystemCleanUpTool , I need to find a good way to know which kernels on the system where manually built and installed, and which ones are provided by the repo
[02:48] <zul> sivang: thats a tough one that im not sure about sorry
[03:18] <thom> hrm, that's a problem
[03:18] <Keybuk> what is?
[03:19] <thom> the bug on sudo i'm just about to report
[03:19] <thom> (i just ran ntpdate on a machine, sudo now thinks the timestamp is too far in the future, and neither -k or -K will remove the timestamp)
[03:21] <thom> i'm open to suggestions short of rebooting :-)
[03:22] <Keybuk> heh, yeah, sudo hates it when you do that
[03:22] <Keybuk> there's a stamp file you can touch
[03:22] <Keybuk> I think
[03:26] <thom> oh fun, and the sudo.ws bugzilla is bust
[03:26] <thom> Keybuk: hrm, i can't see any reference to something like that in the docs
[03:27] <infinity> thom: Only root can read the docs.  Try "sudo sudo --help" :P
[03:28] <ogra> infinity, oh, was sudo patched to accept -- options now ? 
[03:28] <thom> infinity: :P lamer
[03:29] <infinity> ogra: It was a joke. :)
[03:29] <ogra> infinity, i *know* :P
[03:29] <pitti> well, --help works :)
[03:29] <infinity> pitti: Only in the sense that it tell you that it doesn't.
[03:29] <infinity> s/tell/tells/
[03:29] <pitti> (if you ignore the 'please use single character options' complaint)
[03:30] <thom> yes, it's precisely as useful as -h (aka, chocolate fireguard)
[03:31] <Keybuk> it could be worse, it could be designed by a launchpad developer
[03:31] <Keybuk> --force --force-harder --FORCE --FORCE-FORCE-FORCE
[03:32] <pitti> Keybuk: is it that bad? :/
[03:32] <infinity> LP's options are very.. Verbose.. And almost always ALL required.
[03:32] <infinity> Other than that, it's great.
[03:32] <pitti> I guess you guys have a fine collection of aliases by now :)
[03:33] <thom> i might add a FASTER_PUSSYCAT_KILL_KILL_KILL env var to sudo which'll make it kill everything
[03:33] <ogra> apt-get install slay ?
[03:33] <ogra> sudo slay root ?
[03:33] <ogra> :)
[03:33] <thom> ogra: I don't have sudo access, see above
[03:34] <ogra> ah, right
[03:34] <Keybuk> pitti: I tend to dislike multiple "force" options which only differ in case, but have different effects
[03:34] <pitti> urgh
[03:40] <Seveas> --force --dark_side_of_the_force
[03:41] <pitti> "Dark the other side is."
[03:41] <pitti> "Be quiet, Yoda, and eat your toast!"
[03:42] <ogra> *g*
[03:46] <siretart> Keybuk: Do you want to sync the new wpasupplicant from unstable, since we now have madwifi-ng in edgy?
[03:47] <Keybuk> siretart: already done it
[03:47] <siretart> ok.
[03:47] <Hobbsee> Keybuk: thanks a lot!
[03:48] <Hobbsee> w.r.t. endless lots of syncs.
[03:48] <Hobbsee> the count has finally gotten high enough now?  i poked over kdmtheme stuff last night, so feel free to do that one too.
[03:51] <bddebian> Howdy
[03:51] <bddebian> Keybuk: What does [debian multimedia]  imply?
[03:51] <Keybuk> bddebian: that it's a sync from debian multimedia?
[03:52] <bddebian> Keybuk: Oh, you dropped the sync part :-)
[03:52] <bddebian> Thx btw
[03:52] <Keybuk> Hobbsee: how do you mean?
[03:53] <gnomefreak> siretart: i borrowed one of your merges to use for learning but it needs to be synced (just letting you knwo so you dont freak out)
[03:53] <Keybuk> bddebian: *shrug* it's pretty obvious to us that it's a sync request <g>
[03:53] <Hobbsee> Keybuk: how do you mean to which bit?
[03:53] <bddebian> Keybuk: Well I'm kinda slow sometimes :-)
[03:53] <Keybuk> Hobbsee: "do that one"
[03:53] <Hobbsee> Keybuk: ah, sorry.  i'll find it for you
[03:53] <siretart> gnomefreak: sure, go ahead!
[03:54] <Keybuk> Hobbsee: you want kdetheme sync'd?
[03:54] <gnomefreak> Hobbsee: you want mine ;)
[03:54] <gnomefreak> siretart: ty but looks like it will be a week or so i got a out of office reply
[03:54] <Hobbsee> Keybuk: i want whatever we finally decided at the end of that bug report synced.
[03:54] <Hobbsee> gnomefreak: heh, no.
[03:57] <Hobbsee> Keybuk: hmmm.  looks like kcontrol-kdmtheme/kdmtheme has already been fixed, i cant see the bug for it anymore.  we want to sync the debian version, and axe ubuntu's, anyway.
[04:00] <Hobbsee> Keybuk: axe kcontrol-kdmtheme, sync kdmtheme.  that's it.  :)
[04:03] <mako> ge_ubuntu: i see you
[04:20] <pitti> is it just my current incarnation of archive.ubuntu.com that doesn't want to get linux-source-2.6.17_2.6.17-6.17 or everyone's? (the binaries were built > 8 hours ago)
[04:21] <pitti> oh, wait, they could hang in NEW
[04:21] <pitti> Keybuk: are the 2.6.17-6.17 kernel binaries in NEW?
[04:21] <ajmitch> I think they were poked through NEW awhile ago
[04:24] <infinity> pitti: No, they were rejected because BenC messed up.
[04:24] <infinity> pitti: He was meant to upload a fixed source, but hasn't yet.
[04:25] <pitti> ah, ok
[04:25] <infinity> Keybuk: Ignore pitti, I am. :P
[04:26] <bddebian> Heya infinity
[04:27] <Keybuk> infinity: I usually do
[04:27] <Keybuk> ;)
[04:29] <bddebian> HuH/
[04:29] <bddebian> Err Huh?
[04:30] <bddebian> Seen it
[04:32] <infinity> Poeple outside of Australia have seen this film?
[04:32] <infinity> Intentionally?
[04:33] <zul> yes intentionally
[04:34] <infinity> mvo: Correct me if I'm wrong, but isn't update-notifier supposed to notify me of updates?
[04:34] <thom> infinity: that's a bold assumption
[04:34] <infinity> mvo: Oh, never mind, I didn't wait long enough for it to think.  Stupid thing.
[04:37] <mvo> infinity: get a faster machine
[04:39] <infinity> mvo: Thpt.
[04:46] <Keybuk> doko: ping
[04:47] <doko> Keybuk: pong
[04:47] <Keybuk> doko: bug #54916, colour me confused
[04:47] <Ubugtu> Malone bug 54916 in Ubuntu "removal requests for dapper-proposed" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/54916
[04:47] <Keybuk> what exactly is the problem there?
[04:48] <Keybuk> you don't want openoffice.org in dapper-proposed for amd64?
[04:48] <Keybuk> but it's fine to leave it in for i386, powerpc and sparc?
[04:48] <doko> no, just not openoffice.org-evolution and openoffice.org-dev
[04:48] <Keybuk> why just those two?
[04:48] <Keybuk> that makes NO sense
[04:49] <doko> these were built from the OOo source, not the OOo-amd64 source. it DOES make sense
[04:49] <rodarvus> infinity, ping
[04:49] <Keybuk> openoffice.org-gtk-gnome | 2.0.3-3dapper3 | dapper-proposed | amd64
[04:49] <Keybuk> openoffice.org | 2.0.3-4dapper2-1 | dapper-proposed | amd64
[04:49] <Keybuk> what about those two?
[04:49] <rodarvus> it seems we are having chroot problems on powerpc -> https://launchpad.net/+builds/+build/236139
[04:50] <doko> openoffice.org is ok, built from ooo-amd64, yes openoffice.org-gtk-gnome should be removed as well, there should be an binary indep package instead
[04:51] <Keybuk> ah, that makes more sense then
[04:51] <Keybuk> ttf-opensymbol | 2.0.3-4dapper2 | dapper-proposed | all
[04:51] <Keybuk> libmythes-dev | 2.0.3-4dapper2 | dapper-proposed | amd64
[04:51] <Keybuk> what about those two?
[04:51] <doko> they are ok
[04:51] <Keybuk> ok
[04:52] <doko> all *4dapper2* packages should be ok
[04:52] <Keybuk> *nods*
[04:53] <Keybuk> it was that you didn't ask for gtk-gnome to be removed that was confusing
[04:53] <Keybuk> ummmmmm
[04:53] <infinity> rodarvus: pong.
[04:54] <infinity> rodarvus: If you're pinging about LRM, I can't really do anything about getting a fixed one in until BenC uploads a linux-source that doesn't suck.
[04:55] <lucas> Keybuk: is someone monitoring debian removals on a regular basis ? the few packages I asked you to remove were removed from debian a long time ago
[04:55] <Keybuk> lucas: no, nobody monitors those
[04:56] <infinity> elmo used to, in the bad old days.
[04:56] <Keybuk> doko: ya know, I'm not sure I can do this
[04:56] <infinity> We don't seem to have a procedure for that now.
[04:56] <rodarvus> infinity, no, I'm not pinging about LRM :)
 it seems we are having chroot problems on powerpc -> https://launchpad.net/+builds/+build/236139
[04:56] <doko> Keybuk: open a lp bug report ;)
[04:58] <infinity> rodarvus: That's not a chroot problem, it's an archive problem... Which is even worse.
[05:00] <mdke> jdub: ping
[05:09] <Gloubiboulga> seb128, ping
[05:13] <Gloubiboulga> mvo, ping as well :)
[05:16] <mvo> hello Gloubiboulga
[05:18] <Gloubiboulga> hi mvo, I've seen that you've merged gksu
[05:18] <mvo> Gloubiboulga: yes
[05:18] <Gloubiboulga> the package now builds a nautilus extension which brings all the gnome libs
[05:18] <Gloubiboulga> and it's a little problem for xubuntu...
[05:19] <mvo> Gloubiboulga: right, I guess that needs to be split out
[05:19] <Gloubiboulga> mvo, ok, I can take care of this if you want
[05:21] <mvo> Gloubiboulga: yes please!
[05:24] <mvo> Gloubiboulga: let me know if the upload needs sponsoring 
[05:24] <Gloubiboulga> mvo, not a sponsoring, but a review yes, I've never touch nautilus packages yet :)
[05:28] <rouzic> Hi all
[05:32] <doko> $ sudo debootstrap dapper /home/chroot/d2
[05:32] <doko> I: Retrieving Packages
[05:32] <doko> I: Retrieving Packages
[05:32] <doko> I: Retrieving Packages
[05:32] <doko> I: Resolving dependencies of required packages...
[05:32] <doko> I: Resolving dependencies of base packages...
[05:32] <doko> W: Failure trying to run: chroot /home/chroot/d2 mount -t proc proc /proc
[05:32] <doko> why?
[05:32] <doko> that's ia64
[05:33] <kylem> doko, don't you need to specify ports.ubuntu.com as the mirror?
[05:35] <Spads> proc proc proc
[05:35] <doko> kylem: yes, I remember ... but the error message ... is ... interesting ...
[05:36] <kylem> yeah, i saw that too while rebuilding some chroots. :\
[05:38] <bddebian> w00t, Keybuk r0x!!
[05:39] <Toadstool> Keybuk: about the gnomebaker sync, the only remaining difference with the current debian version is a tiny modification for the desktop file which is not worth a -XubuntuY version to my mind
[05:39] <Toadstool> hey bddebian 
[05:39] <bddebian> Hi Toadstool
[05:39] <Keybuk> Toadstool: what's the modification?
[05:39] <Keybuk> bddebian: for?
[05:40] <Toadstool> Keybuk: Icon=/full/path -> Icon=icon.png, both validate with desktop-file-validate
[05:44] <bddebian> Keybuk: Everything :-)
[06:19] <Keybuk> damn, emacs really hates __attribute__ :-/
[06:22] <Spads> heh, feh fullscreen has the same problem
[06:22] <infinity> Now, do I blame gnome-panel, metacity, or mplayer?
[06:23] <Keybuk> infinity: seb
[06:23] <Keybuk> I find it more gratifying to blame him than the software
[06:33] <Amaranth> infinity: metacity's fullscreen code got some changes recently, i guess it's to blame
[06:34] <infinity> Amaranth: Well, poo.  Any clues where to start looking?
[06:35] <Amaranth> xprop | grep _NET_WM_STATE
[06:36] <Amaranth> '[if]  it shows "_NET_WM_STATE_FULLSCREEN" as one of the states and then focus that window and observe it to not be on top of all other windows (including panels), then it sounds like you've discovered a bug in Metacity.  If so, I would be interested to learn how to reproduce so that I can fix it.'
[06:37] <Amaranth> Basically they did some things to fix it for WINE and it must have broken other apps.
[06:40] <infinity> Amaranth: I assume this is going to require another machine and an SSH session.  Which means getting up off the couch.
[06:40] <Amaranth> hmm
[06:41] <Amaranth> no, you have to click on the window after running that to make it spit out anything
[06:41] <Amaranth> only the panels go on top of it, nothing else can?
[06:42] <Amaranth> http://cvs.gnome.org/viewcvs/metacity/src/window.c?r1=1.442&r2=1.443 <--what changed
[06:44] <Amaranth> btw, totem and vlc still work just fine ;)
[06:47] <kozz> is there a page somewhere describing all packages that are included on the cd/dvd?
[06:47] <Amaranth> infinity: Can you try reverting that change and seeing if you still have problems?
[06:47] <kozz> who decides which packges are to be included?
[06:48] <LaserJock> kozz: that is done by seeds
[06:48] <kozz> seeds?
[06:48] <infinity> Amaranth: Yeah, just the top panel, in fact.
[06:48] <Amaranth> anyone on the ubuntu-core-dev launchpad team can change the seeds, apparently
[06:48] <LaserJock> kozz: https://wiki.ubuntu.com/SeedManagement
[06:48] <infinity> Amaranth: But I'm far too drunk to debug anything right now.
[06:48] <Amaranth> infinity: alright :)
[06:48] <Amaranth> infinity: Just use totem then. ;)
[06:49] <infinity> Amaranth: Will you be around on the weekend?  I may want to debug this when sober.
[06:49] <Amaranth> I should be around part of the time.
[06:49] <infinity> Amaranth: Kay, cool.  I'll ping you later.  I'm GNOME ignorant, and preder to stay that way.
[06:49] <kozz> LaserJock: thanks for pointing me in the right direction
[06:50] <infinity> prefer too.
[06:50] <LaserJock> kozz: np
[06:53] <kagou> hi
[06:56] <ge_ubuntu> hi
[07:03] <rouzic> Hi kagou :)
[07:04] <KhanReaper> I am trying to write a package that provides a wrapper for the /usr/bin/firefox script. I placed a dpkg-divert in the new package's preinst script, but dpkg complains that the replacement firefox script already exists in the firefox package. What can I do to overcome this? The dpkg-divert call works successfully, but why is this happening? Is there any way around this?
[07:04] <KhanReaper> By the way, there is a very good reason that this is being handled via a Debian package.
[07:06] <Chipzz> KhanReaper: why?
[07:08] <sivang> Keybuk: You use emacs ?
[07:08] <Keybuk> sivang: yes
[07:09] <sivang> Keybuk: did you notice that was a font change in one of the previous edgy updates? Do you know how it can be put back to the normal default one? or is this smaller font the new default?
[07:09] <KhanReaper> It would be waste of time to discuss it here, as the reasons are very exhaustive. Simply put, there are some problems with a plugin that Firefox uses, and it needs some special modifications to accommodate this special plugin. This is used on about 100+ workstations, and the configurations for these machines need to be kept in-sync via a local Debian package repository. As I said, this is a waste of time to discuss.
[07:10] <sladen> sivang: fixed font size change?
[07:10] <KhanReaper> Shouldn't a diversion, mark that file's name as free for when a new package is installed?
[07:10] <kagou> who can i ask for doing a sync for new debian blender package ?
[07:10] <Keybuk> sivang: "font change" ?
[07:11] <sivang> Keybuk: ah, you're not using Xemacs then.
[07:11] <Keybuk> right
[07:11] <sivang> Keybuk: I'll try that as well :-)
[07:11] <sladen> sivang: I'm not clear what you're doing, but could it be related to which aliases file is used?  bug #54809
[07:11] <Ubugtu> Malone bug 54809 in xfonts-utils "X server cannot find default font `fixed'" [Untriaged,Confirmed]  http://launchpad.net/bugs/54809
[07:13] <Kaleo|work> rodarvus: hmmm
[07:13] <Kaleo|work> I still have problems with i810
[07:14] <rodarvus> Kaleo|work, latest i810 is not published yet
[07:14] <rodarvus> (though it is uploaded)
[07:14] <Kaleo|work> ok
[07:14] <Kaleo|work> sorry I was reading that in your comment
[07:15] <sivang> sladen: my server doesn't crash though
[07:17] <sladen> sivang: ignore the bug report itself.  look at the notes about the fonts.aliases file and which is used
[07:18] <rodarvus> sladen, nice, thanks for tracing the root of the fonts issue
[07:18] <sladen> sivang: one thing I did notice is that I had differently sized fonts afterwards
[07:18] <rodarvus> sladen, do you want to update xfonts-base?
[07:19] <sladen> rodarvus: what's your thoughts, _just_ fixing xfonts-base shoulds fix it for all intents and purposes---you happy if I just do a fresh upload of that?
[07:19] <rodarvus> sure, seems just fine for me
[07:19] <rodarvus> maybe other font packages might need it too
[07:19] <rodarvus> alternatively, have a versioned depends on xorg could do it.
[07:20] <sladen> rodarvus: yeah, but they won't actually *break* the X server
[07:23] <sladen> rodarvus: should it actually be pre-depends?
[07:23] <rodarvus> it should, but debian policy really really hates pre-depends
[07:30] <nags> seb128, hi
[07:30] <nags> seb128, we got some review comments regarding gedit scripts from gedit authors
[07:30] <seb128> re
[07:31] <seb128> ah, nice comments?
[07:31] <nags> seb128, we are incorporating them
[07:31] <nags> seb128, yes :)
[07:31] <seb128> cool
[07:31] <seb128> I tried to make it working
[07:31] <nags> seb128, we would like to hear from you too :)
[07:31] <seb128> but it does nothing on my box
[07:31] <nags> seb128, oops
[07:31] <seb128> maybe I'm not using it rightly
[07:31] <sivang> sladen: I also get Warning: locale not supported by Xlib, locale set to C
[07:31] <sivang> sladen: btw
[07:32] <nags> seb128, infact all those were developed in Ubuntu box
[07:32] <seb128> I don't have ldtp installed to my system, I run if from the srcdir, I'm not sure if that's right
[07:32] <nags> seb128, enabled accessibility ?
[07:32] <seb128> no
[07:32] <nags> seb128, ldtp has to be installed and accessibility should be enabled
[07:32] <seb128> it needs to be installed? 
[07:32] <seb128> or from the srcdir is fine?
[07:33] <sladen> sivang: sounds like encodings isn't being setup ---could be related
[07:35] <nags> seb128, installed
[07:35] <nags> G0SUB, hi
[07:36] <sladen> rodarvus: in the spirit of blows-your-X-server-away I've uploaded it with and I'll wait for the complaints to come in
[07:37] <rodarvus> :)
[07:38] <rodarvus> sladen, btw, I have uploaded new libdrm, mesa, xserver-xorg and xserver-xorg-video-i810 this morning, you might be intersted on it
[07:38] <rodarvus> (since you seem to care about mesa and i810)
[07:38] <seb128> nags: is there a ldtp package somewhere?
[07:39] <nags> seb128, yes, based on cr3 advice Casanova created it
[07:39] <nags> Casanova, Ubuntu LDTP deb
[07:39] <Casanova> nags: the deb?
[07:39] <Casanova> ah ok :)
[07:39] <nags> Casanova, yes
[07:40] <nags> Casanova, link ?
[07:40] <seb128> nags: http://download.freedesktop.org/ldtp/0.x/0.5.x/ldtp_0.5.0-2_i386.deb ?
[07:40] <seb128> I've just found that with google
[07:40] <Casanova> http://people.freedesktop.org/~prashmohan/latest/
[07:40] <Casanova> seb128: the source package is in the above link :)
[07:40] <seb128> ok, thank you
[07:40] <seb128> I tried to build from the ldtp src dir I have but the deb had no content out of the documentation of the package
[07:41] <iwj> Oh, how amusing, this thing labelled as `breezy' in my grub is in fact a half-completed install.  I've just booted it and it's `Installing packages' :-).
[07:41] <nags> Casanova, for running addict3d / narasim_7 scripts, do we need to enable accessibility explicitly ?
[07:41] <nags> seb128, ah okay
[07:41] <Casanova> seb128: i dont get you
[07:41] <Casanova> nags: explicitly in the sense?
[07:42] <seb128> Casanova: I runned debuild to http://people.freedesktop.org/~prashmohan/latest/ldtp-0.5.0/ copied on my disk
[07:42] <seb128> Casanova: and the .deb only had /usr/share/doc/ldtp, no binary
[07:42] <Casanova> oops
[07:42] <Casanova> seb128: i use dpkg-buildpackage -rfakeroot
[07:42] <seb128> should be the same
[07:42] <Casanova> i guess
[07:42] <seb128> anyway just installed the deb from where you pointed
[07:42] <nags> Casanova, we set from gnome-control-center right ? instead just launchapp ('gedit', 1) ?
[07:43] <Casanova> i am building gnome now... i will try running debuild on it after that
[07:43] <Casanova> nags: the scripts start gedit using launchapp
[07:43] <Casanova> nags: yes
[07:43] <nags> Casanova, cool
[07:43] <nags> Casanova, also narasim_7 here ?
[07:44] <nags> Casanova, maybe he can directly get comments from seb128 :)
[07:44] <seb128> bah, still doing nothing
[07:44] <Casanova> seb128: doing nothing in the sense?
[07:44] <seb128> I run "ldtprunner gedit-main.xml" but it stays on the command doing nothing
[07:44] <nags> seb128, Casanova did jhbuild + LDTP integration and Evolution automation as part of Google SoC
[07:44] <seb128> Casanova: like no gedit opened, no message, no action, no CPU used, ... :)
[07:44] <nags> seb128, narasim_7 who has developed those 70 diff scenarios for gedit with his friend addict3d
[07:45] <Casanova> oh oh
[07:45] <Casanova> this is bad
[07:45] <seb128> nags: ah, ok :)
[07:45] <narasim_7> seb128: hai
[07:45] <seb128> hey narasim_7
[07:45] <Casanova> nags: same issue frederic was facing?
[07:45] <seb128> but I don't know how to use that thing
[07:45] <seb128> should I opened gedit first?
[07:45] <Casanova> seb128: could why try 1 thing?
[07:45] <Casanova> seb128: no
[07:45] <seb128> or is that supposed to be done by the script?
[07:45] <seb128> Casanova: sure
[07:45] <Casanova> seb128: could you start `ldtp' in a terminal and run `ldtprunner gedit-main.xml' in another terminal?
[07:46] <nags> Casanova, ah okay
[07:46] <Casanova> seb128: before that
[07:46] <Casanova> seb128: `export LDTP_DEBUG=1' in both terminals
[07:46] <nags> Casanova, maybe LDTP_DEBUG=2
[07:46] <nags> it will give more verbose info
[07:47] <narasim_7> gnome-accessibility
[07:47] <Casanova> narasim_7: ?
[07:47] <narasim_7> shuldnt that be exported that first
[07:47] <Casanova> narasim_7: that shouldnt be necessary
[07:48] <Casanova> if accessibility is enabled
[07:48] <seb128> http://paste.ubuntu-nl.org/20310
[07:48] <seb128> that's the log
[07:48] <seb128> from the ldtprunner 
[07:48] <narasim_7> Casanova: ohh.. okie
[07:48] <seb128> ldtp has nothing printed
[07:48] <Casanova> seb128: did you export the variable?
[07:48] <Casanova> on the ldtp terminal too?
[07:48] <seb128> yep
[07:48] <seb128> but =1
[07:48] <nags> seb128, try =2
[07:49] <seb128> ah
[07:49] <Casanova> nags: hmm connection wasnt made even?
[07:49] <seb128> there we go :)
[07:49] <nags> Casanova, actually =1 gives min info, that too in client only
[07:49] <Casanova> ah
[07:49] <nags> Casanova, and =2 gives verbose info
[07:49] <seb128> Packet received from client is not valid
[07:49] <Casanova> seb128: you started `ldtp' before running `ldtprunner' right?
[07:49] <nags> Casanova, recent change, based on dobey's request
[07:49] <seb128> Cliend disconnected
[07:49] <seb128> Casanova: yep
[07:50] <Casanova> seb128: is accessibility enabled?
[07:50] <seb128> how do I check?
[07:50] <seb128> I've clicked on the box for that and restarted the session
[07:50] <seb128> so it should
[07:51] <Casanova> seb128: System --> Preferences --> Assistive Technology Support
[07:51] <seb128> just the support option
[07:51] <seb128> no app
[07:52] <narasim_7> seb128: perhaps some more servers of ldtp are running?
[07:52] <Casanova> seb128: thats ok
[07:52] <Casanova> seb128: killall -KILL ldtp
[07:52] <Casanova> seb128: http://paste.ubuntu-nl.org/20311
[07:55] <seb128> Casanova: client connection: accepted
[07:55] <seb128> Locale languages: fr_FR.UTF-8
[07:55] <Casanova> ah :)
[07:55] <Casanova> close python
[07:55] <Casanova> and run the script now?
[07:55] <seb128> when I close I get the
[07:55] <seb128> "Packet received from client is not valid"
[07:56] <Casanova> thats ok
[07:56] <nags> narasim_7, Casanova, seb128, we have some issues in localized version as of now...
[07:56] <seb128> lot and lot of log to the ldtp but nothing at screen
[07:56] <nags> prem is working on that
[07:56] <nags> seb128, that will be fixed in our next release
[07:57] <seb128> do I need some a11y app?
[07:57] <nags> seb128, hmmm
[07:57] <nags> seb128, no
[07:57] <nags> seb128, you need to try them in default locale - English
[07:57] <nags> then things will work fine
[07:59] <nags> seb128, in GNOME metacity, if I want to bring a window to foreground, can we do something ? I mean write a small program etc ? We tried with gok code, but we were not able to succeed
[07:59] <nags> Casanova, narasim_7, maybe we need to update the doc, that it can be run in English lang only as of now ?
[08:00] <Casanova> i knew only now ;)
[08:00] <seb128> nags: right, that works now ;)
[08:00] <narasim_7> nags: Srini said he would take up updating the doc once gedit was over
[08:00] <narasim_7> nags: i promised him my help
[08:00] <nags> narasim_7, cool
[08:01] <nags> seb128, wow !
[08:01] <nags> narasim_7, you are going to get good review comments from seb128 :)
[08:01] <seb128> nags: I'm not good with metacity focus stealing prevention things, /j #bugs on GIMPnet and ask to elijah, he's the metacity maintainer and reply to questions in a detailled and friendly way usually
[08:01] <narasim_7> nags: :-)
[08:02] <seb128> ok, guy, thank you for giving a hand to set that up ;)
[08:02] <nags> seb128, ya, but these days could not get in touch with him
[08:02] <seb128> I've to go for dinner now but I'll play with that and let you know how it works for me
[08:02] <nags> seb128, wow !
[08:02] <nags> seb128, sure
[08:03] <seb128> nags: he spoke 20min ago on #bugs so he might be around ;)
[08:03] <seb128> nags: better to ask what you have to ask than just "ping" though, some people tend to not reply to ping without context when they are busy because they don't know if they are interested to start a discussion on an unknow topic ;)
[08:03] <nags> seb128, let me check now
[08:03] <seb128> anyway I've to run
[08:03] <seb128> thank you again
[08:03] <seb128> bbl
[08:04] <nags> seb128, ya right :)
[08:44] <Royal-team> 190 personnes et personne parle ?
[08:44] <Royal-team> Mdr
[08:48] <BenC> linux-source-2.6.17: edgy ia64   Chroot problem
[08:48] <BenC> in case anyone wants to know
[08:59] <infinity> BenC: Already fixed.
[09:17] <bluefoxicy> So if i wanted to suggest something really dumb
[09:18] <bluefoxicy> like say
[09:18] <bluefoxicy> Allowing ShipIt users to request CD-RWs instead of real pressed CDs
[09:18] <bluefoxicy> where would I suggest this?
[09:18] <bluefoxicy> (Ubuntu 6.06.1?  What am I supposed to order new CDs or something?)
[09:18] <Keybuk> I'm not entirely sure that is possible ... the CDs are pressed by a CD distribution plant
[09:18] <mjr> not gonna happen
[09:18] <Keybuk> I doubt they can even do CD-RW
[09:19] <bddebian> How about asking in #IwannauseGentoo
[09:19] <Keybuk> and you can get all of 6.06.1 from the archive as usual
[09:23] <Surak> I noticed that cdimage.ubuntu.com has only 6.06.1 DVD images.
[09:23] <Chipzz> I know this is most likely more an upstream issue than an ubuntu issue
[09:24] <Chipzz> but is there a reason why the network applet in gnome-system-tools doesn't show my profiles anymore?
[09:24] <Chipzz> are we going to switch to the network-manager applet for edgy?
[09:32] <seb128> Casanova: is the ldtp script supposed to give any sort of output or just use gedit?
[09:33] <Casanova> seb128: in the gedit-run.xml file there is a tag called <logfile>
[09:33] <Casanova> the output is an xml file specified within that tag
[09:33] <seb128> ah, ok
[09:34] <seb128> is there some documentation on ldtp or some sort of faq?
[09:34] <Casanova> seb128: http://ldtp.freedesktop.org/user-doc/index.html
[09:34] <Casanova> seb128: http://ldtp.freedesktop.org/wiki/FAQ
[09:34] <seb128> thank you
[09:34] <Casanova> seb128: http://ldtp.freedesktop.org/wiki/LDTPEditor/GeneratingLDTPDataXMLfileformat http://ldtp.freedesktop.org/wiki/LDTPEditor/GeneratingldtprunnerXMLfileformat
[09:35] <seb128> that should be enough, thank you :)
[09:35] <Casanova> hehe
[10:52] <pygi> sivang, poke
[11:03] <jdahlin> doko: ping
[11:08] <doko> jdahlin: pong
[11:08] <jdahlin> doko: what's the primary use case for linking python dynamically instead of statically?
[11:09] <jdahlin> this number 6% worries me a little bit
[11:09] <doko> jdahlin: ask jamesh :-)
[11:09] <jdahlin> doko: then I probably know what it is
[11:10] <jdahlin> being able to use a library from python which has python plugins
[11:10] <doko> yeah, you need to avoid two copies of the interpreter, which can happen, if things get dynamically loaded
[11:10] <jdahlin> to avoid starting two interpreters
[11:10] <doko> exactly
[11:10] <jdahlin> however, it's possible to avoid that by dlopening libpython
[11:10] <jdahlin> which is what nautilus-python does
[11:11] <jdahlin> gstreamer and gnome-vfs are two real uses cases otherwise
[11:11] <doko> or building python as a PIE, which I didn't try yet
[11:11] <jdahlin> both of them have python bindings and python plugin support
[11:18] <jamesh> jdahlin: if you have two pieces of code that have linked Python statically (or one static + one dynamic), there are two copies of the Python interpreter
[11:19] <jamesh> building python as a PIE is a way to make /usr/bin/python the same as /usr/lib/pythonX.Y.so
[11:20] <jdahlin> jamesh: right, but are there any important libraries or program that does that apart from /usr/bin/python ?
[11:20] <jdahlin> and should you not avoid linking to libpython.so directly just because of that?
[11:20] <jdahlin> [for extension modules] 
[11:20] <jamesh> jdahlin: ideally everything would link to libpython.so (including python)
[11:21] <jamesh> although there is a performance penalty
[11:21] <jdahlin> jamesh: ideally because it's good practice or because it creates real world problems?
[11:23] <jamesh> jdahlin: because it avoids the problem all together.  If everything links to a dynamic libpython, then there is only one copy of the interpreter
[11:23] <jdahlin> jamesh: right, but it is possible to workaround it
[11:23] <jamesh> really?
[11:23] <jdahlin> yeah
[11:24] <jdahlin> so, in the case of gst-python
[11:24] <jdahlin> which can be loaded by /usr/bin/python and by gstpythonloader through a normal C application
[11:25] <jdahlin> in the case of /usr/bin/python you can access the symbols directly
[11:25] <jdahlin> for the gst plugin you can check if it's available dlsym(dlopen(NULL), "Py_None")
[11:25] <jdahlin> and if it's not just dlopen(/usr/lib/libpython.so)
[11:25] <jamesh> ugly hack ...
[11:26] <jdahlin> I found it quite neat ;-)
[11:27] <jdahlin> but the use case is limited, I don't think any real applications has actually run into this
[11:28] <jamesh> what if another copy of the interpreter gets loaded without RTLD_GLOBAL?
[11:28] <jamesh> your solution works in a limited number of situations (arguably a useful number of situations though)
[11:29] <jdahlin> wouldn't that be a bug?
[11:29] <jdahlin> in the specific library or application
[11:29] <jamesh> it is a bug that can occur because there is a static libpython :)
[11:30] <jdahlin> true
[11:31] <jdahlin> but why would anyone do that anyway?
[11:34] <jamesh> jdahlin: people do all kinds of crazy shit.
[11:35] <jamesh> jdahlin: anyway, I need to pack up for my flight tomorrow
[11:35] <jdahlin> jamesh: good luck with the laptop
[11:35] <jamesh> thanks