[12:08] <Keybuk> Trae: that looks correct to me
[12:08] <sladen> Keybuk: I was thinking of using /dev/input to draw a moveable mouse cursor.  The Mac has one and it's good UI feedback to the user that the system hasn't actually crashed
[12:08] <Trae> Keybuk, nub nub :)  danke
[12:09] <Keybuk> sladen: sick
[12:09] <Trae> Keybuk, ok, let me reboot and try it... and see if I can get this thing to crash
[12:09] <Trae> *chuckle*
[12:09] <Keybuk> but yes
[12:09] <sladen> Keybuk: Maclike!
[12:09] <Keybuk> you can play "bounce the mouse cursor around the screen"
[12:09] <Keybuk> which is my favourite hourglass game
[12:09] <Trae> bbiab gang
[12:10] <Keybuk> I'm guessing by the sudden requirement of an INPUT option that usplash somehow displays on something *other* than /dev/console ?
[12:10] <Keybuk> mjg59, Seveas ^ ?
[12:10] <Seveas> Keybuk, -ENOIDEA, the INPUT isn't my work
[12:10] <mjg59> Keybuk: Nor mine
[12:11] <Keybuk> who did the INPUT stuff?
[12:11] <Seveas> (although I just patched it to support INPUTENTER for "Press enter to reboot" messages
[12:11] <mjg59> But we've always run usplash with -c, which changes it away from the active terminal
[12:11] <mjg59> Keybuk: Read the changelog :p
[12:11] <Seveas> David Hrdeman implemented it according to debian/changelog
[12:11] <Keybuk> mjg59: it's not mentioned
[12:12] <Keybuk> mjg59: active terminal != /dev/console
[12:12] <Kamion> gosh, d-i built with console-setup
[12:12] <Kamion> now, will it even work a little bit
[12:12] <Keybuk> mjg59: ah, yes it is, and it has your name by it
[12:12] <mjg59> Keybuk: I really don't know what you're talking about
[12:12] <Keybuk> revno 40
[12:12] <mjg59> No, about /dev/console
[12:12] <Keybuk> well
[12:13] <Keybuk> what I'm trying to work out is why usplash suddenly needs an INPUT method
[12:13] <Keybuk> why can't they just type as they always did, and whatever script was running get the key presses?
[12:13] <mjg59> So that it can be done graphically
[12:13] <Keybuk> "done graphically" ?  the typing bit
[12:13] <mjg59> Yes
[12:13] <Keybuk> right, but this is damned hard to deal with
[12:13] <Keybuk> the script would have to do something like
[12:13] <mjg59> Debian wanted it
[12:14] <Keybuk> if usplash_is_running; then
[12:14] <mjg59> We don't need to care
[12:14] <Keybuk>    usplash_input ...
[12:14] <Keybuk> else
[12:14] <Keybuk>    read ...
[12:14] <Keybuk> fi
[12:14] <mjg59> But I still don't understand your comments about /dev/console
[12:14] <_ion> . /some/file
[12:14] <Seveas> Keybuk, or use someting like log_end_msg, but for inout
[12:14] <_ion> read_maybe_usplash ...
[12:14] <Seveas> where log_end_msg checks for usplash
[12:14] <_ion> (Why am i still awake?)
[12:14] <Keybuk> mjg59: so the old usplash ran in a virtual terminal (tty8)
[12:14] <Keybuk> which happened to be framebuffered
[12:15] <mjg59> Keybuk: It still does run in a virtual terminal
[12:15] <Keybuk> right
[12:15] <Kamion> if it's David Hrdeman, chances are it was for cryptsetup
[12:15] <Keybuk> and usplash doesn't read() from that terminal?
[12:15] <mjg59> No
[12:15] <Keybuk> ok, it's probably fine then
[12:15] <Seveas> I saw getchar() in the code
[12:15] <Keybuk> that means any script read()ing from /dev/console (ie. any init script) would win and get the chars
[12:17] <Keybuk> probably /dev/console, not whatever-tty-it-changes-to :p
[12:17] <Trae> ok.... let's give this joker a shot
[12:18] <sladen> Keybuk: it may well be, ogra was having issues with it on ltsp machines?
[12:19] <Seveas> Keybuk, sprintf(/dev/tty%d, vtnumber)
[12:19] <Trae> hmm I get no volume.... I did fuser -k /dev/dsp
[12:19] <Seveas> (if I interprete the code correctly)
[12:20] <Keybuk> Seveas: ah yes, somebody's fixed that then
[12:20] <Seveas> close(STDIN_FILENO);
[12:20] <Seveas> open(foo)
[12:20] <Seveas> will that open foo as stdin?
[12:20] <Keybuk> yes
[12:20] <ogra> sladen, my prob is rather the weird handling of start/stop in the usplash initscript
[12:20] <Seveas> then it's indeed /dev/tty%d
[12:20] <Keybuk> open always returns the lowest numbered unused file descriptor
[12:20] <Keybuk> if you've just closed stdin, that will always be zero
[12:21] <ogra> apart from that usplash is perfectly fine on ltsp
[12:21] <Seveas> Keybuk, what's more useful: INPUTENTER, which waits specifically for \n or INPUTKEY which waits for the any key?
[12:22] <Keybuk> where's the any key? :p
[12:22] <Seveas> 
[12:22] <Seveas> for cases like "press .... to reboot"
[12:22] <_ion> keybuk: It's the : http://johan.kiviniemi.name/pictures/kb/naeppaeimistoe02.jpeg
[12:24] <Trae> is there a reason why I would be able to play mp3 and ogg with Rhythmbox and not have sound with flash?  I don't have RB running, and I did fuser -k /dev/dsp
[12:26] <Seveas> mjg59, new revision pushed (INPUTENTER and usplash_down)
[12:26] <mjg59> Seveas: Thanks
[12:30] <Trae> Keybuk I'm getting no audio, but I'll let it play to see what happens.
[12:30] <Trae> that's a different problem for a different time
[12:30] <Trae> heh
[12:41] <Trae> Keybuk nope... it powered off
[12:41] <Trae> :(
[12:42] <Trae> Keybuk I like how gentoo does thekernel mesgs
[12:42] <Trae> heh
[12:42] <Trae> hmm
[12:42] <Trae> maybe I'm thinking about something else
[12:46] <Keybuk> how do gentoo do the kernel messages?
[12:46] <mjg59> Keybuk: They have a hacked kernel
[12:46] <mjg59> But anyway
[12:47] <mjg59> Keybuk: Uhm. I'm not convinced that the problem you're describing "usplash results in my tty being invisible" is the same as "upstart doesn't always seem to start a getty on tty1"
[12:47] <Seveas> I don't like "hey, let's put jpg decoding in the kernel" at all
[12:47] <Keybuk> mjg59: the user reported the former 
[12:47] <mjg59> Oh well
[12:48] <Keybuk> I've just also seen the latter "go away" if usplash isn't used
[12:48] <mjg59> Keybuk: Given that usplash is running on tty8, I really don't see how it could have any direct influence
[12:48] <Keybuk> nonetheless, it seems to
[12:49] <mjg59> Well, how is this usplash's fault?
[12:49] <sladen> Keybuk: how just upstart /manage/ to fail to start a getty on tty1?
[12:49] <Keybuk> I don't know, yet if you turn usplash off, it seems to work just fine
[12:49] <mjg59> If upstart launches gettys on everything other than tty1, how is an application on tty8 influencing it?
[12:50] <Keybuk> mjg59: ah, but getty is running on tty1 as well
[12:50] <Keybuk> something nukes it
[12:50] <Seveas> mjg59, something in restore_console?
[12:50] <Keybuk> and not in a "kill" way, but in a "fucks it up so it wedges" way
[12:50] <mjg59> Doesn't happen with sysvinit
[12:50] <Keybuk> I've seen it happen with sysvinit also
[12:51] <Keybuk> which is where I stopped believing it was definitely upstart
[12:52] <Seveas> usplash does nothing to tty1 but VT_GETSTATE and trying to switch back to it -- maybe svgalib or a video driver f*s up during the switch back?
[12:53] <mjg59> Seveas: Wouldn't kill getty
 and not in a "kill" way, but in a "fucks it up so it wedges" way
[12:53] <mjg59> Still won't matter
[12:53] <Keybuk> mjg59: it doesn't _kill_ getty, is the point
[12:53] <Keybuk> getty is still running
[12:53] <Keybuk> and still select()ing on tty1
[12:53] <Keybuk> just getting nothing
[12:56] <mjg59> Seveas: FTBFS
[12:56] <mjg59> helvB10.c:2:27: error: usplash-theme.h: No such file or directory
[12:56] <mjg59> Looks like it needs a -I
[12:56] <Seveas> odd
[12:56] <Seveas> what were you trying to build?
[12:57] <mjg59> dpkg-buildpackage
[12:57] <Seveas> hmmm more
[12:57] <Seveas> as that works here
[12:57] <mjg59> Seveas: You've presumably got a usplash-theme.h in /usr/include?
[12:57] <Seveas> urgh, yes
[12:57] <mjg59> Fixed here
[12:57] <Seveas> if only pbuilder wasn't too slow to use it for all tests
[12:58] <Seveas> ok, thanks
[12:58] <Seveas> mjg59, would vga_setmode(TEXT); in usplash_done() hurt?
[12:58] <mjg59> Seveas: Just added
[12:59] <mjg59> It's actually slightly implicit due to the crack way that svgalib works, but
[01:11] <mjg59> Seveas: Themes need to be built -fPIC
[01:12] <Seveas> ok
[01:12] <mjg59> Seveas: Other than that, all looks *very* good
[01:12] <Seveas> good, that means my programming skills are still present
[01:13] <Seveas> (/me actually does glibc ld.so hacking for work and graduation project)
[01:13] <mjg59> Seveas: How's the different png sizes handled?
[01:14] <Seveas> if no -x / -y given, it inspects the image struct to select a mode
[01:14] <mjg59> Ok
[01:14] <Seveas> if -x -y given, then usplash will fallback to testcard if image is too big
[01:14] <mjg59> Right
[01:14] <mjg59> Do we have support for images of multiple sizes?
[01:14] <Seveas> if you call "create the same theme a few times" then yes ;)
[01:15] <Seveas> but that would be a nice edgy+1 feature
[01:15] <mjg59> Well, and have it as a single file on the filesystem?
[01:15] <mjg59> Erm
[01:15] <mjg59> I think it's actually an edgy *necessity* :)
[01:15] <Seveas> hmm
[01:16] <sladen> multiple images.  Just grab the biggest one that fits
[01:16] <Seveas> shouldn't be too hard to implement (allow several struct_themes to be present in eg a linked list), but I need very strong coffee to finish that before FF
[01:16] <Keybuk> especially if we could have 16:9 14:9 4:3 etc. in the same theme
[01:16] <mjg59> Seveas: I've just uploaded this, so if on some machines it doesn't work properly, that's clearly a bug
[01:17] <Seveas> mjg59, how about the ppc bits?
[01:17] <mjg59> So we have time to fix it :)
[01:18] <mjg59> Seveas: What about them?
[01:18] <Seveas> well, they're not there yet (modesetting in bogl and having bogl put 256 colors on screen) unless either magic happened or I misunderstood you earlier
[01:19] <mjg59> Oh, right
[01:19] <mjg59> Well, that's just a matter of
[01:19] <mjg59> a) Adding a bogl_setmode (just needs to ioctl the framebuffer)
[01:19] <mjg59> b) Fixing bogl not to care about there being more than 16 colours
[01:21] <mjg59> Yeah, I can probably manage that
[01:21] <Seveas> although fixing bogl for more colors may have happened through magic
[01:21] <Seveas> because I simply ripped out runlenght crack and made it not do &/| tricks on colors
[01:22] <Seveas> (see the diff for the first commit on my branch)
[01:23] <Seveas> I'll work on multiple-themes-per-file then adding a struct usplash_theme* next right after the version and code to figure out ratio (if x/y given on command line) and biggest fitting image. If no x-y given it will use first image available or lowest res (which would you prefer?)
[01:24] <Seveas> s/then/then,/ for readability of that line
[01:25] <mjg59> So it's not clear to me /why/, but
[01:25] <mjg59> Running usplash now adds 40 seconds to my boot time
[01:26] <sladen> t'fsck what
[01:26] <sladen> that's almost up there with the 58 seconds that redsplash was adding to the Fedora boot process
[01:26] <mjg59> Probably because it's using 100% of the CPU
[01:27] <Seveas> mjg59, the setitimer/signal crack for animation?
[01:27] <sladen> while(getchar(0))?
[01:27] <mjg59> It's selecting insanely fast
[01:27] <Seveas> hmm
[01:27] <mjg59> select(4, [3] , NULL, NULL, {0, 0})      = 0 (Timeout)
[01:27] <mjg59> Yeah, unsurprising
[01:27] <Seveas> ah
[01:27] <Seveas> lol
[01:27] <Seveas> my bad, that should be an insanely high value
[01:27] <Seveas> I set it to 0 to investigate an unrelated (and already fixed) bug
[01:28] <sladen> ETIMEOUTTOOSMALL
[01:28] <Seveas> ESTUPIDSEVEAS
[01:30] <mjg59> Seveas: Hm. No, it seems to be 40000
[01:31] <Seveas> 40000 microseconds
[01:31] <mjg59> That's 1/25 of a second, isn't it?
[01:31] <Seveas> yes
[01:31] <Seveas> at first I tried to get the animation integrated with the select loop
[01:31] <Seveas> horror
[01:31] <mjg59> But it's firing much more often than that
[01:32] <mjg59> Or is it just that the timeout is never reset?
[01:32] <Seveas> it's set once
[01:32] <Seveas> so inded
[01:32] <mjg59> Ok
[01:32] <Seveas> fallout of that horror attempt to not use signals
[01:32] <mjg59> So what does it need to be?
[01:33] <Seveas> 15
[01:33] <Seveas> or actually: timeout
[01:33] <mjg59> Right
[01:33] <Seveas> because the TIMEOUT command changes it
[01:33] <mjg59> And no framedelay?
[01:33] <Seveas> there's another thing still missing
[01:33] <Seveas> I'll create a quick patch and pastebin it
[01:33] <Seveas> hang on
[01:35] <mjg59> Seveas: Nah, sorted
[01:36] <Seveas> could you poke me when you push to LP so I can merge again
[01:36] <mjg59> Just done
[01:36] <Seveas> ok, then I just need to wait for $launchpad_delay 
[01:36] <mjg59> Seveas: Oh, you're using http to get it?
[01:37] <Seveas> yes, I don't think I can reach it via sftp as I am not a member of core-dev (or dev for that matter)
[01:37] <mjg59> Yeah
[01:38] <Keybuk> Seveas: yeah, it's a common "WTF?!" why you can't sftp-get things read-only if you have a LP account
[01:39] <mjg59> Ok, that seems /much/ better
[01:48] <Seveas> mjg59, did you see usplash-test.sh ?
[01:49] <Seveas> I use that a lot to test shininess
[01:49] <mjg59> Nope
[01:49] <mjg59> I'll play now
[01:50] <mjg59> So, the good news is, usplash appears to add no time to the boot
[01:50] <Seveas> nice
[01:50] <Seveas> btw: multi-theme themes is almost working (code is there and simple enough that it should just work, will add a few extra resolutions to the example theme
[01:54] <Seveas> hmm, I'm getting syntax errors in code of which I'm fairly sure that I didn't change since the last push
[01:55] <Seveas> so either bzr is being weird or you didn't tell me about a few stupid bugs you found
[01:55] <mjg59> Oh, yeah
[01:55] <mjg59> I wasn't sure if they were merge bugs or not
[01:56] <Seveas> missing (..) and extraneous }
[01:56] <mjg59> Yup
[01:56] <Seveas> so not bzr bugs
[01:56] <Seveas> I thiught i did dpkg-buildpackage before pushing though
[01:56] <mjg59> Ok
[01:58] <mjg59> Seveas: The other thing that would be nice...
[01:59] <mjg59> Seveas: Have some "squashed" artwork that we can use on widescreen machines
[01:59] <mjg59> So it'll be stretched out to roughly the correct aspect
[02:00] <Seveas> currently I'm doing the using the crude method of selecting the theme that covers the maximum area -- it's up to the artists to create themes for various resolutions 
[02:00] <Seveas> and no stretching happens at all
[02:00] <mjg59> Well, the stretching is implicit
[02:01] <mjg59> 1024x768 on a 16:9 screen will be stretched
[02:01] <Seveas> no
[02:01] <mjg59> No?
[02:01] <Seveas> if -x and -y are present on the command line, those x and y are used as resolution
[02:01] <Seveas> the theme is centered in that resolution
[02:01] <Seveas> assuming -x and -y are correct, no stretching should occur
[02:01] <mjg59> No, I mean if you try to display a 4:3 screenmode on a 16:9 display, it'll be stretched out by the hardware
[02:02] <mjg59> So your image needs to be squashed to take that into account
[02:02] <jdub> worst is 640x400 vs. 640x480
[02:02] <Seveas> ah right, only those screenmodes are supported
[02:02] <mjg59> Seveas: Right, we have no widescreen video modes
[02:02] <Seveas> hmm
[02:02] <Seveas> would it be hard or not a good idea to add them?
[02:02] <mjg59> We can't add them
[02:02] <mjg59> VESA doesn't generally support it
[02:03] <Seveas> ah
[02:03] <Seveas> bummer
[02:03] <Seveas> but -x and -y will still be 16:9 right?
[02:03] <mjg59> Right
[02:04] <Seveas> ok, some thinking needed
[02:05] <Seveas> I really don't want to create 3 linked lists (a 4:3 one, a 16:9 and a 14:9(is taht really used?) one), so i'd suggest adding something in the theme struct to indicate whether it's squashed or not
[02:06] <mjg59> Right
[02:06] <Seveas> and the selection function should take that into account
[02:06] <mjg59> Yeah
[02:06] <Seveas> fair enough
[02:06] <Seveas> let's just support 4:3 and 16:9 now and select the ration that's closest to -x:-y
[02:06] <mjg59> Sure
[02:07] <Chipzz> btw
[02:08] <Chipzz> what happened to the graphical grub?
[02:08] <mjg59> Theme incompatible with usplash_bogl
[02:08] <mjg59> Hm
[02:08] <Chipzz> I still have it on a few boxes here, but that's because I didn't run grub-install yet
[02:08] <Seveas> mjg59, heh
[02:08] <Chipzz> iirc it also had issues with widescreen?
[02:08] <Seveas> usplash_bogl hardcodes "NONONO NO HIRES THEMES YET!" as safety measure
[02:09] <Seveas> line 32, usplash_bogl.c
[02:09] <mjg59> Ok, so that's a good start
[02:09] <mjg59> Colours are utterly wrong, but still
[02:11] <Seveas> are you ripping out my changelog entries after merging or am I misunderstanding what bzr is trying to do when I merge from your branch?
[02:11] <mjg59> I'm leaving them there
[02:11] <Seveas> hmm
[02:11] <mjg59> Then adding a new version above
[02:12] <Seveas> odd, in the changelog.OTHER there is no 0.4-16dennis 
[02:12] <Seveas> maybe I merged before it was realy there
[02:12] <mjg59> usplash (0.4-16dennis) edgy; urgency=low
[02:12] <mjg59> Is what's in the tree
[02:13] <Seveas> then something is confusing bzr 
[02:13] <Keybuk> dennit? :p
[02:13] <Seveas> that's a dch -i without paying attention ;)
[02:13] <mjg59> Ha
[02:14] <Seveas> hmm, I thin I merged too soon
[02:14] <Seveas> reverting
[02:15] <Seveas> yes, this merge is better
[02:20] <mjg59> Wow.
[02:20] <mjg59> That's an, uh, interesting effect.
[02:20] <sladen> mm?
[02:21] <Keybuk> hmm
[02:22] <Keybuk> ya know, this "make shutdown work with upstart and sysvinit" is actually damned hard
[02:22] <mjg59> Thinking about it, that's hardly unexpected
[02:22] <mjg59> I'm writing a one byte value without actually turning it into a useful colour
[02:23] <mjg59> Wait, no
[02:25] <mjg59> Ah, right, i see
[02:28] <Seveas> hmm, 14:9 is actually *exactly* between 16:9 and. Which one should be used for that resolution?
[02:28] <Seveas> s/resolution/ratio/
[02:28] <sladen> Keybuk: don't you just set each task goal to be 'stopped'.
[02:28] <sladen> Keybuk: and the task that runs 'rc' calls the whole contents of rc6.d ?
[02:30] <Keybuk> errr?
[02:30] <Keybuk> sladen: explain?
[02:31] <sladen> Keybuk: the sysv-compat is just a job that runs 'rc', yes?
[02:31] <Keybuk> no
[02:31] <sladen> Keybuk: (entirely possible the answer is not "yes", in which case you'll need to explain
[02:31] <Keybuk> it's ~10 jobs that arrange for rc to be run
[02:32] <sladen> Keybuk: I suspect you have a better handle on it too...
[02:32] <Keybuk> oh, you're talking about shutdown?
[02:33] <Keybuk> shutdown in upstart is easy, it sends out a "shutdown" event, waits for the system to go idle, then sends out whatever additional event you asked for
[02:33] <Keybuk> usually you do "shutdown" ... "reboot"
[02:33] <Keybuk> or "shutdown" ... "power-off"
[02:33] <Keybuk> etc.
[02:33] <Keybuk> so yes, it runs rc6, etc.
[02:34] <Keybuk> but that's not the problem I have here
[02:34] <sladen> Keybuk: so, 'this "make shutdown work with upstart and sysvinit" is actually damned hard' ?
[02:34] <Keybuk> the problem is what happens when you were running dapper, upgraded to edgy
[02:34] <Keybuk> you're running sysvinit
[02:34] <Keybuk> but have upstart installed
[02:34] <Keybuk> how do you tell sysvinit to shutdown? :p
[02:34] <sladen> Keybuk: sync, umount, reboot -f
[02:34] <Keybuk> lol
[02:34] <infinity> There are people who may be unimpressed with that solution.
[02:35] <infinity> And I'm using the word "solution" very liberally here. :)
[02:36] <Seveas> Keybuk, keep some parts of sysvinit around in /tmp and use them? + Be windows lik and really force people to reboot
[02:36] <sladen> Keybuk: my suspicion is that to provide compatibility with /dev/initctl you're going to have to copy a certian chunk of sysvinit into upstart
[02:36] <infinity> And, for the record, even "reboot -f" didn't do a damned thing after I upgraded to upstart.
[02:36] <Keybuk> infinity: it should do, that just does sync() reboot() :p
[02:36] <infinity> Keybuk: It just sat there like a dumb shit.  I had to cut the power.
[02:37] <Keybuk> how odd
[02:37] <sladen> wonder whether '-f' causing sync is actually out of spec
[02:37] <infinity> I thought so too.
[02:38] <sladen> sync() can potentially hang.  In which case, the purpose of trying to purpose a 'reboot -f' has been defeated (yes, I've had machines in that state)
[02:38] <sladen> s/purpose a/perform a/
[02:42] <Seveas> what's a common 16:9 resolution?
[02:45] <infinity> 1280x720
[02:45] <Seveas> mjg59, I'll merge from you tomorrow and then push multi-res-themes support
[02:45] <Seveas> infinity, thanks
[02:47] <Keybuk> sladen: reboot -n -f
[02:47] <Keybuk> sladen: aye, the general solution appears to be either
[02:47] <Keybuk> a) stash the old /sbin/init somewhere on install and use it to shut down the running one
[02:47] <Keybuk> b) copy enough from the old /sbin/init to send the initctl shutdown message ourselves
[02:48] <infinity> Keybuk: http://librarian.launchpad.net/4160730/buildlog_ubuntu-edgy-sparc.linux-restricted-modules-2.6.17_2.6.17.5-1_CHROOTWAIT.txt.gz
[02:48] <Keybuk> infinity: *sigh*  MVO!
[02:48] <infinity> Keybuk: Have we really thought this through?
[02:48] <Keybuk> it's not essential
[02:49] <infinity> Keybuk: It is in the installed system.
[02:49] <Keybuk> right, but it isn't in the archive
[02:49] <infinity> Anyhow, I'll update all the chroots manually, but this needs ot be addressed (obviously) for dapper->edgy
[02:49] <Keybuk> right
[02:50] <Keybuk> Colin thought making upstart essential would "fix" it
[02:50] <Keybuk> apparently not
[02:50] <infinity> Just made it worse, in my case. :)
[02:50] <infinity> Since nothing was forcing the upgrade in the chroots until that happened.
[02:50] <Keybuk> yeah
[02:55] <infinity> This misfeature is precisely the reason why we never make libraries Essential:yes
[02:56] <infinity> I suspect no one had really thought that we'd be replacing init anytime soon.
[02:56] <Keybuk> right
[02:57] <Keybuk> it's why I didn't make upstart essential originally, as I remembered it being difficult to get them off again
[02:58] <infinity> If we were willing to deal with not having an alternate init anymore, the easy way out would be to make sysvinit a transitional package.
[03:04] <LaserJock> hmm, I seem to have a problem with postrm vs. prerm here
[03:04] <infinity> What would that problem be?
[03:04] <LaserJock> I want to do some stuff postrm, but it need to check a file that is being removed in the postrm
[03:04] <Keybuk> infinity: the thought is increasingly occuring to me <g>
[03:05] <Keybuk> I have one standing by, in fact
[03:05] <Amaranth> Do it. :)
[03:05] <LaserJock> but if I put it in prerm then it doesn't get run at --purge right?
[03:05] <infinity> Only postrm remains when a package is removed but not purged, yes.
[03:06] <infinity> Of course, the only other things left should be conffiles.
[03:06] <infinity> So you can't rely on any non-conffile during the purge phase.
[03:06] <infinity> http://www.debian.org/doc/debian-policy/ch-maintainerscripts.html#s-removedetails
[03:06] <infinity> LaserJock: Policy is your friend.  It details how allof this works.
[03:07] <LaserJock> infinity: ah, I was looking in 6.5 ;-)
[03:07] <infinity> Well, all of section 6 is a good read. :)
[03:07] <infinity> s/section/chapter/
[03:07] <LaserJock> ok, well then I don't really need to worry about --purge then as I don't have any conf files
[03:26] <infinity> Has anyone else been noticing weird keyboard-grabbing issues in edgy over the last week or two?
[03:26] <infinity> Like, you'll focus a window, but it won't take keyboard input.  Switching VCs or even workspaces will fix it.
[03:26] <Keybuk> yeah
[03:26] <infinity> Oh good, I'm not alone.
[03:27] <infinity> I thought I was going insane.
[03:27] <Keybuk> usually x-chat vs gaim for me
[03:27] <Keybuk> switch between the two and one won't take keyboard until you switch back and forth a bit
[03:27] <infinity> gnome-terminal and GAIM being my two big keyboard-not-grabbers.
[03:27] <Burgundavia> infinity: yep, I have noticed that as well
[03:28] <infinity> Keybuk: A quick Ctrl-Alt-Right/Ctrl-Alt-Left always clears it up, BTW, if you don't want to much with Alt-Tabbing for a few minutes.
[03:28] <infinity> s/much/muck/
[03:29] <Keybuk> The "tags" portlet is my BEST FRIEND
[03:30] <infinity> Smells GTKish... Or even Xish.
[03:30] <Keybuk> gtk :p
[03:30] <Keybuk> or X, yes
[03:30] <infinity> If KDE people aren't whining, I'll assume GTK.
[03:30] <Keybuk> the KDE user is asleep
[03:30] <Burgundavia> the have users? ;)
[03:30] <Burgundavia> *hey
[03:31] <infinity> Hobbsee should be awake.  Slacker.
[03:31] <infinity> It's 11:30am, FFS.  Where's my token Australian Kubuntu user?
[03:31] <Riddell> ahem
[03:31] <infinity> Oh, hey.  I have a Scottish one.
[03:31] <infinity> Riddell: Noticed any keyboard grabbing (or lack of grabbing) oddities recently?
[03:32] <Riddell> infinity: nope
[03:32] <infinity> Alright, I'll blame GTK then.  Thanks.
[03:33] <Riddell> infinity, Keybuk: any thoughts on having KDE 4 packages in edgy, the files are installed to /usr/lib/kde4/{bin,lib,share,etc} as the only way to avoid them clashing with kde 3
[03:33] <ajmitch> sweet, workaround for f-spot crasher works
[03:33] <infinity> Riddell: In universe, co-installable, you mean?
[03:33] <Riddell> infinity: yes
[03:34] <infinity> Riddell: I'd probably be okay with that.  It's big enough that I'd recommend pinging mdz when he's done catching up on email, though.
[03:34] <Riddell> maybe at the distro meeting tomorrow then, time for sleep now
[03:35] <mdz> I'm able to poke my head above water from time to time at this point. what's up?
[03:35] <infinity> mdz: 4 lines up.
[03:35] <infinity> Or 6.  Or something.  Counting is hard.
[03:35] <Riddell> mdz: KDE 4 in edgy, installed to /usr/lib/kde4/
[03:36] <mdz> is the question about paths or about having the packages in edgy?
[03:36] <mdz> the latter, no problem
[03:36] <infinity> The latter.
[03:36] <Riddell> mdz: mostly if having it in that path will be allowed in
[03:36] <Riddell> since it's not really debian-policy compliant
[03:37] <infinity> If it's in universe as a set of "experimental packages for testing", I say policy-be-damned, to a certain extent.
[03:37] <Burgundavia> ajmitch: are there release notes for fspot 2.0?
[03:37] <infinity> gcc-snapshot isn't all that policy-compliant either. :)
[03:37] <ajmitch> Burgundavia: not really
[03:37] <Riddell> infinity: that's just the answer I'm looking for :)
[03:37] <Burgundavia> ajmitch: can you tell me in one sentence about what makes it 2.0?
[03:37] <Keybuk> why not /opt/kde4?
[03:37] <ajmitch> Burgundavia: 0.2.0, not 2.0
[03:38] <Riddell> Keybuk: too suse-ish?
[03:38] <infinity> Keybuk: Packaging systems messing in /opt is rather bad form.
[03:38] <Burgundavia> ajmitch: right
[03:38] <Keybuk> yeah, true
[03:38] <Keybuk> it's a bit /usr/local
[03:38] <Keybuk> unless you're Solaris of course
[03:38] <ajmitch> Burgundavia: lots of fixes, 0.1.x was branched & left for SLED
[03:38] <Keybuk> in which case /opt is the package manager's domain
[03:38] <infinity> Yes, well.  Thankfully, we're not Solaris.
[03:38] <ajmitch> when all the activity was on 0.2.0, which lewing recommended we have in edgy
[03:39] <infinity> If we were, I'd see a long walk and a short cliff in my future.
[03:39] <ajmitch> Burgundavia: though we're currently bitten by bug 59166 :)
[03:39] <Ubugtu> Malone bug 59166 in f-spot "Mono segfaults in dbus_pending_call_get_completed()" [Unknown,Unknown]  http://launchpad.net/bugs/59166
[03:39] <Burgundavia> ajmitch: right, this is for UWN
[03:40] <mdz> Keybuk: FHS forbids us from installing it in /opt
[03:40] <ajmitch> Burgundavia: I figured it would be
[03:40] <mdz> I don't know that there's an FHS-blessed place to include sub-hierarchies a la /opt though
[03:40] <mdz> I think their answer is "don't do that"
[03:40] <Keybuk> infinity: hmm, making upstart not Essential means that APT just refuses to update ubuntu-minimal
[03:40] <mdz> Riddell: what breaks if it just conflicts with kde3?  they shouldn't need to be parallel installable 
[03:41] <infinity> Keybuk: Special.
[03:41] <Burgundavia> Keybuk: system-services and startup-tasks are provided by upstart?
[03:41] <Keybuk> Burgundavia: "provided by" ?
[03:41] <Riddell> mdz: kde4 isn't stable enough to use as an actualy desktop yet, most developers will want the kde 4 libs for whatever app they're working on but still be running kde 3
[03:41] <infinity> Keybuk: I suspect this conversation needs an mvo and a bit of a thinktank.
[03:41] <Burgundavia> Keybuk: you just added those packages to ubuntu-desktop and I wondered where they came from
[03:42] <Keybuk> Burgundavia: no, I added them to ubuntu-minimal
[03:42] <Burgundavia> Keybuk: yes, I was about to point out my error
[03:42] <Keybuk> :)
[03:42] <Keybuk> they will be the upstart equivalent of "initscripts"
[03:42] <Burgundavia> and they don't exist yet?
[03:42] <Keybuk> they exist
[03:43] <Burgundavia> oh, right, that was my bad reading
[03:54] <Burgundavia> is it must me, or has performance when apt is loading the system quite bad?
[03:55] <infinity> Burgundavia: That sentence is lacking a few crucial words, I think.
[03:56] <jelmer> Keybuk: thanks for the quick reply. upstart seems like an interesting project; keep up the good work :-)
[03:56] <Burgundavia> infinity: hmm, ok: When apt is busy grinding away updating packages, responsiveness decreases significantly
[03:57] <infinity> Burgundavia: s/apt/dpkg/, surely.  And yes, it can be a bit evil. Not sure it's really gotten any worse, though.
[03:58] <Keybuk> infinity: at least if everything's not Essential, the second time you run dist-upgrade, everything works
[03:58] <Burgundavia> infinity: the break between apt and dpkg always confuses me, but yes
[03:58] <Keybuk> the first time it holds ubuntu-minimal, but upgrades sysvinit (to the non-Essential one)
[03:58] <Keybuk> then the second time it can upgrade ubuntu-minimal and install upstart
[03:58] <infinity> Burgundavia: apt is just an acquisition tool.  It doesn't install anything.
[03:59] <Burgundavia> infinity: ah, ok
[03:59] <infinity> Keybuk: Well, I guess that could be worked around in update-manager by just doing two passes, so perhaps that's the best solution.
[03:59] <infinity> Keybuk: And people doing it by hand have been long trained to keep trying to upgrade untilit "sticks" anyway. :)
[03:59] <Keybuk> indeed
[04:00] <Keybuk> #define INIT_MAGIC 0x03091969
[04:00] <Keybuk> or maybe he's American, and can't get dates right, and it was in March
[04:01] <jdub> Keybuk: heh
[04:04] <jdub> infinity: put in a proposal for lca yet?
[04:11] <theCore> does GNOME removed the Shift-Ctrl-[Unicode]  shortcut?
[04:13] <Keybuk> with any luck, this might work
[04:13] <Keybuk> /lib/libc.so.6: version `GLIBC_2.4' not found
[04:13] <Keybuk> grr
[04:17] <Keybuk> \o/
[04:36] <infinity> jdub: Give me something interesting to talk about, and I'll write something. :)
[04:36] <infinity> jdub: I'm terrible at coming up with topics.
[04:37] <infinity> Okay, migrating diversions from one package to another is a *pain*
[04:37] <infinity> Doubly-so, when the diverted file is dpkg-deb.  *cough*
[04:40] <mjg59> infinity: Maintaining a buildd network
[04:40] <mjg59> How to avoid implementing specs
[04:41] <infinity> mjg59: Yeah, I thought so too, then they cleverly assigned me specs that vaguely relate to buildds.
[04:46] <infinity> mjg59: When mangling maintainer addresses, do you have any you'd like whitelisted?  (ie: LaMont requested that I not mangle lamont@debian.org to another contact, as he's happy to support his Ubuntu packages via that address)
[04:47] <infinity> mjg59: Reference here: https://wiki.ubuntu.com/DebianMaintainerField
[04:47] <infinity> mjg59: About halfway down the page, the mangling strategy (minus whitelists) is described.
[04:47] <mjg59> Erm
[04:48] <mjg59> I think all my debian maintainer entries are likely to go away soon
[04:50] <infinity> Right.  Just looking for people to whitelist off the bat.
[04:50] <infinity> I already have lamont, so one example in the config file is good enough.
[05:15] <robertj> has anyone looked at this patch for adding an Empty Trash view to trash://? http://pages.cpsc.ucalgary.ca/~weckerl/nautilus_patch.html
[05:19] <robertj> sorry, Emptry Trash widget
[06:50] <cge> Is there some reason besides tradition that ed doesn't display any error messages by default? Would changing the behaviour break something?
[06:51] <Burgundavia> cge: does anybody still use ed?
[06:51] <StevenK> My boss does
[06:53] <mneptok> dum dee dee
[06:53] <cge> Burgundavia: Ed is THE editor!
[06:54] <mneptok> that's bizarre. i popped in here to start a main inclusion discussion for another editor.
[06:54] <cge> More seriously, I've used it a few times with scripts, like sed.
[06:54] <Burgundavia> mneptok: which one?
[06:54] <mneptok> ne
[06:55] <Burgundavia> and what do you want done with it?
[06:55] <mneptok> http://ne.dsi.unimi.it/
[06:55] <cge> mneptok: ne is already in universe
[06:55] <mneptok> Burgundavia: i thought "main inclusion" would be a giveaway there
[06:55] <mneptok> :)
[06:56] <Burgundavia> do you mean main inclusion or installation by deafult?
[06:56] <mneptok> inclusion
[06:56] <Burgundavia> right
[06:56] <mneptok> installation would be nice, but space is tight.
[06:56] <Burgundavia> what does it have over nano?
[06:56] <Burgundavia> which is already both
[06:56] <mneptok> ease of use for shell newbies
[06:56] <cge> mneptok: Installation will never happen. There are even people who want to remove vim!
[06:57] <Burgundavia> used nano? (not saying it is a bad idea, just want to see why it is good"
[06:57] <Burgundavia> cge: vim is already gone, vim-tiny is in
[06:57] <cge> Burgundavia: Yes, but there were those in that discussion who didn't want vim in at all.
[06:57] <mneptok> Burgundavia: i'm a nano user. i want this in main so we can officially support it, and have it something we can recommend to more novice users.
[06:57] <Burgundavia> cge: like myself
[06:58] <Burgundavia> mneptok: to justify it, what does it have over nano?
[06:58] <Burgundavia> better features? easier to use? smaller install?
[06:58] <mneptok> 00:56 < mneptok> ease of use for shell newbies
[06:58] <Burgundavia> in waht way?
[06:58] <mneptok> i.e. an menu-driven command system that GUI users find more approachable.
[06:58] <mneptok> s/an/a/
[06:59] <cge> Burgundavia: ne is about twice as big when installed as nano
[06:59] <Burgundavia> hmm
[06:59] <cge> Burgundavia: around 500k
[06:59] <mneptok> 260 here
[06:59] <mneptok> -rwxr-xr-x 1 root root 267624 2005-10-25 09:43 /usr/bin/ne
[06:59] <mneptok> ok, 270
[07:00] <cge> ed is only 40 kilobytes, and I think that is a rather large implementation.
[07:01] <mneptok> cge: i don't want to be put in the position of telling as new Linux user coming from a more GUI environment that they have to edit hidden or root-owned files with ed. they'll melt down.
[07:01] <Burgundavia> right, that is why the default editor is nano
[07:02] <mneptok> i have already dealt with people that find nano confusing, and want to try editing /etc stuff with gedit, which leaves foo.txt~ temp files all over /etc
[07:03] <Burgundavia> according to packages.ubuntu.com, nano is 100k larger than ne
[07:03] <mneptok> i'd like to be able to tell them we support a more GUI-like shell editor.
[07:03] <Burgundavia> so here is what you need to do: create a spec talking about the reasons why we shoudl replace nano with ne
[07:03] <mneptok> and if it's not in Main, we tend not to support it.
[07:03] <Burgundavia> then get some people like myself to look voer it
[07:03] <mneptok> (at least for the time being)
[07:05] <mneptok> yeah, i'm going to be doing an inclusion report. but first stop is here.
[07:05] <mneptok> https://wiki.ubuntu.com/UbuntuMainInclusionQueue?highlight=%28inclusion%29%7C%28main%29
[07:05] <Burgundavia> I would do the spec first
[07:05] <mneptok> figured i'd idle here until the lazy euro co-workers unidle :)
[07:05] <Burgundavia> main inclusion is unlikely without a good rationale
[07:18] <cge> mneptok: Have you considered that gedit's behaviour could be changed to not leave backup files around like that? Editing /etc/ things with gedit should be supported.
[07:43] <infinity> mjg59: Still around?
[07:44] <Seveas> isn't he in UTC+1?
[07:44] <infinity> Yeah, but he was here not that long ago. :)
[07:44] <infinity> Oh, perhaps longer thanI remember.
[07:45] <Burgundavia> infinity: mjg59 last spoke about 6 hours ago, just as I left work
[07:45] <Seveas> no, he spoke arond 2:something here too
[07:45] <Seveas> that's around 5h
[07:46] <Seveas> and exactly the amount of sleep I've had 
[07:47] <Burgundavia> hmm, have to be at work tomorrow at 6am
[07:48] <Burgundavia> hmm, it appears we have lost Stephen Vaughan to SLED
[07:49] <jdub> hopefully we never really had him :)
[07:49] <mneptok> oy jdub
[07:50] <Burgundavia> jdub: he does appear to write a lot
[07:53] <jdub> Burgundavia: i suppose, being a journalist, he is somewhat compelled to.
[07:54] <Burgundavia> jdub: there is a certain respectability that is gained from pure volume
[07:54] <jdub> Burgundavia: you appear to be talking from the incorrect orifice.
[07:54] <mneptok> although *why* anybody listens to that loon any more escapes me
[07:55] <desrt> jdub; with popularity does come a certain degree of responsibility...
[07:55] <infinity> Seveas: Actually, since you've been recently involved with usplash hacking, you may be able to answer this one for me.
[07:55] <desrt> oh.   he said respectability
[07:55] <desrt> no.  i don't agree with that at all :p
[07:56] <infinity> Seveas: Is usplash-dev going to end up being a build-dep of anything in main in the near future, or should I send it to universe?
[07:56] <jdub> sjvn hasn't really earned much of my respect
[07:56] <Seveas> infinity, main -- I split it off of the main binary so artwork packages don't need to depend on usplash
[07:56] <desrt> Burgundavia; did you do any more work on the mentorship thing?
[07:56] <jdub> at one point, i pointed out that his description of portland was incorrect
[07:56] <Seveas> it'll have to be a builddep of artwork packages
[07:57] <Burgundavia> desrt: no, been insanely busy with work and a few personal projects
[07:57] <Burgundavia> desrt: can you fire me an email to remind me?
[07:57] <desrt> Burgundavia; it's cool.  i'll keep politely reminding you every week or so :)
[07:57] <desrt> ok.  i can do that too.
[07:57] <jdub> and he decided to disagree, assert even more outlandish ideas, and ignore the fact that i was a participant in the initial portland specification meeting
[07:57] <infinity> Seveas: Okay, great.
[07:57] <infinity> Seveas: Thanks.
[07:58] <mneptok> jdub: cut him some slack. omniscience is a heavy burden.
[07:59] <desrt> Burgundavia; ok sent.  i'm gonna go to bed now.  ciao.
[07:59] <Seveas> mneptok, please don't do that when I am just drinking something
[07:59] <Seveas> my keyboard doesn't like that
[08:00] <jdub> willful ignorance is a heavy burden. :-)
[08:01] <mneptok> Seveas: coffee | nose > keyboard
[08:01] <infinity> I'd suspect that wouldn't do much harm without a 2>&1
[08:02] <infinity> Surely, snorting coffee is a standard error.
[08:02] <lifeless> IO overflow I think
[08:04] <infinity> That nick gives me a headache...
[08:06] <mneptok> infinity: try eating an aptly-named user
[08:53] <pitti> Good morning
[08:54] <tseng> hello pitti 
[08:56] <jdub> ber, laptop still 'in production', before 'delivery prep'
[08:56] <jdub> hello tseng and pitti
[08:56] <Lathiat> jdub: which one did you get?
[08:57] <Lathiat> dell or the ibm?
[08:57] <jdub> Dell D420
[08:57] <jdub> + mediabase with DVI!
[08:57] <Lathiat> ah nice
[08:57] <Lathiat> for the 24"? :)
[08:57] <jdub> saves getting an apple :)
[08:57] <jdub> yeah
[08:57] <pitti> tseng: hi
[08:58] <pitti> tseng: yes, wv is fine with me, I was just concerned about code duplication
[08:58] <tseng> pitti: ok, i put it back on the page, no need to look at it right now
[08:58] <tseng> pitti: im on a business trip, anyway
[08:59] <tseng> on that note, 3am, done working for the day
[08:59] <tseng> good night
[08:59] <jdub> night tseng!
[08:59] <pitti> tseng: sleep well!
[09:00] <pitti> jdub: pants off!
[09:00] <mneptok> jdub: dunno about Oz, but in the US IBM shareholders get some nice shareholder discounts on Lenovo stuff. worth the price of one share to pay US$1299 for a laptop that is US$1899 retail. :)
[09:00] <mneptok> (kinda late with the tip, but oh well)
[09:01] <jdub> mneptok: a discount for being punched in the face?
[09:03] <mneptok> uhhhh .... i'll just pretend i understood that and say "i guess so"
[09:04] <jdub> mneptok: i am not a fan of the thinkpad keyboards (esc/f1, fn/ctrl)
[09:05] <mneptok> ah, mine has a seperate <esc> and <fn> keys. the meta for the <fn> keys are all things like battery display, brightness, power savings, etc.
[09:05] <jdub> i always hit f1 instead of esc on pia's
[09:05] <jdub> and fn is toooootally in the wrong place
[09:06] <mneptok> you could always remap the keys with software or epoxy. :)
[09:07] <jdub> fn/ctrl isn't doable
[09:07] <HiddenWolf> mneptok: and put stickers on the keys with the correct keycodes, yeah right. :P
[09:07] <jdub> due to the nature of the fn key
[09:07] <mneptok> hrmf
[09:12] <mneptok> here's an ugly choice. sl-modem is in Multiverse. we can either move it to restricted and actually support it (*shudder*) or have no official support for any Winmodems.
[09:12] <mneptok> ugh
[09:14] <HiddenWolf> There is actually some software that can make those ugly hardware hacks work?
[09:15] <mneptok> sl uses ALSA to communicate with the Winmodem (which basically a sound card)
[09:16] <mneptok> apparently you just have to have a working ALSA and a Winmodem that actually exposes itself as a modem.
[09:16] <mneptok> but the whole thing smells of dead hookers and vomit.
[09:16] <lifeless> mneptok: we have a support department to wear such pain, no ? :)
[09:16] <mneptok> i love you, too.
[09:16] <lifeless> of course you do
[09:17] <mneptok> *muah*
[09:17] <jdub> lifeless: to wear the pain of dead hookers and vomit?
[09:19] <dholbach> good morning
[09:22] <cbx33> hi dholbach 
[09:23] <dholbach> hi cbx33
[09:28] <infinity> mneptok: slmodem has probably wanted some main love for ages now, but no one in core-dev uses a modem, so it keeps getting ignored.
[09:39] <jsgotangco> mvo!
[09:41] <mvo> jsgotangco: good morning!
[09:43] <mneptok> infinity: are you using some Jedi mind trick to try to make *me* be the poor stooge that's submits Main inclusion for sl-modem?
[09:43] <mneptok> huh huh huh huh. mettu steeny barana muddah Jedi mind trick d'vulus andaska.
[09:50] <Kamion> ok, that's console-setup more or less working in d-i, anyway
[10:08] <tepsipakki> why are cdroms mounted as noexec by default?
[10:08] <tepsipakki> annoying
[10:13] <mempf-edgy> any news on when the next kernel update will be?
[10:14] <Seveas> mjg59, usplash -17 is messing up my consoles: nothing on the screen, getty stuck in read() according to strace
[10:14] <ogra> Kamion, 
[10:14] <ogra> Reading package lists...
[10:14] <ogra> W: Couldn't stat source package list file: edgy/main Packages (/srv/cdimage.no-name-yet.com/scratch/edubuntu/daily/apt/edgy-powerpc/apt-state/list
[10:14] <ogra> s/_srv_cdimage.no-name-yet.com_ftp_dists_edgy_main_binary-powerpc_Packages) - stat (2 No such file or directory)
[10:14] <ogra> W: Couldn't stat source package list file: edgy/restricted Packages (/srv/cdimage.no-name-yet.com/scratch/edubuntu/daily/apt/edgy-powerpc/apt-stat
[10:14] <ogra> e/lists/_srv_cdimage.no-name-yet.com_ftp_dists_edgy_restricted_binary-powerpc_Packages) - stat (2 No such file or directory)
[10:14] <ogra> W: You may want to run apt-get update to correct these problems
[10:14] <ogra> E: Some index files failed to download, they have been ignored, or old ones used instead.
[10:15] <ogra> make: *** [apt-update]  Error 100
[10:15] <ogra> seems lithium isnt well
[10:16] <Kamion> Failed to fetch file:/srv/cdimage.no-name-yet.com/ftp/dists/edgy/main/binary-pow
[10:16] <Kamion> erpc/Packages.gz  MD5Sum mismatch
[10:16] <Kamion> Failed to fetch file:/srv/cdimage.no-name-yet.com/ftp/dists/edgy/restricted/bina
[10:16] <Kamion> ry-powerpc/Packages.gz  MD5Sum mismatch
[10:16] <Kamion> try again later
[10:16] <Kamion> this happens sometimes when you get unlucky with the archive
[10:16] <ogra> ah, ok
[10:16] <Kamion> it's a soyuz bug of course - that shouldn't be possible
[10:16] <ogra> i can wait till it sorted itself :)
[10:17] <Kamion> if it hasn't sorted itself by the next publisher run, let me know
[10:17] <ogra> yup
[10:28] <Kamion> infinity: how're live-cd-stacked-filesystems/larger-livefs going?
[10:30] <sivang> morning
[10:52] <doko> ajmitch: there already are zope sync requests filed for zope
[10:54] <mvo> Kamion: did you got my germinate mail?
[10:55] <Kamion> mvo: yeah, haven't read the diff yet though
[10:56] <Kamion> I'm battering through sane-installer-keyboard
[10:56] <mvo> Kamion: ok, no problem
[10:56] <Kamion> infinity: could you give-back console-data on all architectures please? I can't see why it failed but my guess is that it's transient ...
[10:56] <Kamion> mvo: I'll try to get to it this morning thoug
[10:56] <Kamion> h
[10:56] <mvo> cool!
[10:57] <Kamion> infinity: oh, maybe not
[10:57] <Kamion> i386 was transient I think, but powerpc looks more substantial
[10:58] <infinity> Kamion: So, no give-backs?
[10:58] <Kamion> infinity: looks like I need to reupload anyway
[10:58] <Kamion> so no, spoke too soon
[10:58] <infinity> Kamion: Also, live-cd-whatever should be complete by the meeting, except for the whole "blocking on kernels on the buildd" thing.
[10:59] <Kamion> infinity: can that be expedited somehow?
[10:59] <infinity> Kamion: Indulging in a bit of evening wine right now, then back to work between now and the meeting to polish up.
[11:00] <Kamion> right, but there's also the cdimage/ubiquity end, and I'd really like to be able to test that
[11:02] <infinity> Well, royal is running a kernel with unionfs support, so when I'm done playing here, I could roll it out on powerpc.  I can't do amd64/i386 without elmo's team.
[11:02] <Kamion> I've made a note to ping sysadmin about that
[11:02] <Kamion> is it just unionfs, or squashfs too?
[11:03] <infinity> Should be just unionfs, though squashfs would be nice to be able to do local testing without dragging the images elsewhere.
[11:03] <Kamion> infinity: did you already file an RT request?
[11:04] <Kamion> since they're gonna ask me. :)
[11:04] <infinity> I think I did a month or so ago, and now I'm questioning that.  I know I spoke at length with elmo about it.
[11:04] <infinity> I can always file another. :)
[11:06] <Kamion> I've asked
[11:07] <zyga> mvo: hi, I hope you didn't waste your time on running that script
[11:07] <mvo> zyga: I did, why? something wrong?
[11:08] <mvo> zyga: it is still runing
[11:08] <zyga> yeah, one fix broke sometime later
[11:08] <zyga> some package in universe
[11:08] <zyga> anyway I'm running the updated script since 9 CET
[11:08] <mvo> zyga: ok, let me know when I can merge/re-run. not a big issue :)
[11:08] <zyga> otherwise everything is great, data format is the same and it just works :)
[11:09] <zyga> ok
[11:09] <mvo> cool :)
[11:31] <infinity> Why do I push so hard to get new kernels through the machinery and into the archives, and then spend 5 minutes fretting about whether or not I should reboot after installing them?
[11:31] <infinity> Question for the ages, I'm sure.
[11:32] <Fade> Mithrandir: did that xlib update get pushed out yesterday?
[11:33] <jdub> infinity: sadism.
[11:34] <mneptok> infinity: welcome to life inside the sausage factory. "yeah! new flavor! let's ship it!" :: later, in the men's room :: "Bob, we *did* clear up that botchelism thing, right?"
[11:35] <Spads> botulism is a harsh mistress
[11:35] <jdub> 'botulism'
[11:35] <mneptok> roit
[11:35] <Spads> jdub: but kernels have botchelism
[11:35] <mneptok> very late in my day. i'm running on 2 pistons.
[11:36] <jdub> Spads: infinity's do!
[11:36] <Spads> NO PISTONS HERE!!!
[11:36] <infinity> My kernels will not stand for this.
[11:36] <thom> Spads: haven't you been recalled by mazda? :P
[11:36] <jdub> Spads: enjoying the new gig?
[11:36] <Spads> jdub: aye
[11:37] <jdub> Spads: saw a great television interview with one of your colleagues -
[11:37] <jdub> http://youtube.com/watch?v=AxW6-_Qx1JA
[11:37] <jdub> absolutely priceless
[11:38] <StevenK> jdub: Bwahaha
[11:38] <infinity> jdub: I'm disturbed that you watch Rove.
[11:38] <Spads> http://www.youtube.com/get_video?video_id=AxW6-_Qx1JA&l=597&t=OEgsToPDskJTVZZ4VhY9GgyiVPAN2Tcf&nc=6724044
[11:39] <Spads> of course, I'm in the office
[11:39] <Spads> so I will not fall for this obvious troll
[11:39] <mneptok> Spads: "Wankel" ;)
[11:39] <jdub> Spads: he should see it
[11:40] <mneptok> jdub: so what *is* your new gig? been owrking too hard to keep on top of such news.
[11:40] <Spads> mneptok: Speak for yourself.
[11:40] <jdub> my new gig is hitting reload on the dell shipping page
[11:40] <jdub> so i can find out when my new laptop arrives
[11:41] <Spads> ha ha dell
[11:41] <mneptok> training for that must have taken a while.
[11:41] <Spads> mneptok: he kept failing his back but
[11:42] <Spads> button exams
[11:42] <Seveas> mjg59, new usplash revisions uploaded: multi-variant themes, a fix against non-working TIMEOUT and some doc/copyright fixes
[11:42] <Mithrandir> Fade: yes, it should.
[11:42] <jdub> the website kept failing those exams!
[11:42] <Spads> okay, when will ubuntu get the "disable middle click on touchpad that rests under the palm of your hand" gui control?
[11:42] <mneptok> Spads: when you get a new laptop
[11:42] <thom> touchpads are evil, kthx
[11:43] <Spads> yeah they are
[11:43] <Spads> I wouldn't mind if the clicktaps were disabled
[11:44] <Gman> jdub, you bought a dell laptop? nutball.
[11:44] <jdub> Spads: seems the only way to configure that stuff is via insecure shmem crack
[11:44] <jdub> Gman: a good one. :)
[11:44] <jdub> i had tough requirements
[11:44] <mneptok> dholbach: only because someone labelled your middle-click "SEND HUGZ0RZ!!!!"
[11:44] <infinity> jdub: So, it's an IBM with a Dell logo glued on it?
[11:44] <thom> jdub: oxymoron. or just moron. which ever :-)
[11:44] <Gman> jdub, one that will allow you to take on planes?
[11:44] <dholbach> jdub: what? you replaced your quebecistani laptop?
[11:44] <StevenK> infinity: Hah
[11:44] <Spads> I have a button for middle click
[11:44] <Gman> hey thom
[11:44] <jdub> ibm lost, and i totally didn't want to get an apple
[11:44] <dholbach> mneptok: :-)
[11:44] <thom> Gman: dude :-)
[11:45] <Spads> and I can chord either set of L/R buttons
[11:45] <Spads> I don't need the tapmiddle
[11:45] <ogra> ENOSEB :(
[11:45] <jdub> dholbach: hopefully, it will complete its duty when i visit the UK office later in the month.
[11:45] <Fade> I saw thre libx11 updates, but no xlib.
[11:45] <ogra> dholbach, do you have an idea how far seb is with the 2.16 gnome-session package  ?
[11:45] <Gman> jdub, nod, ibm is ovely expensive, apple is evil :)
[11:45] <Fade> s/thre/three
[11:45] <dholbach> ogra: no, I'm sorry - why?
[11:45] <jdub> dholbach: despite being a complete bastard of a laptop, it has been a trooper under demanding conditions.
[11:45] <Fade> xemacs is still dying horribly.
[11:46] <Fade> should I relink it?
 xemacs is still horrible.
[11:46] <sladen> Spads: as jdub notes, it needs a (Safe) X-extension writing for runtime Synaptics configuration---or leaving the user to enable SHMConfig on themselves
[11:46] <dholbach> jdub: demanding conditions like what? spilling water into the keyboard?
[11:46] <Spads> ah
[11:46] <ogra> dholbach, he told me to just upload a patch that disables fading on logout for ltsp clients, but i suspect he's working on the new package and dont want to come in his way 
[11:46] <jdub> dholbach: being my laptop during and after canonical is a demanding condition. :)
[11:46] <Seveas> dholbach, having beagle crapping ovr your FS during a presentation ;)
[11:46] <dholbach> ogra: just send him the patch by mail and he'll integrate it with his upload
[11:47] <dholbach> Seveas: right :)
[11:47] <dholbach> jdub: I see :-)
[11:47] <jdub> Seveas: that was pre-craptop :)
[11:47] <Seveas> dholbach, or having jdub stp on it while searching for pants
[11:47] <jdub> wasn't it?
[11:47] <jdub> hrm
[11:48] <Seveas> it was october last year
[11:48] <Seveas> during eurooscon
[11:48] <jdub> post ubz?
[11:48] <Seveas> (btw: I'll be at eurooscon this year)
[11:48] <jdub> yeah
[11:48] <Seveas> pre-ubz
[11:49] <jdub> oh
[11:49] <jdub> yeah
[11:49] <jdub> i'll be at eurooscon too
[11:49] <jdub> should be fun
[11:49] <thom> yeah, i'm thinking about euroscon
[11:50] <jdub> dude! come!
[11:50] <Seveas> sabdfl is coming too, accorfing to the agenda
[11:50] <jdub> he's doing closing
[11:51] <jdub> dholbach: interesting, gentoo ported all the rh admin tools :)
[11:56] <pygi> sivang, poke? :)
[11:56] <sivang> pygi: hi
[11:56] <pygi> sivang, asero 0.4.9 with libburn/libisofs backend is out, and it rocks :)
[11:57] <pygi> Brasero*
[11:57] <Seveas> Kamion, ping
[11:57] <dholbach> jdub: that's interesting
[11:57] <mneptok> sivang: Jane pinged me last week about DB/2 progress. gave her an update, and she should have renewed Canonical's agreement with Big Blue.
[11:57] <Kamion> Seveas: hi
[11:58] <Seveas> Kamion, usplash uses /etc/usplash.conf which is generated by the usplash postinst -- is it regenerated after a ubiquity install to match the installed system?
[12:00] <Kamion> Seveas: no - what about it is system-specific?
[12:01] <Seveas> X-detected resolution is grabbed from debconf and stored in /etc/usplash.conf
[12:01] <Kamion> file a ubiquity bug please, I'll make it do that
[12:01] <Seveas> ok, will do -- I'll attach the postinst
[12:02] <infinity> Seveas: Did you make the postinst less ugly, or is it still my hideous "It's 5am, and I don't care, I just want it to work" shell?
[12:02] <Seveas> infinity, I didn't prettify it, don't know about mjg59
[12:02] <sivang> mneptok: cool, great.
[12:02] <infinity> Seveas: Just checked, it's still scary. :)
[12:02] <Seveas> heh ;)
[12:02] <zyga> pitti:ping
[12:03] <zyga> pitti: some .desktop and .png files in langpacks have +x set, that doesn't make any sense
[12:03] <infinity> Easily discoverable because I'ma backtick whore, and people keep slapping me, telling me that I'm living in 1975, and I should use $()
[12:04] <ogra> infinity, we'll need to talk about https://features.launchpad.net/distros/ubuntu/+spec/ltsp-daily-image-tarballs soon
[12:04] <StevenK> infinity: You're living in 1975, use $() *slap*
[12:04] <infinity> StevenK: Thpt.
[12:04] <StevenK> infinity: Just try and nest your backticks.
[12:04] <Kamion> infinity: I suggested cut instead of sed
[12:05] <pitti> zyga: oh, hmm; thanks for the hint
[12:05] <mneptok> so, my biggest gripe about Jono thus far is ..... HEY HEY! hi!
[12:05] <infinity> Kamion: I think there was a specific reason I didn't use cut, and it wasn't cause I wanted to waste CPU/RAM.
[12:05] <jono> mneptok, hehe
[12:05] <infinity> Kamion: Damned if I can remember why now.
[12:06] <sladen> mneptok: we should have a voice-o-meter competition between jdub and jono to see who's the loudest :)
[12:06] <mneptok> aha! physically assaulting me! now you really *are* a Canonical employee!
[12:06] <infinity> ogra: Yes, we should.  You're certainly leaving it to the wire, though. :P
[12:06] <Seveas> sladen, please distribute ear protectioin generously before doing that
[12:06] <jono> sladen, I will kick his ass, and he knows it :P
[12:06] <infinity> ogra: It may end up having to happen shortly after FF, since I'm flat out right now on getting my own stuff done.
[12:06] <mneptok> sladen: in the end it doesn't matter. i have bigger boobs.
[12:07] <jono> mneptok, haha
[12:08] <ogra> infinity, i'm fine with that, the scripts are done and lie under http://people.ubuntu.com/~ogra/ltsp-tarballs/
[12:08] <ogra> (in the bzr archive)
[12:08] <ogra> so you can grab them any time yu ike 
[12:08] <ogra> *you like
[12:13] <pygi> pitti, poke, I wanna bug you ;)
[12:26] <pitti> pygi: I'm a bit busy ATM, but feel free to /msg me, I'll read scrollback
[12:26] <pygi> pitti, just poke me when you're free 
[12:27] <dholbach> Kamion: can you have a look at  http://daniel.holba.ch/temp/debhelper-pngcrush.patch ? (my attempt to let dh_compress make use of pngcrush) - if I could do it cleverer, I'd be happy to do so
[12:34] <zyga> mvo: better kill it, it loops on template.extract for some reason and eats all memory
[12:34] <zyga> s/template.extract/template.substitute/
[12:35] <mvo> zyga: hm, ok
[12:44] <mvo> zyga: killed
[12:45] <Kamion> dholbach: TBH I'd rather we did that by hand for now in the few packages that really benefit, and send that patch up to Joey to comment
[12:47] <dholbach> Kamion: Ok. I'll follow up on the post on ubuntu-devel@ saying so.
[12:47] <Kamion> dholbach: I'm not sure I like the extra dependency
[12:47] <dholbach> Kamion: I knew it was going to be controversial :/
[12:48] <Kamion> maybe it would be better for dh_compress to use pngcrush if it's there and require packages that want to use it to build-depend on it themselves? or make it a command-line option? I really don't know, and it seems like the sort of thing that would best be discussed in a Debian debhelper bug
[12:48] <infinity> pngcrush as a debhelper dependency does seem somewhat unacceptable to me as well.
[12:48] <infinity> I do like the idea of the "if it's there" option, though.
[12:49] <dholbach> I'm not sure that it makes sense for package to build-dep on pngcrush - it's the same amount of work to just run pngcrush on your stuff or prod upstream to do it and then happily live on
[12:49] <Kamion> only trouble with that is that it's a bit nondeterministic
[12:49] <infinity> Not nondeterminstic in a way that really matters.
[12:49] <Kamion> true
[12:50] <Kamion> still, I think it's worth discussing
[12:50] <Kamion> dholbach: oh also, pngcrush is in universe right now
[12:50] <dholbach> Kamion: I'm aware of that.
[12:50] <infinity> Hrm, pngcrush, libpng12-0, zlib1g... That's not SO bad.  I was expecting a longer chain.
[12:50] <dholbach> Kamion: I just thought it'd be better to get it into main as opposed to let debhelper depend on imagemagick. :-)
[12:51] <Kamion> dholbach: can we just crush the artwork manually for now?
[12:52] <dholbach> Kamion: It's not only about artwork
[12:52] <dholbach> Kamion: documentation, icons, etc etc
[12:52] <dholbach> but I can do it for artwork, yes
[12:52] <Kamion> sure, but a lot of those are small enough not to matter
[01:06] <pitti> pygi: ok, next part of interview just finished, go ahead
[01:08] <pygi> pitti, well, it was about creating a libburn, libisofs and cdrskin package
[01:09] <pygi> since sivang probably won't have time to create them, I would like to aks you if you could do it
[01:10] <pygi> ask*
[01:13] <pitti> pygi: I'm afraid I won't have time in the next days
[01:13] <pitti> pygi: you don't want to do it yourself?
[01:13] <pygi> pitti, busy with hacking on libburn, and I ain't MOTU so won't be able to maintain 
[01:15] <pygi> I could probably become MOTU If I had someone for key signing, but this way I'll never be MOTU :)
[01:15] <pitti> pygi: maybe you can remind me again next week; this week I have a lot of other stuff going on
[01:15] <pygi> pitti, no worries, all god
[01:15] <pygi> good*
[01:39] <mneptok> eeek! a cr3!
[01:47] <pitti> hey sabdfl, how's it going?
[01:47] <sabdfl> pitti: happily so far, though my own bed is looking more and more attractive! just got to tokyo
[01:48] <Fujitsu> Hi, sabdfl.
[01:48] <pitti> big in Japan :)
[01:48] <sivang> sabdfl: woo, tokyo :-)
[01:48] <sabdfl> hi Fujitsu, sivang
[01:48] <sivang> pitti++
[01:49] <cbx33> hey sabdfl 
[01:49] <sabdfl> good work on the sounds, cbx33
[01:49] <cbx33> thanks you're happy with them
[01:49] <cbx33> ?
[01:49] <cbx33> I just got SCP finished up too :D
[01:49] <cbx33> Pessulus is now integrated :D
[01:49] <sabdfl> cbx33: some concern about length, but nonetheless happy to have someone on it!
[01:50] <cbx33> sabdfl: well we're gonna put them on Knot3 - I'll just add on this AMD64 machine as soon as the sound finishes....the desktop appears :p
[01:51] <cbx33> so we can get feedback on them after that
[02:00] <sivang> ah damn, mistakengly uploaded bin pkg :-/
[02:02] <pitti> sivang: they will
[02:02] <sivang> pitti: good
[02:04] <pitti> hi till 
[02:04] <pitti> tillkamppeter_: maybe freenode requires you to use a nick for half an hour before you can register it?
[02:04] <cbx33> I'm just hoping SCP can make it into main
[02:04] <cbx33> but i doubt it will now....
[02:06] <sivang> pitti: do I need to reupload ?
[02:06] <pitti> sivang: yes
[02:09] <tillkamppeter_> pitti, I succeeded to register tillkamppeter now (seems really to be needed to be logged in for 30 min, they should write this in the FAQ) and I am trying to link tillkamppeter_ to it.
[02:09] <pitti> tillkamppeter_: great
[02:10] <tillkamppeter_>  /msg nickserv link tillkamppeter NISfra#3
[02:10] <pitti> tillkamppeter_: ouch
[02:12] <_ion> Hmm. /me put "splash" back to the kernel command line. Usplash didn't break the resolution of the virtual consoles anymore, but it didn't show any splash either. How should i debug this?
[02:12] <tillkamppeter_> pitti, so this wa the test and now I will start over with everything new, as the system is full of traps. Probably I am now better off to take a nick without till and kamppeter and another password.
[02:13] <tillkamppeter_> Or is there a possibility to change the password?\
[02:13] <Fujitsu> You can change it.
[02:14] <tillkamppeter_> Or can I unregister and then start over on another day?
[02:14] <Fujitsu> Why would you want to?
[02:14] <jdong> oh keybuk.... where are you.... I think we need to have a talk about the meaning of /sbin/halt... :)
[02:15] <_ion> jdong: What do you mean?
[02:15] <jdong> as in, it quite literally performs a halt, instead of the poweroff we're all used to
[02:16] <jdong> so performing a shutdown from kde/kdm performs a halt
[02:16] <jdong> leaving my laptop batteries draining all night :)
[02:16] <Fujitsu> KDE/KDM shouldn't be using /sbin/halt, should it?
[02:16] <jdong> _ion: bug 59134
[02:16] <Ubug2> Malone bug 59134 in upstart "upstart fails to power off my system" [Untriaged,Confirmed]  http://launchpad.net/bugs/59134
[02:17] <jdong> Fujitsu: I don't know what mechanism kdm uses to shut down, but with upstart it's broken :)
[02:17] <jdong> with sysvinit it worked
[02:17] <jdong> heck with sysvinit, init 0 worked just fine, too :)
[02:17] <_ion> Hmm, sysvinit's shutdown(8) says:        -h     Halt or poweroff after shutdown.
[02:18] <robertj> has anyone seen the patch I mentioned in bug #59327 previously?
[02:18] <Ubug2> Malone bug 59327 in nautilus "please include this patch to add an Empty Trash button to trash:/// simila to 'Write CD' in Burn://" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59327
[02:18] <_ion> The use of "or" is interesting. :-)
[02:18] <jdong> in the 21st century, we tend to power off systems instead of halt them :)
[02:18] <jdong> isn't upstart supposed to be futuristic? :)
[02:20] <pygi> jdong, !!!
[02:20] <jdong> pygi: !!!
[02:21] <robertj> jdong: we need to to add a nice mechnical halting sound on shutdown :)
[02:21] <jdong> lol
[02:22] <Fujitsu_> A lot of machines have one anyway :P
[02:22] <Treenaks> robertj: or a nice, mechanical crash sound on crashing :)
[02:22] <jdong> or put up the Windows 95 brown "It's now safe to turn off your computer" via usplash :)
[02:22] <Spads> screeeeeeeeeCRUNCH
[02:22] <jdong> that'd be great for april fools
[02:22] <Fujitsu_> Yup, sounds good, Spads.
[02:22] <Fujitsu_> No Windows!
[02:22] <jdong> but it's brown :)
[02:22] <Spads> jdong: the SF lugs tried to promote "It is now safe to turn ON your computer" as a Linux slogan, few years back.
[02:22] <Fujitsu_> True.
[02:23] <robertj> Treenaks: if the sound system was still up we could have a national geographic style announcer "Watch as the program pounces on the process and brings it to the ground, dead in its jaws"
[02:23] <cbx33> "It's now safe to plug and play" 
[02:23] <Treenaks> robertj: too bad Keybuk is speeding up shutdown as well as startup :P
[02:23] <robertj> hehe
[02:23] <Fujitsu_> Yeah, shutdown is really quick!
[02:23] <_ion> if (is_april_1st) shutdown_sound = SOUNDS_CRASH;
[02:23] <robertj> "How quick is it!"
[02:23] <Fujitsu_> Sorry, teardown.
[02:25] <robertj> I wonder how long it takes before most soundcards will be online with upstart
[02:26] <robertj> I've just not got enough guts (nor time) to try it out just yet
[02:26] <robertj> Ride of the Valkyries
[02:26] <robertj> hrmm, silly paste buffer, silly xchat-aqua :(
[02:29] <_ion> jdong: Strange, from a quick look at sysvinit's halt.c it looks like it shouldn't poweroff either if it's called as "halt" without -p. I might be reading it wrong, though.
[02:30] <jdong> _ion: well, I'm not sure of the explanation, but kdm used to power down my system before upstart, and now it doesn't
[02:30] <jdong> so I'm blaming it on keybuk :D
[02:31] <_ion> jdong: Is there a reason for instead of "halt -p" or "poweroff" calling "halt"?
[02:31] <jdong> _ion: I'm not sure how kdm shuts down... I'm not sure it even uses halt
[02:32] <jdong> _ion: I'm just observing that halt with sysv did poweroff
[02:32] <jdong> even with no arguments
[02:33] <Kamion> halt calls shutdown to shut down properly - halt -p just tells the kernel to die NOW
[02:33] <Kamion> kdm is right to call halt
[02:33] <Mithrandir> halt should just halt the system, not power it off, iirc.
[02:33] <jdong> Mithrandir: not from my experience... it powers down 
[02:34] <jdong> Mithrandir: but upstart seems to agree with you :)
[02:34] <_ion> kamion: I thought it was -f, not -p.
[02:34] <Fade> Mithrandir -- what is the package/revision of the xlibs you sent up?
[02:34] <Mithrandir> jdong: *shrug*; you probably had buggy hardware or something, then.  It shouldn't.
[02:34] <Kamion> _ion: oh, hmm, you're right
[02:34] <jdong> lol
[02:34] <jdong> so... the verdict... upstart bug or kdm bug? :)
[02:34] <Kamion> _ion: but see /etc/default/halt, which controls whether /etc/rc0.d/S90halt uses -p
[02:34] <Mithrandir> Fade: libx11 (2:1.0.3-0ubuntu3) edgy; urgency=low
[02:35] <Kamion> _ion: kdm still shouldn't use halt -p
[02:35] <Kamion> jdong: upstart
[02:35] <mneptok> IIRC, halt only halts the OS, but the BIOS can have a function to power-off on system halt.
[02:35] <Fade> oh, yeah, I got those. xemacs and gnuemacs are still bjorked.
[02:35] <mneptok> thus, a hlat effectively is a power down on such machines.
[02:35] <robertj> btw, can someone enlighten me on what circumstances can leave ext3 requiring a fsck?
[02:35] <jdong> well, either way, kdm is not turning off my computer :)
[02:36] <Kamion> Mithrandir: halt with no arguments should power-down if /etc/default/halt says "poweroff".
[02:36] <mneptok> jdong: check the BIOS
[02:36] <ogra> jdong, get a new one 
[02:36] <jdong> mneptok: it's not a bios problem
[02:36] <Kamion> the chain is quite twisty but that's what it's meant to do
[02:36] <jdong> if I switch back to sysvinit, kdm happily shuts off my computer
[02:36] <robertj> I thought the whole idea of these fancy-pants new file-systems was that if the power does go off, you shouldn't need to fsck the drive?
[02:36] <jdong> robertj: hardware failure
[02:36] <jdong> robertj: IDE write cache
[02:37] <jdong> robertj: mount-count/day-count fscking set via tune2fs
[02:37] <jdong> robertj: the user typing in fsck.ext3 /dev/hda1
[02:37] <robertj> jdong: hehe :)
[02:37] <Kamion> jdong: my guess is that upstart's halt's option processing is subtly wrong
[02:37] <Kamion> it doesn't *look* wrong to me, but that's the only explanation I can think of
[02:37] <jdong> k
[02:38] <jdong> you gnome users say that gdm powers down correctly?
[02:38] <mempf-edgy> hey guys
[02:38] <jdong> kdm doesn't power down on two of my systems, both using upstart :-/
[02:38] <jdong> so that leads me to believe it's pretty hardware independent
[02:38] <mempf-edgy> i read that Texis insterment card readers were going to be fixed in next kernel update
[02:38] <mempf-edgy> im running the nw kernel but it still doesent ork
[02:38] <jdong> mempf-edgy: what kind of TI card readers, though?
[02:39] <Spads> robertj: http://teh.entar.net/~nick/mail/why-reiserfs-is-teh-sukc has some comments from ted t'so about where many jfses introduce errors
[02:39] <mempf-edgy> Texas Instruments PCIxx21/x515 Cardbus Controller
[02:39] <jdong> mempf-edgy: TI FlashMedia is pretty depressing
[02:39] <robertj> Spads: thanks
[02:39] <mempf-edgy> it was working in dapper....
[02:39] <jdong> mempf-edgy: cardbus or x-in-1 card readers?
[02:39] <mempf-edgy> x-in-1
[02:40] <jdong> and it worked in dapper???
[02:40] <robertj> is the scheduled fsck off in edgy?
[02:40] <mempf-edgy> yes
[02:40] <robertj> because quite simply, that scares the living daylights out of people
[02:40] <mempf-edgy> perfectly
[02:40] <jdong> 0a:09.2 Mass storage controller: Texas Instruments Unknown device 803b
[02:40] <jdong> nvm, mine is different
[02:40] <jdong> mempf-edgy: I'm not sure then... I got a different TI card reader, and it requires the tifm21xx drivers
[02:41] <jdong> robertj: isn't it considered healthy paranoia though?
[02:41] <robertj> nop
[02:41] <robertj> its my mom "whats wrong with my computer"
[02:41] <Fujitsu> robertj, yeah.
[02:41] <robertj> in fact, the black screen being normal is so bad, it vastly outweighs any eror message that might be given
[02:41] <shenki> I can confirm that; the pci id is 0180: 104c:8033, and it used to work with dapper - today discovered it wasn't working with edgy
[02:41] <mempf-edgy> wait hold on
[02:42] <mempf-edgy> when i put a card in it
[02:42] <mempf-edgy> the light goes on for a split second
[02:42] <mempf-edgy> it did not before
[02:42] <Fujitsu> mempf-edgy, can you check syslog?
[02:42] <robertj> if you tell your mom that the black screen is normal, she is likely to not call you if it says you have bad this, or bad that
[02:42] <mempf-edgy> where is it?
[02:42] <Kamion> jdong: the bug is probably that upstart's halt/poweroff implementation doesn't do reboot(RB_ENABLE_CAD) first
[02:42] <robertj> when SMART status would likely be a much better indicator
[02:42] <Fujitsu> tail /var/log/syslog
[02:42] <shenki> Fujitsu, there was nothing in the syslog on insertion for me
[02:42] <Kamion> which sysvinit does
[02:42] <shenki> but I haven't rebooted since installing (todays?) new kernel
[02:43] <Fujitsu> shenki, my Ricoh doesn't mount in Edgy any more, but I get an entry in syslog from the reader.
[02:43] <mempf-edgy> where do you want it?
[02:43] <Fujitsu> (when I insert it)
[02:43] <jdong> Kamion: you want to comment in the bug report?
[02:43] <Fujitsu> mempf-edgy, is there anything regarding the card in there?
[02:43] <mempf-edgy> yep
[02:43] <mempf-edgy> Sep  7 05:42:39 Mempftop kernel: [17180242.956000]  tifm_7xx1: sd card detected in socket 3
[02:43] <mempf-edgy> Sep  7 05:42:39 Mempftop kernel: [17180243.436000]  tifm_7xx1: demand removing card from socket 3
[02:43] <mempf-edgy> Sep  7 05:42:40 Mempftop kernel: [17180244.132000]  tifm_7xx1: sd card detected in socket 3
[02:43] <Kamion> jdong: yes, I'm just doing so
[02:43] <jdong> robertj: smart doesn't say everything....
[02:44] <robertj> jdong: indeed but the current situation just isn't useful to most end users
[02:44] <jdong> holy crap, tifm is in the ubuntu kernel?
[02:44] <jdong> robertj: I'm all for the ext3 periodic fsck, just we need some nicer way of presenting it to the user
[02:44] <robertj> they don't know what the implications of these errors are non-clear to alot of people
[02:44] <jdong> i.e. an explanation of what's going on
[02:44] <jdong> I have seen periodic fscks pick up stuff on ext3 volumes
[02:45] <jdong> especially on older hardware
[02:45] <jdong> and smart did not trigger or come close to triggering
[02:45] <robertj> jdong: it may happen, but people don't want it on by default
[02:45] <robertj> jdong: its dumb
[02:45] <jdong> umm, I suppose most people also want their data?
[02:45] <Kamion> jdong: hmm, actually, may not be quite that, there are a few things involved. I'll comment, anyway
[02:45] <robertj> jdong: how many back it up?
[02:46] <jdong> not that many to be honest
[02:46] <robertj> jdong: not many, ergo they really don't
[02:46] <jdong> why not switch everyone to reiserfs or xfs while we're at it?
[02:46] <robertj> jdong: that's the proof that their time is worth more to their data to them
[02:46] <jdong> heck let's get BenC to add the reiser4 patch to the kernel :)
[02:46] <mempf-edgy> so now the card is being detected
[02:46] <mempf-edgy> but not mounted?
[02:49] <robertj> jdong: what do you think of the windows-way of giving them option to skip?
[02:49] <pygi> anyone willing to test libburn in an application? :)
[02:50] <jdong> robertj: I've got no problem with that
[02:50] <jdong> robertj: right now, doesn't ctrl+c skip the check?
[02:51] <mempf-edgy> oh, there is another line here in lspci
[02:51] <mempf-edgy> 05:09.4 Class 0805: Texas Instruments PCI6411, PCI6421, PCI6611, PCI6621, PCI7411, PCI7421, PCI7611, PCI7621 Secure Digital (SD) Controller
[02:51] <mempf-edgy> not there before
[02:53] <robertj> jdong: I think it mounts it read-only and halts the boot
[02:53] <jdong> ugh :-/
[02:53] <robertj> jdong: I feel like scheduled fscks is a band-aid for an underlying problem
[02:54] <jdong> robertj: yes. x86's are cheap.
[02:54] <jdong> the fix would be a battery-backed-up storage controller
[02:54] <jdong> but who has one of those? ;)
[02:54] <robertj> jdong: but you only need to do that on dirty right?
[02:55] <jdong> well, would you rather fsck on every dirty boot? :)
[02:55] <robertj> jdong: yes
[02:55] <robertj> jdong: that's the way it works on windows
[02:55] <robertj> its good for a number of reasons
[02:55] <robertj> 1. you have limited control over that except for power, etc
[02:55] <jdong> no it's not... ntfs doesn't fsck ever :)
[02:55] <robertj> jdong: it's doing some kind of disk checkup
[02:55] <jdong> not ntfs....
[02:55] <robertj> yes
[02:55] <jdong> it doesn't check unless it gets marked dirty
[02:55] <robertj> jdong: I know
[02:55] <jdong> and that's only when the fs encounters an error
[02:56] <jdong> ext3 has that feature too
[02:56] <jdong> but the problem is, by the time that gets triggered, you've built up enough errors for substantial data loss
[02:56] <robertj> jdong: it seems to happen 100% of the time if I turn off while Windows is running
[02:56] <robertj> jdong: so apparently it is pretty effective in whatever metric it's using
[02:56] <jdong> then something's wrong with your computer
[02:57] <Amaranth> that reminds me, windows seems to check fat32 partitions every 3 boots
[02:57] <jdong> I regularly zap power to my ntfs'es, and they never chkdsk
[02:57] <jdong> Amaranth: huh? it shouldn't
[02:57] <jdong> fat32 is checked on every dirty mount
[02:57] <jdong> what we need is a fsck capable of operating online
[02:57] <robertj> jdong: maybe its because I usually only hard power-off during boot?
[02:57] <jdong> that could do it
[02:58] <jdong> but I've damaged my indexes doing that before
[02:58] <jdong> the worst time to hard-power-off windows is during boot
[02:58] <robertj> jdong: no
[02:58] <robertj> while your installing SP2 is the worst
[02:58] <robertj> (stupid cat)
[02:58] <jdong> lol
[02:59] <jdong> but I do see your point, robertj... the way the periodic fsck is presented to the user currently is anything but user friendly :)
[02:59] <jdong> and while I wish I had the authority to do something about it, I really don't :P
[02:59] <jdong> launchpad is probably gonna be the best place to complain about this one
[03:00] <_ion> jdong: I found the culprit. /etc/event.d/rc0-halt contains "env INIT_HALT=HALT", which (in rc0.d/S90halt) overrides the setting in /etc/default/halt.
[03:00] <Spads> When FreeBSD got softupdates, they added hooks to fsck so that it would snapshot the filesystem during the fsck and release after.  Don't people have tools to do something like this for ext3?
[03:00] <jdong> _ion: aha
[03:00] <jdong> Spads: not without lvm and free space
[03:00] <Spads> a friend of mine ran fsck from cron.monthly, as I recall
[03:00] <Spads> ah, lvm...
[03:01] <jdong> Spads: that doesn't work though
[03:01] <jdong> Spads: you must snapshot it while it's mounted ro
[03:01] <jdong> then mount it rw and continue bootup
[03:01] <jdong> but once you're booted, you can't put it back to ro easily
[03:01] <jdong> without disrupting everyone
[03:01] <Mithrandir> that only works if you don't get any errors.
[03:01] <jdong> and fscking an rw'ed ext3 partition will always get errors
[03:01] <Spads> because of smearing
[03:01] <Spads> yeah
[03:02] <jdong> and softupdates are not magical on x86 either
[03:02] <jdong> I've had cases where after a 2nd reboot FreeBSD went ahead and did a full fsck
[03:02] <jdong> and found problems, too
[03:02] <Spads> oh, believe me, I've no illusions about softupdates
[03:03] <jdong> I am guessing the ide write cache is the culprit there, too
[03:03] <Spads> I took Kirk McKusick's kernel class during a period where he'd pretty much come to the realization that the tradeoffs for SU vs a JFS just plain weren't worth it
[03:06] <mempf-edgy> il file a bug if you guys can tell me what logs and stuff  i should include
[03:07] <jdong> _ion: you gonna put what you found in the bug report?
[03:07] <jdong> _ion: 59134
[03:09] <_ion> jdong: Done.
[03:09] <mempf-edgy> how do i manully mount a sd card?
[03:23] <cbx33> hey seb128 
[03:31] <synic> How does one go about getting an app into the repos?
[03:31] <Kamion> mvo_: ok, I've been looking over that branch. Just from inspection, it looks mostly correct, although the control flow with regard to whether '?' was seen in front of a seed entry is a little confusing; I don't know if anything can be done about that. Maybe add a comment
[03:31] <Kamion> mvo_: have you been able to test it at all?
[03:32] <Hobbsee> synic: see the /topic - ask in #ubuntu-motu
[03:32] <mneptok> synic: all repo modifications are delivered via Jarts
[03:32] <Hobbsee> mneptok: jarts?
[03:32] <mneptok> http://mowabb.com/aimages/images/07-06-04.jpg
[03:33] <Hobbsee> ah
[03:33] <mneptok> and here's a set - http://tonova.typepad.com/thesuddencurve/images/Jarts-thumb.jpg
[03:34] <mneptok> lawn darts. now illegal in the US, iirc.
[03:36] <Spads> illegal how?
[03:37] <zul> probably you cant take them on airplanes for one
[03:37] <seb128> cbx33: hi
[03:37] <dholbach> re
[03:37] <seb128> hi dholbach
[03:37] <cbx33> hi again dholbach 
[03:38] <mneptok> Spads: illegal as in "you cannot manufacture or sell them new"
[03:38] <Spads> hmmm
[03:39] <mneptok> Spads: i doubt the FBI would kick your door in for owning Jarts, but then this is 2006 ....
[03:39] <Spads> right
[03:39] <cbx33> we used to make warts!
[03:39] <cbx33> wall-darts
[03:40] <Spads> it's just that saying "these are illegal in the US" makes it sound like they're contraband items across the country
[03:40] <cbx33> drawing pin and a little origami, mixed with some blu-tack !
[03:40] <Spads> saying "these were recalled by court order due to infant injury claims" or whatever is something else
[03:41] <mneptok> Spads: if you take off your tin-foil hat the sentences sound pretty similar. :)
[03:41] <mneptok> but then, this is 2006 ...
[03:42] <Spads> mneptok: so you're saying that if you allow in the government elf-rays, your brain interprets only two levels of legality?
[03:42] <Spads> I would say that you are a conspiracy nut!
[03:42] <mneptok> Spads: i would say you are either entirely correct, or should be shipped to Guantanamo Bay
[03:43] <mneptok> mmm .... binary.
[03:43] <Spads> you're either with us or you're with the aliens
[03:44] <mneptok> i'm with you. i tried the aliens, and the hours sucked. and they have crazy unions.
[03:47] <sladen> mneptok: crazier hours than Canonical?
[03:47] <mneptok> touche.
[03:52] <Whoopie> Hi, did anybody look at the issue that some packages don't arrive in dapper-backports or dapper-proposed although they were build successfully? e.g. xchat in dapper-backports or openoffice.org-l10n in dapper-proposed.
[03:53] <jdong_> Whoopie: it's an effect of bug 58144, last I heard
[03:53] <Ubug2> Malone bug 58144 in soyuz "Backport is rejected if an older backport is already there" [Critical,In progress]  http://launchpad.net/bugs/58144
[03:54] <Whoopie> jdong_: ok, thanks.
[03:55] <Whoopie> jdong_: strange, that it also affects dapper-proposed
[03:55] <jdong_> Whoopie: I'm not sure about proposed, though
[03:55] <jdong_> Whoopie: I wouldn't be surprised if it's the same bug though
[03:57] <Kamion> jdong_: no, it's not
[03:57] <Kamion> er, well, would need to check but it seems unlikely
[04:00] <Kamion> oh, it's easy
[04:01] <Kamion> Whoopie: openoffice.org-l10n is just in NEW, that's all - it built binaries which weren't previously in dapper
[04:01] <jdong_> hehe
[04:01] <Kamion> xchat is as jdong says, an entirely different issue; those were rejected
[04:02] <pitti> Kamion: can you please approve the hoary-updates langpacks?
[04:02] <Whoopie> Kamion: so openoffice-org-l10n needs to be approved?
[04:02] <Kamion> to be precise, openoffice.org-help-pl was added
[04:02] <Kamion> Whoopie: yes; I'm doing it now
[04:03] <Whoopie> cool, thanks.
[04:03] <Kamion> pitti: I'm reluctant to do (non-proposed) stable release updates until we've finished doing the post-mortem of the X update
[04:03] <Kamion> pitti: mdz explicitly asked me not to process dapper-updates until that was done, and while he wasn't explicit about hoary-updates, I'd rather be conservative
[04:04] <pitti> Kamion: alright
[04:04] <pitti> Kamion: post-mortem means, the exercise how to quickly disable an -updates version by adding a fake version into *-security Packages.gz?
[04:04] <bluefoxicy> sweet, compressed caching for 2.6.18-rc6 has page cache compression
[04:05] <Mithrandir> pitti: more like "what went wrong and how are we going to prevent it from happening in the future".
[04:05] <bluefoxicy> Too bad Nitan doesn't want to shoot this at mainline when it's done.
[04:15] <Kamion> pitti: what Mithrandir said. A post-mortem (colloquially) is a study you do after something went wrong to find out what Mithrandir said.
[04:23] <Kamion> wow, running setupcon from within ubiquity is rather fun
[04:23] <Kamion> especially the screen flicker as it switches from tty to tty
[04:23] <Kamion> must fix that
[04:23] <Mithrandir> Kamion: just don't load the font in ubiquity?
[04:24] <Kamion> whoa, and it's totally buggered my keymap
[04:26] <Kamion> Mithrandir: surely loadkeys needs to change tty as well
[04:26] <Mithrandir> Kamion: hmm, maybe.
[04:28] <Hobbsee> who's working on usplash, and do they know that it's not displaying boot messages at all now, since 0.4-17?
[04:29] <Hobbsee> or should i just file a bug over this?
[04:29] <Hobbsee> mjg59 it looks like
[04:30] <mvo_> Kamion: thanks for looking over it. I did did it and marked openoffice as a recommend. that worked nicely. I also rebuild ubuntu-meta and it produce correct new relations. I have not tested apt yet with it because we would need the new section "metapackages" first. but I did test the apt behaviour for other packages
[04:31] <mjg59> Hobbsee: That's correct behaviour
[04:31] <mjg59> Remove quiet if you want messages
[04:32] <Hobbsee> mjg59: ah right, so what's the point of the black box then?  just artwork?
[04:32] <mjg59> Yes
[04:32] <Hobbsee> gotcha.  cool :)
[04:32] <seb128> infinity: did you plan to try getting a new samba version of edgy? nautilus-share (which is a cool tool to share folders on smb from nautilus) requires 3.0.23 and we have 3.0.22 
[04:32] <gnomefreak> Hobbsee: i just filed bug on that :)
[04:33] <gnomefreak> ill close it
[04:33] <Hobbsee> gnomefreak: guess you'd better reject it.  you must have only just done it, as when i checked a min or so ago, there was nothing there
[04:33] <gnomefreak> i did
[04:33] <Hobbsee> gnomefreak: maybe with an explanation - that's one that's likely to get plenty of dupes
[04:33] <mvo_> Kamion: one issue with the "?" is that filterPackages support shell-like globing for packages and "?" is than not a good choice. me have have to use something else
[04:33] <gnomefreak> true
[04:36] <Kamion> mvo_: let's just do "optional: package"
[04:36] <Kamion> oh, no, let's not
[04:36] <Kamion> that syntax is used for fields
[04:37] <Kamion> mvo_: "(package)"?
[04:37] <mvo_> maybe "weak:"? "~"
[04:37] <mvo_> I like the "(pkg)" idea!
[04:38] <Kamion> yeah, feels right
[04:38] <giftnudel> is there a known reason (upstart?) that my laptop doesn't try to resume from my swap partition after hibernate?
[04:38] <Kamion> Mithrandir: hmm, setupcon isn't even supposed to do anything unless it's run from a console ...
[04:39] <madduck> giftnudel: cat /etc/initramfs-tools/conf.d/resume
[04:39] <mvo_> Kamion: ok, added. I will do some more testing now. how should I proceed? just upload it if I think its fine? or wait for you to merge and upload?
[04:40] <giftnudel> madduck: no such file
[04:40] <Kamion> mvo_: upload it as 0.21ubuntu1, please - I'll sort out the rest later
[04:40] <madduck> create it ("RESUME=/dev/sdb2" here), then update-initramfs -u
[04:40] <madduck> the try again
[04:40] <giftnudel> madduck: thanks, will try that
[04:41] <madduck> giftnudel: at least this is what Debian does
[04:41] <Kamion> mvo_: fortunately in this case it isn't important to have it installed on DC systems
[04:41] <Kamion> madduck: yes, same for Ubuntu
[04:41] <giftnudel> hmm, why don't I have that file then
[04:42] <giftnudel> and I don't have it in dapper either
[04:42] <Kamion> giftnudel: it's created on installation; upgrades from older systems may not have it
[04:42] <Kamion> or you may not have had swap around during installation
[04:42] <giftnudel> funny, but it always worked ....
[04:42] <Kamion> there are several possible reason
[04:42] <Kamion> s
[04:42] <giftnudel> hmm
[04:43] <Kamion> or there's a bug and something removed it, or or or ...
[04:49] <Mithrandir> Kamion: true, but it's explicitly given one from the config script.
[04:51] <Kamion> Mithrandir: OH
[04:52] <Mithrandir> Kamion: maybe that's a bit too big a sledgehammer
[04:52] <Kamion> nadgers
[04:52] <Kamion> let me ponder that over late lunch
[04:53] <Mithrandir> the evil solution is have ubiquity set UBIQUITY_RUNNING=1 or something and then poke that.
[04:53] <bddebian> Hello
[04:58] <Kamion> Mithrandir: I think the right answer is to have rootskel's /etc/profile call setupcon (it already sets the console to UTF-8 mode) and remove that change from console-setup.postinst.
[05:00] <Mithrandir> Kamion: ok
[05:02] <doko> Kamion: please could you process #59049, translate-toolkit sync request. needed for the next OOo upload
[05:02] <netdur> guys!!! you forgot to release ubuntu preview!!!!!!
[05:02] <mdz> Kamion: yes, that applies to *-updates
[05:02] <Mithrandir> netdur: hm?
[05:02] <bluefoxicy> Cool, you can request feedback on specifications.
[05:03] <netdur> you used to release ubuntu preview the same day gnome release
[05:03] <Mithrandir> netdur: betarelease isn't until sept 28th
[05:03] <Kamion> Mithrandir: it does mean that if you have a running shell then it won't be reconfigured, but I think that's ok
[05:03] <Mithrandir> Kamion: agreed
[05:03] <jdong> netdur, go grab a nightly :)
[05:05] <mdz> bluefoxicy: you can, but it doesn't email the person, and as such isn't very useful
[05:05] <bluefoxicy> mdz:  ah, so it's just until you happen to click the spec and see "hey this guy wants feedback from you"?  bleh :P
[05:06] <netdur> jdong, sweet
[05:06] <bluefoxicy> I don't have a real reason anyway, I'm just fooling around
[05:06] <mjg59> Right
[05:06] <mjg59> So
[05:06] <mjg59> usplash should now be happier
[05:06] <bluefoxicy> mjg59:  did you teach it to not arserape my screen?
[05:07] <mjg59> I believe so
[05:07] <jdong> mjg59: you mean it'll stop turning off my screen and/or corrupting all my vt's? :)
[05:07] <bluefoxicy> I'm tired of shutting down and going from gnome desktop to a pink screen
[05:07] <jdong> at least the former masks the latter :)
[05:07] <mjg59> It works on the hardware I have here
[05:07] <jdong> wow, when have we heard that before....
[05:07] <jdong> *cough* xorg * cough*
[05:07] <mjg59> I would expect it to work elsewhere, but it's hard to be certain
[05:08] <mjg59> Oh, unless you're using vga=. Then you lose.
[05:08] <jdong> lol
[05:08] <jdong> nope
[05:08] <mjg59> I'll worry about that at some later date
[05:08] <bluefoxicy> jdong:  I suggested kdrive before but :>  The overhead of bringing up a fully functional vesa X server is probably insurmountable.
[05:08] <mjg59> What would kdrive buy us?
[05:08] <jdong> well, when -18 comes rolling in, I'll test out usplash
[05:08] <bluefoxicy> Fedora does that but uses full out Xorg and it takes 42 seconds longer to boot because of it I hear
[05:09] <mjg59> usplash overhead is between 0 and a second, based on my measurements
[05:09] <bluefoxicy> mjg59:  it's an actual X server, rather than an attempt to write on the frame buffer console or use of svgalib etc; I doubt it'd buy you anything aside from being able to say you're incredibly lazy.
[05:09] <bluefoxicy> Probably slow, like with Fedora
[05:10] <jdong> lol
[05:10] <mjg59> bluefoxicy: It's an actual X server that makes vesa calls in an identical manner to svgalib
[05:10] <mjg59> Anyway. svgalib is much nicer now that I've thrown out all the crack.
[05:10] <bluefoxicy> as long as it's behaving now I'm happy.
[05:11] <bluefoxicy> oh, what the hell
[05:11] <bluefoxicy> I should find out what happened to xscreen
[05:11] <jdong> hey, anyone knows what it takes to get those cool discharge graphs in gnome-power-manager?
[05:11] <bluefoxicy> that was a google SoC this year
[05:12] <Kamion> Mithrandir: actually, hmm, running setupcon at shell startup will do the vt switch thing, which will be annoying; I guess it would be ok to check whether we're really in d-i by means of checking some other path
[05:12] <jdong> my core duo doesn't show em, while my toshiba does
[05:12] <Kamion> e.g. /lib/debian-installer
[05:13] <mjg59> jdong: Is your core duo a laptop?
[05:13] <jdong> mjg59: yes...
[05:13] <jdong> I get a device information and event log tab
[05:13] <mjg59> And if you run on battery, does something appear under "current rate" in /proc/acpi/battery/whatever?
[05:14] <bluefoxicy> http://code.google.com/soc/xorg/appinfo.html?csaid=73A89F18E7770493
[05:14] <bluefoxicy> There we go.
[05:14] <jdong> mjg59: present rate, yes...
[05:14] <mjg59> jdong: Unsure, then
[05:14] <mjg59> I'll take a look at some stage
[05:20] <Spads> IS CHICAGO
[05:20] <Spads> IS NOT CHICAGO
[05:21] <sladen> mjg59: is svgalib working okay on amd64?
[05:21] <mjg59> sladen: In general
[05:21] <iwj> seb128: AYT?
[05:21] <mjg59> May die horribly on some nvidias
[05:22] <mjg59> Someone needs to debug that
[05:22] <mjg59> It's x86emu, not svgalib as such
[05:22] <bluefoxicy> hmm.  on i386 it looks like I have "generic kernel for x86-64"
[05:22] <knot_jdong> bluefoxicy: shh.... mornfall is looking for me....
[05:22] <iwj> seb128: You seem to have dropped my patch for SuggestedPackagesForFiletypesSpec from nautilus.  Why ?
[05:23] <mdz> doko: ping
[05:26] <thom> Keybuk: http://www.clearairturbulence.org/edgy-20060907-1.png as requested
[05:26] <doko> mdz: pong
[05:26] <thom> Keybuk: (for ath0 not coming up correctly)
[05:26] <seb128> iwj: ups
[05:26] <Keybuk> thom: thanks
[05:27] <iwj> seb128: It's strange because you left the entry in the changelog.
[05:27] <seb128> iwj: because you didn't put it to debian/patches as we do for GNOME packages usually
[05:27] <seb128> iwj: yeah, and I just copy debian/ over from one version to the new one
[05:27] <iwj> That's not a sound approach.
[05:27] <seb128> iwj: we don't package changes out of the debian dir usually
[05:28] <iwj> Because people (like all Ubuntu developers) are entitled to dpkg-source -x, edit, upload.
[05:28] <seb128> changes to the .diff.gz are evil
[05:28] <iwj> And we don't have time to learn how this package does its stuff.
[05:28] <iwj> If you really think that then you can turn it into a stupid debian/patches thing.
[05:28] <seb128> iwj: simple, drop the diff to debian/patches
[05:28] <Keybuk> iwj: I'd disagree ... if you edit a source, you should take 5s to figure out how to apply a patch to it
[05:28] <Keybuk> that's basic common sense
[05:28] <mjg59> And then figure out if it needs numbers at the front, if it needs to end in .patch, if it needs to be listed in 00files...
[05:28] <iwj> seb128: There are THOUSANDS of packages.  It's NOT ON to require weirdshit knowledge of specific package build systems.
[05:29] <mvo_> iwj: I'm with seb128 here, there is a pretty consitent style over all gnome packages
[05:29] <iwj> Keybuk: There is a standard way to apply patches to sources.  dpkg-source -x; edit; upload.
[05:29] <mvo_> bye zyga
[05:29] <seb128> iwj: so send the patch on launchpad next time and somebody knowing the package will apply it
[05:29] <Keybuk> iwj: that has not been standard for quite a while
[05:29] <zyga> mvo_: bye, talk to you in the evening :)
[05:29] <Keybuk> iwj: by numbers, debian/patches is now the standard
[05:29] <iwj> Only because of idiocy.
[05:29] <Keybuk> *shrug*
[05:29] <mvo_> *cough*
[05:30] <HiddenWolf> !coc
[05:30] <seb128> iwj: yeah, and 5 people do that with different changes and then you can't figure what changeset is doing what
[05:30] <iwj> seb128: Perhaps you and gnome upstream can't.
[05:30] <seb128> yeah, and I'm maintaining that package
[05:30] <iwj> I and Mike Hommy have been doing firefox like this and figuring out which patch is what is _not_ the problem.
[05:31] <iwj> dpkg-source -x should produce the source code ready for editing.
[05:31] <iwj> That's what it's for.
[05:31] <seb128> debian/patches allow to make clear separation of changes you apply
[05:31] <seb128> and dropping a patch to it is not that hard
[05:31] <iwj> Anyway, edits in the .diff.gz are easy to spot.  If you don't like them you can write some tool to include them in debian/patches or something.
[05:31] <seb128> anyway, I'll fix that one
[05:31] <seb128> if you prefer just mail me the patch next time
[05:31] <seb128> or attach it to a bug
[05:31] <iwj> Yay!  The Maintainer Lock is back!
[05:32] <seb128> yeah, I just didn't think somebody went to change the diff.gz when I updated
[05:32] <seb128> no, it's not
[05:32] <pitti> iwj: personally I consider the lack of a standard patch system a bug in Debianish source packages; that's the sole reason why we do have so many patch systems
[05:32] <Keybuk> iwj: there's no need for a maintainer lock if you take the minimum of care when changing source
[05:32] <iwj> Of course no-one would edit a package in the normal way that packages are edited.
[05:32] <seb128> I was just trying why I dropped your patch by mistake
[05:32] <mvo> iwj: please, as I said, gnome is having a consitent policy for those things. it makes maintaing the big amount of packages eaiser
[05:32] <Kamion> sure it is, if you're not even bothering to check for updates!
[05:32] <Keybuk> iwj: you wouldn't patch C code with a different coding-style to the one upstream used
[05:32] <Kamion> that might as well be a maintainer lock
[05:32] <Keybuk> so why patch packages with a different patch-style differently?
[05:33] <pitti> iwj: I don't consider inline diff.gz patches the 'normal way' either - it's evil, it's hard work to upgrade to a new upstream version, it's hard to check patches and send them around, etc.
[05:33] <gnomefreak> mjg59: you have a sec?
[05:33] <iwj> hard> I don't find it so.
[05:33] <iwj> See rants on debian-devel passim.
[05:33] <Keybuk> seb128: you should really check the package before just copying the debian/ dir out though
[05:33] <iwj> Patch systems are for people who can't grok code.
[05:33] <seb128> Keybuk: I did a mistake, no big deal
[05:34] <seb128> I'm going to fix it now ;)
[05:34] <Keybuk> seb128: in fact, I seem to recall you once didn't even check for a new version of the package in the archive before doing an updated version
[05:34] <seb128> Keybuk: doesn't ring a bell to me
[05:34] <iwj> Keybuk: That happens sometimes too; it's because we have no machinery to prevent it.
[05:34] <pitti> iwj: 'Patch systems are for people who can't grok code.' -> heavy NACK; that might be your personal opinion, but it's not common sense
[05:34] <dholbach> seb128 does no mistakes
[05:34] <seb128> Keybuk: maybe that was "somebody uploaded while I was working on the update and I didn't check again"
[05:34] <iwj> In Debian the maintainer lock prevents it but in Ubuntu you can do it by mistake because the process is racey.
[05:35] <iwj> Anyway:
[05:35] <Keybuk> yes, yes, the process sucks
[05:35] <Keybuk> NEWS AT 11
[05:35] <iwj> seb128: Thanks for fixing it up.
[05:35] <Keybuk> :p
[05:35] <seb128> iwj: np, sorry for dropping it
[05:35] <iwj> I'll try to remember that gnome packages have patch nightmares.
[05:35] <pitti> NoMoreSourcePackages!!!11!one!
[05:35] <pitti> :)
[05:35] <mvo> yeah!
[05:35] <iwj> Do you need any help ?  I have lots of diffs and old versions and stuff lying around here.
[05:35] <Keybuk> pitti: mmmm, vapourware :-/
[05:35] <seb128> I would no call "putting a diff to debian/patches" a nightmare ;)
[05:35] <pitti> Keybuk: 'import future'
[05:36] <iwj> We'll just have to disagree about the patches but obviously I can try to remember for these packages.
[05:36] <pitti> iwj: in general, the existance of debian/patches/ dir indicates a patch system; and IMHO it's not hard to check for that dir
[05:36] <seb128> iwj: no, that's fine, I've the previous version on the disk and the change is fairly trivial
[05:36] <seb128> thank you 
[05:36] <_ion> mvo: Yeah, cdbs-edit-patch is very handy.
[05:37] <iwj> Keybuk: Soyuz should check that the previous ubuntu version is mentioned in the changelog (with a full-text grap).
[05:37] <iwj> s/grap/grep/
[05:37] <Keybuk> iwj: write a spec ;)
[05:37] <iwj> Keybuk: I might do :-).
[05:38] <pitti> Keybuk: NoMoreSourcePackages would make this spec obsolete, too, right?
[05:38] <iwj> seb128: Could you pass me the new .diff/.dsc under the table ?
[05:38] <iwj> I want to do some last-minute faff :-).
[05:38] <Keybuk> pitti: in theory
[05:41] <mjg59> gnomefreak: Hi
[05:42] <gnomefreak> mjg59: hi. is it usplash that has to do with tty?
[05:42] <mjg59> No
[05:42] <Hobbsee> gnomefreak: what about tty?
[05:42] <gnomefreak> it doesnt work
[05:42] <gnomefreak> none of them
[05:42] <Hobbsee> sigh.
[05:42] <Hobbsee> "doesnt work" is useless.
[05:43] <mjg59> Hobbsee: Should be fixed
[05:43] <Hobbsee> mjg59: cool, okay, i'll try rebooting later today.
[05:43] <mjg59> Hobbsee: -18 has saner code
[05:43] <Hobbsee> yay :)
[05:43] <gnomefreak> blinking cursers depending what tty you use depends where curser is
[05:44] <Hobbsee> mjg59: any ETA on when that will hit the archives?
[05:45] <jdong|laptop> Hobbsee: it'll hit my archives in around 45 seconds :)
[05:45] <mjg59> It's uploaded
[05:45] <Hobbsee> mjg59: cant see it on LP
[05:45] <Hobbsee> maybe i'm blind
[05:45] <jdong|laptop> Hobbsee: usually lp lags a few hours
[05:46] <Hobbsee> right
[06:01] <seb128> iwj: http://people.ubuntu.com/~seb128/nautilus-update/
[06:04] <Kamion> back
[06:05] <iwj> seb128: Thanks a lot.
[06:05] <jordi> jono: haha, cool pic :)
[06:05] <jono> jordi, :P
[06:06] <seb128> iwj: np
[06:07] <iwj> seb128: dpkg-source: error: file nautilus_2.16.0-0ubuntu2.diff.gz has size 740219 instead of expected 114796
[06:07] <iwj> !
[06:08] <seb128> iwj: weird
[06:09] <iwj> Strange.
[06:09] <iwj> My w3m downloaded a huge version.
[06:09] <iwj> wget worked.
[06:09] <seb128> it has the correct size on rookery
[06:09] <seb128> k
[06:09] <seb128> weird
[06:11] <Seveas> mjg59, ping
[06:12] <Kamion> doko: translate-toolkit done
[06:13] <lemsx1> i'm glad to see StickyNotes applet again! 
[06:15] <jordi> heya Huahua 
[06:15] <Huahua> hey, jordi 
[06:18] <Amaranth> iwj: ping
[06:19] <iwj> Amaranth: ponb
[06:19] <iwj> Amaranth: pong even
[06:19] <Amaranth> iwj: I'm trying to find a way to force firefox to use a specific proxy (and not let users turn it off), http://in-cider.spaces.live.com/blog/cns!1F17474AB1F2CE52!338.entry is all I've found on it so far. Does that look like a bad solution?
[06:21] <lemsx1> Amaranth: this is not the right place for those questions. but what you need to do is learn to use user_prefs in the .js file. put the proxy information there and make the file read-only (or use a global file from /etc)
[06:21] <Amaranth> lemsx1: Sure it is, I'm trying to make a package do it. :P
[06:21] <lemsx1> Amaranth: in Windows you can do other things like a pseudo-encryption of the .js file after you set the settings as you want. google it
[06:21] <iwj> What stops the user downloading ff from the official site and installing it in their filespace ?
[06:21] <Amaranth> Nothing. :/
[06:21] <Amaranth> I hate firefox.
[06:22] <iwj> So use a firewall.
[06:22] <lemsx1> Amaranth: all apps in UNIX would go through the same issues. not just Fx
[06:22] <Amaranth> Well, one thing stops them: I'm blocking port 80 out for everyone except the daemon user
[06:22] <ogra> iwj, the iptables rule we talked about before wont let them browse unless they set their proxy settings right
[06:23] <ogra> it wont be intercepting, but port 80 will surely be blocked
[06:23] <kristog> ogra: hey :) did you read my query ?
[06:23] <lemsx1> Amaranth: iptables allows redirect to 3128 for transparent proxies
[06:23] <iwj> ogra: Right.
[06:23] <ogra> so even if the install in ~/ it will work 
[06:23] <iwj> Amaranth: So why do you care if the user finds some way to override the proxy settings and breaks their own web browsing ?
[06:23] <lemsx1> Amaranth: it's actually really easy to do with shorewall
[06:23] <Amaranth> iwj: Well, ok, forcing it isn't required.
[06:23] <iwj> lemsx1: Intercepting proxies are Evil, Bad and Wrong.
[06:23] <ogra> he wants it fully automatic
[06:23] <lemsx1> iwj: but it works
[06:23] <iwj> Amaranth: So just drop it in /etc/firefox/pref or whatever the dir is.
[06:24] <iwj> lemsx1: No it doesn't.  RTFRFC.
[06:24] <Amaranth> I just want to automatically set it without touching the user profile.
[06:24] <ogra> lemsx1, it doesnt ... dont start that discussion again pease 
[06:24] <lemsx1> iwj: I've done that setup at home. it works. use shorewall
[06:24] <ogra> *please
[06:24] <lemsx1> ok
[06:24] <ogra> thanks :)
[06:24] <lemsx1> iwj: you are right. it doesn't work
[06:24] <iwj> lemsx1: Come back when you've read and understood the RFC which explains how it fails.
[06:24] <Amaranth> Well, it does work 99% of the time.
[06:25] <Amaranth> That 1% will kill you though.
[06:25] <iwj> Of course the web is randomly flakey anyway so you probably don't notice an increase in flakiness.
[06:25] <iwj> Amaranth: Sure.  /etc/firefox/pref, create a file called your-thing.js containing appropriate pref() things.
[06:25] <Amaranth> alright
[06:25] <Amaranth> now the fun part is boosting the proxy to root to install that file then dropping again :)
[06:27] <ogra> Amaranth, probably keep that for edgy+1 ?
[06:27] <mvo> Kamion: I uploaded germinate and will upload ubuntu-meta next with the support for recommends (no seed changes yet,) just so that the next ./update will do the right thing(tm)
[06:27] <Amaranth> ogra: I suppose
[06:27] <ogra> smells a bit like suid hell
[06:27] <Amaranth> i'll upload the package i have now then
[06:27] <Amaranth> i'll work on that in a separate branch or something
[06:27] <iwj> Amaranth: Why not ship the file as part of the proxy package or the config package or something ?
[06:27] <ogra> right
[06:28] <ogra> iwj, the gui has a checkbox ...
[06:28] <Amaranth> iwj: Well, the configuration GUI for the proxy has this little checkbox to turn the rules on/off
[06:28] <ogra> it will en/disable the proxy
[06:28] <iwj> Ahh.  userv is the answer :-).
[06:28] <Kamion> mvo: ok; please bump the versioned build-dependency on germinate
[06:28] <ogra> hehe
[06:28] <mvo> Kamion: right, will do
[06:29] <ogra> iwj, so why isnt it in main then ? :P
[06:31] <iwj> Nothing in main is using it and we don't care enough about server admins who are weirdos enough about security to want it.
[06:32] <iwj> Yay, seb128's new nautilus breaks it totally as I expected, by exposing the gnome-app-install breakage.
[06:32] <iwj> So excellent, my test case has arrived.
[06:32] <seb128> iwj: break what?
[06:32] <seb128> the feature?
[06:32] <iwj> You unbroke the attempt to use the feature.
[06:32] <iwj> But gnome-app-install is broken too.
[06:33] <iwj> Don't worry about it, it's not your fault :-).
[06:33] <seb128> pfiouuu ;)
[06:33] <seb128> we can blame mvo then :p
[06:33] <iwj> I go away for a week and look what happens ...
[06:35] <gnomefreak> what is the package to file a bug on the alternate installer?
[06:35] <Kamion> gnomefreak: the catch-all package is debian-installer
[06:35] <gnomefreak> ok ty
[06:37] <iwj> Woah, bzr blame is actually called bzr blame.
[06:38] <Spads> haha
[06:39] <iwj> I tried `bzr ann' and that didn't work, and I didn't fancy typing `annotate' every time ...
[06:42] <jdong|laptop> alias*
[06:42] <dholbach> or add it to bash completion :)
[06:48] <LarstiQ> Isn't ann aliased?
[06:49] <LarstiQ> it is in bzr.dev at least
[06:49] <LarstiQ> and also 0.9
[06:52] <doko> seb128, dholbach: is gtk-clearlooks-gperfection2-theme the replacement for gtk2-engines-clearlooks?
[06:53] <dholbach> doko: no
[06:53] <dholbach> doko: gtk2-engines has the "real thing"
[06:54] <Kamion> Keybuk: any idea why udevinfo -q name -p /block/loop0 says that there's no record for /block/loop0 in the database?
[06:54] <Kamion> sorry, make that loop1, better example
[06:54] <Keybuk> Kamion: because nothing interesting happened to it, usually
[06:54] <Kamion> I know udev knows about the device because:
[06:54] <Kamion> brw-rw---- 1 root disk 7, 1 2006-09-07 17:46 /dev/loop1
[06:54] <Kamion> which is about when I did 'modprobe loop'
[06:55] <Keybuk> right
[06:55] <Kamion> Keybuk: shouldn't it be able to tell me the associated device name?
[06:55] <Keybuk> no symlinks, no programs, etc
[06:55] <Keybuk> Kamion: if udevinfo returns no information, then the device name is the kernel name
[06:55] <Kamion> oh, I guess this is why casper has an ||
[06:55] <Kamion> ok, fair enough
[06:55] <Keybuk> it's an odd "optimisation" that udev doesn't store things in its database if it didn't rename them
[06:56] <Keybuk> if udevinfo doesn't return a name, assume basename of the DEVPATH basically
[06:57] <Kamion> ok, thanks
[06:57] <Keybuk> it increasingly rarely comes up these days, because we have so many little helper programs that look at block devices
[06:57] <Keybuk> but it still can come up when it's just a plain device with no interesting features
[06:58] <Keybuk> udevinfo -q name -p /class/mem/null
[06:58] <Keybuk> ^ e.g. :p
[07:00] <jdong|laptop> Keybuk: are you gonna make upstart shut down right? ;-)
[07:00] <Keybuk> jdong|laptop: I think we've figured that bug out :p
[07:00] <jdong|laptop> Keybuk: rc0-halt's envirnment override?
[07:00] <Keybuk> indeed
[07:00] <jdong|laptop> *rephrases* Keybuk: are you gonna upload a new upstart that fixes that problem?

[07:01] <Keybuk> I always assumed "halt" did what it said on the tin, and halted the machine
[07:01] <Keybuk> so I implemented it that way <g>
[07:01] <jdong|laptop> lol, make sense :)
[07:01] <Keybuk> yeah, there will be a new upstart today that fixes most, if not all, of the open bugs
[07:02] <Keybuk> it's a bit kooky that kdm calls "halt" instead of "shutdown" though :-/
[07:02] <Keybuk> on most UNIXen that will literally just halt the machine
[07:03] <jdong|laptop> heh, don't blame me for that one :)
[07:03] <jdong|laptop> it's kdm... what can I say? :)
[07:03] <pitti> G0SUB: ping?
[07:04] <G0SUB> pitti: hello
[07:05] <Riddell> ogra: is willowng the edubuntu web filter thing?
[07:05] <doko> Kamion, mdz: I would like to upload OOo 2.0.4 release candidate 1; I don't see major problems on i386, amd64 (ia32) and powerpc.
[07:05] <ogra> Riddell, yep
[07:05] <Amaranth> Riddell: yeah
[07:06] <Amaranth> now with KDE frontend ;)
[07:06] <Riddell> Amaranth: that's what I just noticed
[07:06] <Riddell> Amaranth: written by you?
[07:06] <lamont> pitti: ping
[07:06] <pitti> lamont: pong
[07:06] <Amaranth> actually, no, i never even tried it :P
[07:06] <Riddell> Amaranth: who made it?
[07:06] <Amaranth> LaserJock pointed me to a repo, said it works for him, i merged :)
[07:07] <lamont> pitti: wondering why you didn't just sync bind9 from sid like I suggested for edgy....
[07:07] <Amaranth> Robert Ray is his name, I think
[07:07] <Amaranth> i've never seen him on irc
[07:07] <Riddell> Amaranth: interesting
[07:07] <Kamion> doko: I'd like to leave that decision to mdz
[07:07] <pitti> lamont: we can still do that if we can drop the ubuntu changes; I didn't check
[07:07] <LaserJock> Robert Day
[07:07] <pitti> lamont: I just uploaded the same version to dapper-security and edgy
[07:07] <Riddell> Amaranth: got a URL?
[07:07] <lamont> the ubuntu changes were all obsoleted by switching to lsb functions in 9.3.2-2
[07:07] <pitti> ah, cool
[07:07] <lamont> pitti: ah, ok
[07:08] <lamont> I guess I'll forgive you then
[07:08] <doko> Kamion: I need just an ok, not a discussion ;-)
[07:08] <Amaranth> the one in the 0.3 package is ugly codewise, i have new patches from him that fix it
[07:08] <pitti> lamont: I'll file a sync request
[07:08] <Amaranth> Riddell: Nope, I only know of him through LaserJock and one email
[07:08] <Riddell> crazy
[07:08] <LaserJock> Riddell: I'm putting willowng on a Kubuntu 6.06.1 project I'm working with with raphink
[07:09] <Amaranth> LaserJock: The KDE GUI will need to be patched to remove the entire first tab for that
[07:09] <dholbach> pitti, iwj: did you hear of http://www.g2zero.com/2006/09/examining_defects_in_the_firef.html already? :-)
[07:09] <Kamion> doko: I'm just telling you it won't be from me
[07:09] <_ion> keybuk: kdm probably calls halt to poweroff because halt has done exactly that forever, at least on Debian and Ubuntu. :-)
[07:09] <LaserJock> Amaranth: really? for the domain filtering, or is that a different tab
[07:09] <zyga> re
[07:10] <pitti> dholbach: ouch
[07:10] <Riddell> LaserJock: is this written using pykde3?
[07:10] <LaserJock> Riddell: it might be pyqt3 actually
[07:10] <LaserJock> I think
[07:11] <lamont> pitti: the debian version is 9.3.2-P1-1....
[07:11] <LaserJock> Riddell and Amaranth: rkd is the guy that wrote it
[07:11] <rkd> hey
[07:11] <lamont> pitti: so yeah, you can sync that
[07:12] <pitti> nice
[07:12] <zyga> bah
[07:12] <zyga> gaim deadlocks when I drag a buddy from one group to another
[07:12] <mjg59> Seveas: Hi
[07:12] <lamont> pitti: oh right.  9.3.2-2ubuntu1 was doko dropping the gcc-3.3 build-dep for powerpc.  that's incorporated in my upload
[07:13] <Amaranth> hey rkd
[07:13] <Seveas> mjg59, did you merge from my branch today?
[07:13] <lamont> pitti: 9.3.2-P1-1 also has a little bit of low hanging (and low risk) changes
[07:13] <rkd> Amaranth, was my cleaned-up willowng-config-kde code ok?
[07:13] <mjg59> Seveas: Earlier on, yes
[07:14] <Amaranth> haven't tried it yet but it looks nice
[07:14] <rkd> great
[07:14] <Seveas> mjg59, including the multi-variant themes and the fix for the rather silly timeout bug?
[07:14] <Amaranth> rkd: Any chance you could just load the .ui file instead of doing code generation?
[07:15] <mjg59> Yup
[07:15] <Riddell> hi rkd, thanks for doing the KDE williowng
[07:15] <Seveas> great -- then my usplash work is done for now 
[07:15] <Seveas> artwork people will probably poke me for theme creation though
[07:16] <sladen> Seveas: out of interest, do all of your lines end to Japanese unicode?
[07:16] <mjg59> Seveas: I wasn't quite clear on the use of the term "cropped"
[07:16] <rkd> Amaranth: it's possible, i guess, but i'm not experienced enough with QtDesigner to know if you can; currently after every UI change I have to manually add in a fair bit of init code to load modules, set up DBUS etc.
[07:16] <Seveas> mjg59, the term cropped is wrong
[07:16] <rkd> Riddell: no problem
[07:16] <Riddell> Amaranth: that's possible but there's a bug which means you can only do it from the same directory
[07:16] <Seveas> my english isn't always 100% -- it should be 'scaled'
[07:16] <_ion> sladen: Yes. 
[07:16] <Amaranth> Riddell: yuck
[07:17] <mjg59> Seveas: Ah, ok
[07:17] <jdong|laptop> god! the unicode! I can't take the unicode!
[07:17] <Seveas> 
[07:17] <Seveas> hi jdong|laptop !
[07:17] <Amaranth> 
[07:17] <jdong|laptop> lol
[07:18] <janimo> seb128, pitti: you know which irc chan is most likely to have gnome-cups commiters?
[07:19] <pitti> janimo: no, sorry
[07:19] <Amaranth> Seveas: That would almost look like a homestar runner thing if the font used for katakana had antialiasing
[07:19] <seb128> janimo: nobody is working on it, bugzilla
[07:19] <seb128> janimo: you can try #gnome-hackers but don't expect too much
[07:20] <janimo> seb128: their bugzilla has >60 open bugs so not a good idea I think.
[07:20] <janimo> I saw hub committed at least once 
[07:20] <seb128> 60 open bugs is nothing
[07:20] <janimo> seb128: nautilus is not a benchmark :)
[07:20] <LaserJock> heh
[07:21] <seb128> still, 60 bugs is nothing for an upstream product
[07:21] <seb128> nautilus has almost 1000
[07:21] <janimo> see what I said above re nautilus
[07:21] <zyga> lol :)
[07:21] <jdong|laptop> Seveas: do you think your freenx packages would work in edgy?
[07:21] <jdong|laptop> (yes, I am that lazy)
[07:21] <zyga> nautilus bug benchmark systedm 
[07:21] <Amaranth> 60 open bugs is small
[07:21] <seb128> janimo: http://bugzilla.gnome.org/reports/weekly-bug-summary.html
[07:22] <Seveas> jdong, only if you set bash as /bin/sh -- I need to fix freenx
[07:22] <seb128> janimo: look the top 15
[07:22] <Seveas> but NX is in a horrible condition
[07:22] <jdong|laptop> Seveas: k, does setting bash as /bin/sh break anything else?
[07:22] <jdong|laptop> and hell why isn't that default?
[07:22] <seb128> janimo: I was not saying that nautilus doesn't has lot of bug, just that 60 is really low
[07:22] <Seveas> because dash is a lot faster
[07:22] <jdong|laptop> ah, ok
[07:23] <seb128> janimo: anyway when there is no activate maintainer not easy to get somebody to commit
[07:23] <_ion> seveas: A quick solution would be modifying the scripts to have #!/bin/bash
[07:23] <Seveas> indeed
[07:23] <seb128> janimo: IRC will not make it better than bugzilla
[07:23] <_ion> Or "solution" at least.
[07:23] <Seveas> patches apprecitated
[07:23] <jdong|laptop> Seveas: is it ok just to symlink bash to sh, or is there a debianish way of doing it?
[07:23] <Seveas> or time
[07:23] <seb128> janimo: hub decided to commit his patch because nobody was replying, I'm not sure he's wanting to commit patches for other people too, that would be taking over the product
[07:23] <janimo> seb128: we should just branch it in LP and go with our own version
[07:24] <_ion> jdong: I'd recommend against changing the symlink, but patching the package instead. :-) Anyway, probably sudo dpkg-reconfigure dash
[07:24] <seb128> janimo: that's sort of what we do with adding all those patches to debian/patches directory, no? ;)
[07:24] <janimo> seb128: this means maintainers took the lock and went away. Not nice
[07:24] <seb128> janimo: nobody "took the lock"
[07:24] <janimo> seb128: yes but it's a lot more work
[07:24] <seb128> janimo: the lock is free if you want to be maintainer
[07:25] <janimo> seb128: really? hmm whero do I apply for that?
[07:25] <janimo> I am not joking
[07:25] <jdong|laptop> _ion: the thing is, a lot of other things are broken by dash, too.... ati, dvd-slideshow, etc etc etc
[07:25] <seb128> janimo: write the previous maintainer and say you are wanting to take over
[07:25] <jdong|laptop> _ion: my core duo can take some bash slowness :P
[07:25] <Kamion> jdong|laptop: it's not that many, and they're all bugs that were present before
[07:25] <Kamion> they should all be fixed
[07:25] <janimo> seb128: you see, when the previous mainatiner is MIA that is hard.
[07:25] <seb128> maybe show a bunch of patches so they can juge if you fit for the job ;)
[07:25] <Kamion> fixing bashisms isn't hard
[07:25] <jdong|laptop> Kamion: yes... fix them :)
[07:25] <jdong|laptop> lol
[07:26] <Kamion> ha ha ha plonk
[07:26] <seb128> janimo: he's not MIA, he just has willing to work on the product or is busy
[07:26] <Kamion> (hint: I've fixed lots)
[07:26] <seb128> s/willing/no willing
[07:26] <janimo> seb128: who is he? fejj? jody?
[07:26] <seb128> janimo: jody
[07:26] <_ion> * cmdline/apt-get.cc: - always show auto-removable packages and give a hint how to remove them
[07:26] <_ion> Rules! Thanks mvo. :-)
[07:26] <jdong|laptop> Kamion: fix them all. I expect my edgy to be rock solid :)
[07:27] <janimo> seb128: anyway the whole point of bzr is not not get locked upon such bottlenecks
[07:27] <seb128> janimo: feel free to branch from it
[07:27] <Kamion> jdong|laptop: then you're crazy ... edgy was never meant to be rock-solid
[07:27] <janimo> I'll see if I can make a bzr branch then
[07:27] <jdong|laptop> Kamion: I'm joking :)
[07:27] <seb128> janimo: then you have to convince people that they should use your brancj
[07:27] <seb128> branch
[07:27] <Kamion> yes, it's not that funny I'm afraid
[07:27] <tkamppeter> Hi,
[07:27] <Kamion> not given the hours we're all currently doing
[07:28] <tkamppeter> I have a problem with my Samsung X10plus laptop.
[07:28] <tkamppeter> I am trying to run the knot-2 live CD, so that I can start with my work here at Ubuntu
[07:29] <tkamppeter> It seems that there is a problem with the NVidia GeForce FX Go 5200
[07:30] <tkamppeter> When it tries to switch into the X display the screen stays simply blank.
[07:30] <jdong|laptop> tkamppeter: does pressing ALT+F7 help?
[07:30] <Mithrandir> tkamppeter: and the failsafe X option doesn't work?
[07:31] <tkamppeter> Thank you for the tips, I have currently booted my laptop with Mandriva to run this session.
[07:32] <tkamppeter> How do I activate the "failsafe X"? Is this the second entry in the main menu of the boot spalsh screen?
[07:32] <jdong|laptop> yes
[07:32] <tkamppeter> This second option I tried and it did not help. I did not try Alt+F7 though, when do I have to press Alt+F7
[07:33] <jdong|laptop> tkamppeter: after the system has "finished booting"
[07:33] <jdong|laptop> sometimes I find that the livecd finishes booting , but doesn't actually switch into X, rather sitting on an empty console
[07:34] <tkamppeter> There I Have tried Ctrl+Alt+F1 (F2, F3?) and then I saw a boot sequence in text mode scrolling, with a lot of "chdir: No such file or directory"
[07:35] <Kamion> that does rather sound like a broken live CD ...
[07:36] <jdong|laptop> tkamppeter: you might want to do a media check
[07:36] <jdong|laptop> also from the bootup menu
[07:36] <Kamion> have you tried the usual things like burning at a lower speed and cleaning the drive lens?
[07:36] <Kamion> since you're using knot-2, an image problem like that would have come up in testing - chdir failing doesn't sound graphics-card-specific
[07:37] <Mithrandir> it sounds like it failed to mount the cdrom
[07:37] <Mithrandir> casper's not always bright enough to see that's what happens and error out.
[07:38] <janimo> mvo: do I need the changes you uploaded in xubuntu-meta as well?
[07:39] <tkamppeter> I have successfully booted the CD on another PC (with ATI graphics card), there it worked, but there was no space for installing it.
[07:40] <_ion> mvo: "The following packages where automatically installed and are no longer required:"
[07:41] <Kamion> tkamppeter: that would suggest a dodgy drive to me, then
[07:41] <Kamion> sadly it's a very common problem
[07:42] <Kamion> we do suggest it to the user in various places, but there are a huge number of different possible failure modes, from the sublime to the ridiculous
[07:42] <Kamion> we need a try/except around the entire operating system, and we need to put the except bit in code that doesn't come from the CD ;-)
[07:43] <tkamppeter> I have now tried on a third PC with Intel i810. There it seems really to be the graphics card.
[07:43] <tkamppeter> Prompts on Ctrl+Alt+F1, F2, F3, but no X screen on F7.
[07:43] <ogra> Kamion, now thats a cool idea
[07:44] <Kamion> tkamppeter: try the VESA safe mode on that machine?
[07:44] <Kamion> might also be worth trying the alternate install CD, and see if it gets any further
[07:44] <tkamppeter> AFAIR Intel i810 has no VESA framebuffer, but I can try.
[07:44] <Kamion> sometimes squashfs seems to exacerbate transient CD read errors
[07:47] <tkamppeter> I am burning the alternate now. I hope it will work.
[07:50] <tkamppeter> While burning the alternate I am now booting the i810 PC with the VESA safe mode (second choice in boot splash menu).
[07:51] <tkamppeter> VESA safe mode does not help on the i810 box.
[07:52] <tkamppeter> What are the differences between the standard and the alternate boot CD?
[07:52] <Treenaks> doesn't work on my HP NW8240 either
[07:52] <Kamion> the standard one is one giant live (squashfs/unionfs) filesystem; the alternate one is a pool of .debs with a small installation image
[07:53] <Kamion> they're about as different as it's possible to get
[07:53] <Kamion> the alternate image is what you'd probably call the traditional Debian/Ubuntu installer
[07:53] <ogra> its the edubuntu installer :)
[07:53] <Treenaks> alternate is what was 'normal' before dapper
[07:53] <Treenaks> afaik :)
[07:53] <Kamion> yes
[07:53] <tkamppeter> So the alternate is the classical way to provide Linux. You boot a CD to directly install.
[07:53] <Kamion> right
[07:54] <jdong|laptop> ooh, freenx now does single-window sessions
[07:54] <jdong|laptop> *thud*
[07:54] <Kamion> the reason the desktop CD was brought to prominence was that many people wanted us to provide a live CD, and it saves us (Canonical) a considerable amount of money to be able to ship a live CD that can be installed as well
[07:54] <tkamppeter> This I will not be able to test on the i810 box, as my co-worker needs Mandriva on it.
[07:55] <tkamppeter> It is like Mandriva One, this is also a live CD which can be installed by an automated copy onto the hard disk. Mandriva 2007 will also be presented this way.
[08:00] <Kamion> Four hours to feature freeze. Better not need more than about 20 or so installs to get this right, then ...
[08:05] <tkamppeter> Will come back from another box, so that I can try again with my laptop.
[08:16] <Kamion> ubiquity 1.1.13. lucky for some
[08:27] <tkamppeter> Now my laptop has booted finally from the life CD. I have only chosen the safe VESA mode. Probably I did not wait long enough the first time. Thanks for all the tips.
[08:30] <geser> what are the chances to get an updated krb5 (main) into edgy?
[08:30] <geser> the current package doesn't contain the latest security fixes?
[08:30] <geser> s/?//
[08:30] <Mithrandir> geser: if so, quite big.
[08:32] <geser> should krb5 1.4.3-9 (current is 1.4.3-5) go in or the latest from debian unstable (1.4.4-1)?
[08:38] <mdz> doko: ooo 2.0.4 -> please send details via email as usual
[08:58] <jdong> ooh, the new startup sound is really jungle-ish :)
[08:58] <LarstiQ> jdong: forest or music genre wise?
[08:58] <jdong> forest :)
[09:02] <bluefoxicy> the /exec command in xchat is awesome.
[09:02] <bluefoxicy> damn thing keeps freezing and if there's no icon on the desktop for a terminal or no existing terminal you can't kill it.
[09:03] <jdong> anyone else get really blurry rendering of OOo?
[09:03] <jdong> my eyes hurt looking at the menus
[09:04] <Kamion> ubiquity 1.1.13> and the verdict is ... unlucky
[09:19] <tkamppeter> Finally, I have knot-2 up and running,
[09:19] <tkamppeter> but already a bug:
[09:20] <tkamppeter> Thunderbird is in the menu, but not on the system.
[09:20] <tkamppeter> Are the menu entries coming from each package, or is there one package with all menu entries?
[09:20] <ivoks> from packages
[09:20] <ivoks> that's odd
[09:21] <ivoks> tkamppeter: is there /usr/share/applications/*thunderbird*.desktop
[09:21] <ivoks> ?
[09:21] <Keybuk> mjg59: a moment of your time?
[09:21] <tkamppeter> I installed from the live CD
[09:21] <jdong> hmm, powernowd is not loading the ondemand governor
[09:21] <tkamppeter> Now I cannot check any more, as now my "apt-get install mozilla-thunderbird" has finished.
[09:22] <ivoks> heh
[09:24] <mjg59> Keybuk: Sure
[09:24] <mjg59> jdong: Why not?
[09:24] <jdong> mjg59: I'm trying to figure that out... only performance is loaded
[09:25] <jdong>  * Starting powernowd...                                                        /etc/init.d/powernowd: line 86: echo: write error: Invalid argument
[09:25] <jdong> as stated by that ^^
[09:25] <mjg59> jdong: What hardware?
[09:25] <jdong> intel core duo
[09:25] <jdong> speedstep_centrino
[09:25] <Keybuk> mjg59: clear > /dev/tty1
[09:25] <Keybuk> mjg59: why? :p
[09:25] <Keybuk> mjg59: (in the S98usplash script)
[09:26] <mjg59> Keybuk: Christ knows
[09:26] <mjg59> Nothing to do with me
[09:26] <Keybuk> that's why tty1 vanishes if you're running usplash
[09:26] <mjg59> jdong: Ok, that's a touch concerning
[09:26] <mjg59> jdong: It means that it's not getting the latency guarantees it wants
[09:27] <jdong> :( hmm
[09:27] <jdong> mjg59: is there any way to get that info for debugging purposes?
[09:27] <mjg59> Probably not trivially
[09:27] <Tonio_> sladen: ping ?
[09:27] <Keybuk> mjg59: mind if I make an upload to fix that?
[09:28] <Keybuk> mjg59: or do you have one planned?
[09:28] <mjg59> Keybuk: Go ahead
[09:28] <mjg59> Just check it in afterwards
[09:28] <Keybuk> yup
[09:29] <jdong> mjg59: cpufreq-detect.sh does identify speedstep-centrino as the MODULE.
[09:29] <jdong> hmm
[09:30] <jdong> mjg59: if I modprobe cpufreq-ondemand then start powernowd, it's happy
[09:30] <mjg59> Oh
[09:30] <mjg59> Hm.
[09:30] <mjg59> So.
[09:30] <mjg59> powernowd-early is supposed to load that automatically
[09:30] <jdong> so it would appear powernowd is failing to modprobe the right modules
[09:30] <jdong> mjg59: powernowd-early is a broken symlink here?
[09:30] <jdong> hmm
[09:30] <mjg59> That would probably be your problem, then
[09:30] <jdong> I don't have powernowd.early
[09:31] <jdong> what package is that in?
[09:31] <mjg59> powernowd
[09:31] <mjg59> If you've ever deleted it, it won't be recreated
[09:31] <jdong> hmm, I don't think I would've deleted it
[09:32] <jdong> alright, restored it...
[09:32] <jdong> thanks for your help, mjg59 
[09:33] <jdong> mjg59: now, one of these days you're gonna have to help me get those shiny graphs in g-p-m
[09:35] <mjg59> Keybuk: Ah - is it possibly so that the tty is clear when usplash flicks back there briefly?
[09:38] <Keybuk> mjg59: why does usplash flick back there?
[09:38] <mjg59> Keybuk: Because it needs to restore text mode somewhere
[09:38] <Keybuk> oh, for the console reset stuff
[09:39] <Keybuk> why can't it restore text mode on tty8 ?
[09:39] <mjg59> Actually, it's possible that it does
[09:39] <mjg59> Hm
[09:39] <Chipzz> Keybuk: it doesn't restore console 1 for me either; but when I press enter it does
[09:39] <Keybuk> this isn't "restore text mode" stuff
[09:39] <mjg59> Oh, maybe it was from back when we didn't run usplash on its own vt
[09:39] <Chipzz> it's just /etc/issue that's not being shown in my case
[09:39] <Keybuk> this is "when we stop usplash, run console-screen.sh"
[09:43] <Keybuk> mjg59: right, so this code does
[09:43] <Keybuk> - clear tty1
[09:43] <Keybuk> - quit usplash (which I assume chvt's to tty1)
[09:43] <Keybuk> - run console-screen.sh
[09:43] <Keybuk> - clear tty1 (again!)
[09:43] <Keybuk> - chvt 1
[09:43] <Keybuk> (the latter if the console is 8)
[09:44] <mjg59> Right
[09:44] <mjg59> I'm afraid I never wrote any of that code
[09:45] <_ion> The _only_ time the usplash splash is visible is during shutdown, from S20sendsigs i think. I wonder how to start debugging this?
[09:46] <Keybuk> mjg59: sending QUIT to usplash does chvt to 1, yes?
[09:53] <mjg59> Keybuk: To whatever vt usplash was started from
[09:54] <Keybuk> mjg59: tty1, as that's the only one that exists in initramfs
[10:07] <crimsun> iiii
[10:07] <crimsun> sorry, stuck keyboard
[10:07] <Seveas> jdub, ping
[10:12] <sladen> does a chvt 1 while on vt1 cause the wrong set of vt-switching signals to be sent?
[10:21] <trappist> do my comments on bug 59402 look right to you guys?  if so, is there somebody I should assign it to?
[10:21] <Ubug2> Malone bug 59402 in libxxf86vm "Cant't build against libxxf86vm-dev" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59402
[10:24] <Keybuk> sladen: how do you mean?
[10:25] <sladen> Keybuk: what is the effect of  chvt 1; chvt 1  and svgalib restoring the state of the console "on switch"
[10:26] <Keybuk> no idea
[10:26] <Keybuk> by the time the second chvt 1 is run, usplash isn't running
[10:26] <Keybuk> the issue is more that tty1 is cleared in the first place
[10:26] <Keybuk> clearing anything important that was on the screen before
[10:26] <Keybuk> including getty :p
[10:27] <sladen> Keybuk: so it's doing  clear; chvt 8;  rather than  chvt 8; clear
[10:28] <Keybuk> eh?
[10:29] <Keybuk> it's doing exactly what I said
[10:29] <Keybuk> it clears tty1
[10:29] <Keybuk> (on a related topic; I still don't understand why console-screen needs to actually, physically, change the vt ... why can't it just reset the vts while usplash is running - but hey-ho)
[10:30] <Keybuk> probably bugs/features in the vt layer
[10:30] <Keybuk> or a bad script
[10:31] <sladen> Keybuk: "it" clears?
[10:31] <Keybuk> S98usplash
[10:31] <Keybuk> has "clear > /dev/tty1"
[10:31] <Keybuk> which is why getty vanishes
[10:32] <sladen> Keybuk: I suggest you strike down on it with great vengeance
[10:33] <zyga> mvo: back
[10:34] <zyga> mvo: I know what is going on but I kind of don't understand why... the template expanding code explodes into infinity (no pun intended)
[10:36] <LaserJock> is it intentional to have update manager only use archive.u.c?
[10:36] <zyga> mvo: there is also a pretty bad problem of $(ls) treated like variable substitution
[10:37] <mvo> zyga: hm ...
[10:37] <mvo> LaserJock: what do you mean? 
[10:37] <zyga> mvo: I can reproduce the crash using one liner, just a sec
[10:38] <mvo> zyga: ok
[10:38] <LaserJock> mvo: sorry, that wasn't very clear. I'm using update-manager to upgrade to edgy
[10:38] <LaserJock> mvo: it appears as if it is replacing any mirror with archive.u.c in the sources.list
[10:39] <mvo> LaserJock: that would be a bug then, it really should not touch the mirror url
[10:39] <mvo> just rewrite the distribution section
[10:39] <LaserJock> it leaves the components
[10:39] <LaserJock> it just adds a new line with archive.u.c
[10:40] <zyga> (that one liner is a particularly long one *g*)
[10:42] <zyga> 1.9MB long line really hogs vim...
[10:42] <lemsx1> zyga: try vile
[10:43] <zyga> lemsx1: gosh, that's emacs?
[10:43] <zyga> well anything that works ...
[10:43] <lemsx1> zyga: nah
[10:43] <lemsx1> zyga: vi for big files
[10:43] <zyga> woooah!!!
[10:43] <zyga> it's just fast!
[10:43] <lemsx1> vile?
[10:44] <zyga> yeah
[10:44] <zyga> hmm, no insert mode/
[10:44] <Kamion> turn off syntax highlighting and vim should behave better
[10:44] <lemsx1> good to know. i never use it myself :-P
[10:44] <lemsx1> vile is just like vi
[10:44] <zyga> oh, true
[10:44] <lemsx1> you might want to use vim in read-only mode or something... like: view foo.txt
[10:45] <Kamion> that won't particularly help
[10:45] <Kamion> IME
[10:45] <lemsx1> Kamion: i wonder if there is an easy way to tell vim to turn off all extra stuff
[10:45] <Kamion> use vim-tiny and run it as vi ...
[10:46] <Kamion> or vi -C
[10:46] <Kamion> hmm, no
[10:46] <lemsx1> Kamion: ah, removing ~/.vimrc and launching vim as vi: ln -s /usr/bin/vim /usr/bin/vi ... or so
[10:46] <zyga> vim in read only mode still dies on that file
[10:46] <zyga> but vile is quite usable
[10:46] <Kamion> try 'view -u NONE'
[10:47] <lemsx1> zyga: vim should do it as well, but the highlighting, syntax detect, and completion really kills you at times
[10:47] <Kamion> behaves much better on big files IME
[10:47] <zyga> I kind of know what crashes in my code now, somehow I the list of iterable variables expanded and expanded into one *BIG* list of variables, I don't know how it did that on small postinst file
[10:47] <lemsx1> Kamion: in vi mode, yes. but in vim/gvim it's very slow
[10:48] <zyga> OTOH anyone who wrote mailman.postinst should DIE ;-)
[10:48] <Kamion> lemsx1: works fine for me in vim mode
[10:48] <Kamion> (I use vim, not vim-tiny, so I don't really *have* a vi mode)
[10:48] <lemsx1> zyga: lol
[10:49] <Kamion> -u NONE turns off all initialisations, including syntax highlighting etc.
[10:50] <lemsx1> Kamion: i have a spiffy .vimrc/.vim/.gvimrc setup... sometimes i hit one of those huge (gigabytes in size) files who really mess up vim
[10:50] <lemsx1> Kamion: any XML file over 20mb is a killer for me
[10:51] <lemsx1> somebody should write a Howto in howtoforge on the "perfect vim setup" :-P
[10:51] <zyga> lemsx1: nah, ubuntu should ship one :)
[10:52] <lemsx1> zyga: you got a point there... but, vim will read your own $HOME settings after the system-wide ones... 
[10:53] <zyga> lemsx1: so/
[10:53] <zyga> there is no .vimrc in skel
[10:53] <_ion> I'm sure ubuntu wouldn't ship my perfect vim setup, which e.g. contains "set guioptions=Lacir" to hide the menubar and the toolbar in gvim. :-)
[10:53] <lemsx1> zyga: it will screw up things
[10:53] <zyga> lemsx1: anyone who uses vim knows how to get past that ;-)
[10:53] <lemsx1> zyga: yeah, but i mean people who already customize their $HOMEs (or use NFS-mounted home's)
[10:53] <zyga> I'm for removing vim/emacs off the CD for that case :)
[10:54] <zyga> they are not user tools in any way
[10:54] <Kamion> emacs is not on the CD
[10:54] <jdong_> hmm, how does g-p-m determine whether or not to hide its power graphs? is that a hal thing?
[10:54] <lemsx1> zyga: no. that's not true
[10:54] <Kamion> only vim-tiny is on the CD, and I will vehemently oppose removing that
[10:54] <Kamion> the small amount of space it takes up is well worth avoiding the WTF factor for any long-time Unix user upon finding that there's no vi installed
[10:54] <lemsx1> zyga: you need vi in all CDs no matter what. and vim is the best vi implementation on the planet ;-)
[10:55] <zyga> lemsx1: and you need them for what?
[10:55] <jdong_>   system.formfactor = 'unknown'  (string)
[10:55] <jdong_> hmm :(
[10:55] <lemsx1> zyga: in UNIX all config files are plain text
[10:55] <zyga> lemsx1: just think - anyone can install vim, no user uses vim (where user = grandma user)
[10:55] <lemsx1> zyga: adding users is one
[10:55] <zyga> lemsx1: we already ship a good number of editors by default
[10:56] <zyga> lemsx1: I am only referring to the default ubuntu desktop install
[10:56] <lemsx1> zyga: i know newbies and what have you, are a totally different thing... vi takes like 1mb of space
[10:56] <zyga> OTOH ubuntu should be added to all dictionaries :P
[10:56] <jdong_> how do I force hal to believe I have a laptop??
[10:56] <Amaranth> zyga: Removing vi(m) is blasphemous
[10:56] <zyga> lemsx1: hmm, I are oyu sure?
[10:56] <zyga> Amaranth: not removing, moving off the precious CD to the internet archive
[10:56] <lemsx1> jdong_: install laptop-mode scripts ?
[10:57] <Amaranth> zyga: That's removing from the CD
[10:57] <Amaranth> from the default install, rather
[10:57] <zyga> lemsx1: apt-cache show vim | grep Size
[10:57] <zyga> hmm/
[10:57] <Amaranth> 700k
[10:57] <jdong_> lemsx1: does that actually work?
[10:57] <lemsx1> zyga: think networking and you will understand why vi can't be removed. transfering data on slow links is painful and remote management on somebody else's system is done usually on a console running some text editor that's lite
[10:58] <jdong_> lemsx1:   system.formfactor = 'unknown'  (string)
[10:58] <lemsx1> zyga: i was refering to just the app
[10:58] <lemsx1> zyga: $> du -s /usr/bin/vim.tiny 
[10:58] <lemsx1> 868K    /usr/bin/vim.tiny
[10:59] <Amaranth> vim-tiny is 470750 package size
[10:59] <lemsx1> jdong_: i use the laptop mode scripts. they work for me
[10:59] <zyga> lemsx1: you have nano for remote access, and if you have remote access you might as well just install vim-tiny :)
[10:59] <jdong_> lemonade: hal doesn't believe my laptop is a laptop, so g-p-m doesn't show the discharge graphs and such
[10:59] <Amaranth> so we're talking about 460k
[10:59] <jdong_> err lemsx1 
[10:59] <lemsx1> Amaranth: the package is compressed right. gzip'd
[10:59] <zyga> Amaranth: hmm, true
[11:00] <Amaranth> lemsx1: Size is the package size
[11:00] <lemsx1> zyga: and you then will remember to type nano? vi is EVERYWHERE. i'm a solaris admin as well
[11:00] <zyga> I still kind-of think that it's like build-essential, it should not be on a desktop CD
[11:00] <zyga> lemsx1: I'm not telling anyone to remove vi. altogether, I'm just suggesting it's not needed on the desktop CD
[11:00] <Amaranth> zyga: You risk pissing off/confusing a lot of people to save 460k
[11:01] <LaserJock> mvo: bug #59410
[11:01] <Ubugtu> Malone bug 59410 in update-manager "[dist-upgrader]  new a.u.c deb line in sources.list" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59410
[11:01] <lemsx1> Amaranth: package size after install? isn't that "installed size"
[11:01] <zyga> Amaranth: hmm, maybe but with command-not-found it's not that much of an issue really :D
[11:01] <mvo> LaserJock: thanks!
[11:01] <lemsx1> Amaranth: Installed-Size: 1392
[11:01] <Amaranth> oh yeah, installed size is what we want since it's the desktop cd
[11:01] <Amaranth> although even then it's squashfs compressed
[11:02] <Amaranth> lemsx1: vim-tiny is what you want to look at
[11:02] <lemsx1> zyga: the desktop CD is "live" right? and it also allows you to install the OS. so, if i use it to troubleshoot a system, you expect me to remember to type all known text editors until i hit one. or just use "vi" ?
[11:03] <lemsx1> zyga: there are other ways to save space on the CD... vi is 1.x mb...
[11:03] <trappist> I sure don't want to be the first *nix in history to not be able to find room for vi
[11:03] <zyga> lemsx1: to troubleshoot anything use the alternate CD IMHO that's more appropriate
[11:03] <zyga> trappist: again, I'm only for removing them from the default install - nobody is hurting vi :)
[11:03] <Amaranth> lemsx1: it's 940k
[11:04] <Amaranth> lemsx1: vim-tiny is what's installed right now
[11:04] <lemsx1> zyga: well. let's just say that you win. we won't change the distrib today anyway
[11:04] <Amaranth> you want to look at it
[11:04] <trappist> zyga: vi is probably the 1st thing I use on a default install, to edit sources.list
[11:04] <zyga> lemsx1: that's okay, there is a spec about that, might just as well happen for edgy+1 :)
[11:04] <Amaranth> 940k before it's compressed in the squashfs (or whatever is used)
[11:04] <lemsx1> trappist: me too, but not just that, /etc/network/interfaces and what have you
[11:04] <zyga> trappist: me too
[11:05] <zyga> trappist: but you can, on a default install type apt-get install vim-$flavor (and you probably do that by default too)
[11:05] <zyga> right?
[11:05] <lemsx1> zyga: i betcha that when vi gets removed from the desktop CD there will be somebody offering their own CDs "done right" which includes vi
[11:05] <trappist> I do
[11:05] <lemsx1> zyga: if you have networking, yes
[11:05] <zyga> lemsx1: I say, let them :)
[11:06] <zyga> I'm sure slashdot and digg will make it front page stories too ;)
[11:06] <lemsx1> zyga: if you are in an enterprise, you might not have that choice (firewalls)
[11:06] <trappist> but man, if there's one thing that's always been guaranteed to be on any linux/unix box, it's vi
[11:06] <zyga> lemsx1: if you don't you can use nano, it's there too :)
[11:06] <Amaranth> trappist: exactly
[11:06] <zyga> oh guys - but first off 
[11:06] <zyga> there is no remote access without ssh-server, we don't ship it installed by default
[11:07] <lemsx1> zyga: instead of Nano it should be evi (vim in easy mode)
[11:07] <Amaranth> i'm not talking about remote access
[11:07] <zyga> any server will have vim/vi on the cd and in default install
[11:07] <lemsx1> zyga: don't see the point on changing vim with anything else
[11:07] <zyga> IMHO one trivial text mode editor is enough, how many we put today? 5 or something?
[11:07] <Amaranth> i'm talking about 'vi' existing on every Unix/Linux in existance
[11:07] <LarstiQ> zyga: fsvo "use nano"
[11:07] <trappist> not all desktop users are gedit/kate people.  I don't want to alienate the people who have come to accept the availability of vi as a matter of faith.
[11:08] <zyga> Amaranth: that's true if you notice that most of those installations are not desktop systems now, think different as some company used to say
[11:08] <Amaranth> zyga: Does that company include 'vi' on their systems? :)
[11:08] <zyga> Amaranth: yeah, but they ship them on DVD's you know ;-)
[11:09] <zyga> :D
[11:10] <trappist> I've used *floppy* distros that saw fit to include at least ye olde vi
[11:10] <zyga> and besides, they do not have apt-get like system that would get you one with one line of shell
[11:10] <LarstiQ> if your sources.list is working
[11:10] <zyga> trappist: it's not about getting every last bit of space, think about purpose of vim on the desktop - I see none :)
[11:11] <zyga> LarstiQ: again, there are many editors in the default install that just work and are far easier to save and quit
[11:11] <LarstiQ> zyga: on a totally installed system? It's the only editor a lot of people use, including the desktop
[11:11] <trappist> zyga: see my previous comment - not all desktop users are consolephobes.  I think it would be a mistake to alienate the old-school linux-desktop console users
[11:11] <zyga> LarstiQ: stop thinking about people = ubuntu developers, you are wrong in that assumption
[11:11] <trappist> LarstiQ: +1
[11:12] <LarstiQ> zyga: I'm not even an ubuntu developer
[11:12] <trappist> zyga: people = windows users is also a bad assumption
[11:12] <pitti> having *any* vi-ish editor available is really a must
[11:12] <zyga> trappist: both are wrong, but statistics wins over my case
[11:12] <lemsx1> zyga: don't be a hater. if you don't want to use vi, use nano. but let vi exist in the CD for the rest of us
[11:13] <LarstiQ> pitti: agreed, I can live with pure vi
[11:13] <zyga> lemsx1: don't get me wrong - I love vim
[11:13] <pitti> but vim users are generally the kind of people who know how to install the fully version, I'd assume
[11:13] <zyga> guys guys
[11:13] <zyga> pitti: +1
[11:13] <zyga> 1) I love vim, i never use nano
[11:13] <trappist> zyga: 90% of people probably wouldn't notice if we dropped vim, but the 10% who would, would be seriously unimpressed, and potential contributors are kind of concentrated into that 10%
[11:13] <zyga> 2) space is not that important, true but it might be someday or when we want to make a sub 700MB CD
[11:13] <zyga> 3) there is *no reason* to ship half a dozen of text only editors for a desktop CD
[11:14] <lemsx1> zyga: then make people vote
[11:14] <LarstiQ> while I agree on the half dozen part, I see no connection between 'text only' and desktop
[11:14] <zyga> trappist: I agree, so we should do this in a way that they understand is correct
[11:14] <lemsx1> zyga: there is no way nano would win over vi
[11:14] <LarstiQ> or rather, reason not to
[11:15] <zyga> LarstiQ: the connection is: there is at least (not to mention another half a dozen probably) graphical text editors available by default that can be used instead, I'm not alientating console or anything
[11:15] <LarstiQ> zyga: are any of those vi?
[11:15] <zyga> lemsx1: nano is just an example: it could be vi but then let's drop nano
[11:16] <zyga> (nano is better for users that *need* to use the console but never used vi to edit their broken /etc/X11/xorg.conf 
[11:16] <lemsx1> zyga: now you are talking
[11:16] <zyga> but my claim that nano is better for that case still holds
[11:16] <lemsx1> zyga: there is easy vim for that
[11:16] <zyga> lemsx1: how is it called?
[11:16] <lemsx1> zyga: if symlink as evi, it just work like nano
[11:16] <zyga> oh :)
[11:17] <zyga> funny - I don't have evi
[11:17] <lemsx1> zyga: evim in the current ubuntu setup
[11:17] <zyga> evim :)
[11:17] <zyga> right
[11:17] <lemsx1> and you don't need to know anything about edit mode visual mode and whatever
[11:17] <zyga> looks decent :)
[11:17] <LarstiQ> funny, I don't know how to exit evi :)
[11:18] <zyga> hehehe
[11:18] <lemsx1> lol
[11:18] <zyga> me either :D
[11:18] <lemsx1> no ESC for you
[11:18] <zyga> fortunatly that was gevim :D
[11:18] <trappist> yeah, I think evi kinda defeats the purpose of keeping vi
[11:18] <zyga> so I clicked quit
[11:18] <zyga> boy that was odd
[11:18] <zyga> trappist: +1
[11:18] <LarstiQ> trappist: it's just argv[0] 
[11:18] <lemsx1> zyga: same here. it opens in gevim
[11:19] <zyga> I'm glad I have shown you my point of view :)
[11:19] <_ion> ^O:q<enter> seems to quit it. :-)
[11:19] <jdong> there. hal is happily forced into believing that I have a laptop. :)
[11:19] <LarstiQ> _ion: thanks!
[11:19] <trappist> I don't know if we have anything besides nano and vim on the cd, but I think we need nano for new users and vim for old schoolers, neither of whom we want to alienate.  I think they're both crucial.
[11:19] <_ion> (Every vim user should learn ^O)
[11:20] <lemsx1> _ion: ah, i didn't know about ^O
[11:20] <bluefoxicy> weird
[11:20] <bluefoxicy> when I rebooted the splash screen didn't print anything about what tasks were starting
[11:20] <zyga> _ion: I didn't knew that either
[11:20] <lemsx1> ttyt guys. nice chat!
[11:20] <zyga> (vim is an amazing editor, every day you learn something new)
[11:22] <bluefoxicy> also the crash reporter's dialog-to-the-face is starting to seriously piss me off.
[11:22] <_ion> Also :g and :norm
[11:22] <bluefoxicy> it bitches about every little thing, and then keeps saying whatever the first thing that ever crashed was has died
[11:23] <jdong> bluefoxicy: likewise; here it always thinks gaim has crashed for some reason
[11:23] <jdong> bluefoxicy: it tells me about it on every login
[11:23] <bluefoxicy> jdong:  I spewed a half a dozen specs based on pitti's crash reporter, one of them mentioned something about throwing dialogs in users' faces being a bad idea
[11:23] <bluefoxicy> I think I said the user would just click through
[11:23] <jdong> yeah, no kidding.... it's just plain annoying
[11:23] <bluefoxicy> I didn't predict that it would aggitate the hell out of me/anyone
[11:24] <jdong> like the windows XP thing
[11:24] <bluefoxicy> the XP one isn't agitating
[11:24] <bluefoxicy> amusingly enough it doesn't come back on every boot or every illicit program close and throw the same crap in your face.
[11:24] <jdong> well, that's a bug I suppose :)
[11:24] <_ion> s/systray/notification area/ ;-)
[11:24] <jdong> I hope
[11:24] <jdong> I pray
[11:25] <zyga> DAMN is the new artwork pretty! :)
[11:26] <jdong> new sounds are sweet, too
[11:26] <zyga> I broke my sound card :/
[11:26] <zyga> anyway
[11:26] <zyga> could we make the gdm background remain there untill the panel shows up?
[11:27] <bluefoxicy> jdong: https://wiki.ubuntu.com/AutomatedProblemReportsNotification in case you're interested
[11:27] <zyga> I'm not talking about fading the background or anything but just waiting for the panel to show up so that the whole process is more fluid
[11:27] <bluefoxicy> jdong:  not that I'd ever write a single line of code in that direction.
[11:27] <Trewas> btw will edgy have networkmanager by default, or what is used for managing network connections?
[11:30] <bluefoxicy> https://features.launchpad.net/distros/ubuntu/+spec/compressed-memory I'm hoping this is stable by Edgy+1 because I freaking love it (I used the 2.4 one back in the day)
[11:30] <bluefoxicy> anyway *does other stuff*
[11:31] <_ion> bluefoxicy: Hmm, a very interesting concept.
[11:31] <bluefoxicy> _ion:  memory compression?  It's old, but good.
[11:49] <Keybuk> tsk
[11:49] <Keybuk> what's wrong with this?
[11:49] <Keybuk> pid = fork ();
[11:49] <Keybuk> if (fork > 0)
[11:49] <Keybuk>     exit (0);
[11:49] <bluefoxicy> Keybuk:  fork?  what the hell?
[11:50] <bluefoxicy> Keybuk:  s/if (fork/if (pid/
[11:50] <Keybuk> yes :p
[11:50] <thom> heh
[11:50] <bluefoxicy> (hint:  you are checking if fork()'s entry point is at an address > NULL)
[11:51] <thom> maybe testing the correct variables would help? :-P
[11:51] <bluefoxicy> Keybuk:  coding while drunk/tired/surrounded by real life cute guys that actually aren't straight/etc?
[11:51] <Keybuk> coding until a deadline
[11:51] <Keybuk> sadly David's buggered off back to work
[11:51] <bluefoxicy> ah, coding under pressure.
[11:52] <_ion> Something upstart-related, btw?
[11:52] <Keybuk> _ion: nah, just testing why usplash fucks the console
[11:52] <bluefoxicy> Keybuk: perhaps because its condoms are reaching their expiration date?
[11:52] <cbx33> hey seb128 
[11:53] <seb128> hi
[11:53] <cbx33> did you get a chance to see my pessulus patch?
[11:53] <Keybuk> bluefoxicy: heh, never use them myself ... and soooo off-topic :p
[11:53] <bluefoxicy> you missed the joke I see :P
[11:53] <Keybuk> yes, I did
[11:53] <seb128> cbx33: not yet
[11:54] <sladen> Keybuk: bugger off as in ->USA ?
[11:54] <cbx33> seb128, think it'll make FF?
[11:54] <sladen> Keybuk: buggered
[11:54] <Keybuk> sladen: hmm?
[11:54] <bluefoxicy> Keybuk: the latex eventually breaks down and they expire; hence occasionally the issue of ... finding a way to use the rest of the box up quickly ... comes up :P
[11:54] <seb128> cbx33: if people stopping pinging me on IRC every 5s maybe :p
[11:54] <cbx33> hahah
[11:54] <sladen> Keybuk: "disappeared off to work" == "flown to the US"?
[11:54] <cbx33> sorry
[11:54] <seb128> np
[11:54] <bluefoxicy> anyway
[11:54] <seb128> I'll have a look on it next
[11:54] <cbx33> ty
[11:54] <Keybuk> sladen: no, just back home to the Cotswolds ?
[11:55] <seb128> but I've been very busy this week
[11:55] <cbx33> I know
[11:55] <seb128> with GNOME 2.16, feature freeze, etc
[11:55] <cbx33> which is why I havn't buged you
[11:55] <cbx33> this is just a curious...do you think it'll get in
[11:55] <cbx33> not a pressue push :p
[11:55] <cbx33> I don;t do those
[11:55] <cbx33> ;)
[11:55] <seb128> ok, so I think there will be no issue if the patch is correct ;)
[11:55] <seb128> :)
[11:55] <cbx33> ;)
[11:58] <robertj_> should we add Trashes to the places side-bar in Nautilus?
[11:58] <robertj_> (or I guess just Trash:/// since it encompases all .Trash entries now)
[11:59] <seb128> robertj_: no
[11:59] <seb128> we already have it on the panel
[11:59] <seb128> and it's easy to delete from the menu or the keyboard
[11:59] <seb128> no need to clutter places too
[12:00] <robertj_> seb128: I was viewing the applet as the clutter :)
[12:00] <robertj_> because Nautilus is only on-screen while I'm managing files
[12:00] <Kamion> only way I can see to do it is to hope that debian-installer/keymap is there, I think
[12:00] <seb128> robertj_: nautilus manages your desktop too
[12:01] <seb128> robertj_: and some people use a different sidebar or the spatial mode
[12:01] <robertj_> seb128: spatial is more problematic though
[12:02] <seb128> ?
[12:02] <robertj_> seb128: removing trash applet for spatial users
[12:02] <seb128> robertj_: why it's not for people using a different sidebar to browser mode?
[12:02] <robertj_> seb128: because then puting it in the sidebar is an obvious solution
[12:03] <seb128> places sidebar
[12:03] <robertj_> well indeed, I suppose it should be in Tree as well..but still
[12:03] <seb128> what is your issue with the very few pixel it takes on the panel?
[12:04] <robertj_> seb128: im a one-panel heretic
[12:04] <robertj_> seb128: recent convert though, within 7 days every screen I had in the house went from 4x3 to 16x9
[12:04] <robertj_> and then I decided I might should fiddle with my layout a bit
[12:04] <seb128> adding a bookmark to trash:/// is easy enough if you customize your whole desktop anyway