[00:08] sladen: they are not allowed. use .tar.gz [00:14] cjwatson: this is for the .deb, not the .orig Using dh_builddeb -- -Z lzma is saving ~50% [00:15] cjwatson: or is it just gz and bzip2 allowed for .debs in the archive [00:15] sladen: I'm pretty sure mesa is already lzma compressed, which would suggest that it's OK. Check it out? [00:16] sladen: oh, well then those have been allowed for ages with a suitable Pre-Depends [00:17] namely dpkg (>= 1.14.12ubuntu3) [00:33] cjwatson: pre-depends not needed anymore as of lp 2.2.3 [00:34] ok [00:35] sladen: you should be able to just upload lzma compressed debs without doing anything else special now [00:35] sladen: up until ~ april 1 you had to have a pre-depends [00:36] lzma compressed debs? as in data.tar.gz replaced with data.tar.lzma or something? [00:36] yes. [00:36] it makes sense for some particularly large packages [00:36] hyperair: its been supported since gutsy (iirc) [00:37] i see. [00:37] so how come it never really took off? [00:37] or maybe it was hardy, i can't recall [00:37] hyperair: big things use it already such as OOo [00:37] i see [00:37] cjwatson: ta. (And I'm a great fan of Pre-Dpedns for funky compression methods, so I'll add it ;-) [00:37] but how do you trigger it building as lzma instead of gz? [00:37] i was discussing with Keybuk about potentially making it the default for dpkg for karmic, but not sure that it is that helpful without also having lzma for squashfs [00:38] dh_builddeb -- -Zlzma [00:38] calc: I don't think that's suitable [00:38] aah [00:38] I would advise against such a change [00:38] cjwatson: too slow for desktop cd or something else? [00:38] aiui novell already does it by default for rpm (not saying that it is a reason to do it though) [00:39] I'm very concerned about the behaviour on low-memory systems [00:39] and am strongly inclined to veto this [00:39] ok [00:39] how low do we support as far as low-mem is concerned? [00:40] for installation, at least down as far as 64MB; lower is manageable but it gets slow due to localedef [00:40] iirc it uses ~ 30MB to decompress, so it is valid if we support very low-mem situations i suppose [00:40] wow i didn't realize we supported that low [00:41] obviously you can't run a standard Ubuntu desktop on that (hence why I'm not worried about OOo using data.tar.lzma) but there are other valid use cases using the Ubuntu archive that aren't a full desktop [00:41] yea [00:41] I'm pretty sure I raised the same concern a while back [00:42] I do think lzma has been great for getting us some much-needed space on the alternate CD in particular targeted areas, and I'm very grateful for that [00:42] i think in the past when i had talked to you it was more an issue of low-mem ~ 256MB machines installing via the desktop cd, or something to that effect [00:42] but I think we should leave it at that rather than pushing it too far [00:42] yea thats fine [00:42] memory is an issue on desktop CD installs too, certainly [00:43] we've found it difficult to hold it to 256MB [00:43] luckily we have the install without booting into desktop option as well [00:43] which i imagine takes a lot less than 256MB [00:43] the desktop CD is a situation where noticeably more memory can be used during installation than is required for the installed system, which I regard as a fairly major problem [00:43] it's not that much less, unfortunately [00:44] I think it's still clear, but not comfortably so [00:44] cjwatson: someone had requested an option for a gui netinst, would that lower the requirements without having to fall all the way back to ncurses? [00:44] yes, to some extent [00:44] cjwatson: from what i can see debian has that and it only takes ~ 15MB for a netinst cd with gui [00:45] we will be doing that once a current version of GTK gets fixed to be able to support the directfb backend again [00:45] (Debian unstable has run into the exact same problem; it isn't just us) [00:45] /c/c [00:45] of course it takes more than that in ram to run it, but probably less than the current desktop cd direct install method [00:45] cjwatson: ah i see :) [00:46] it's not a replacement for ubiquity; it's (largely) a straight translation of the current d-i dialogs into GTK [00:46] cjwatson: Another ubuntu-autotest run is in progress, BTW (just kicked off), if you want to watch the list. [00:46] which is serviceable, but it isn't a by-hand designed interface. I expect we'll need to put quite a bit of work into some areas [00:46] cjwatson: Seeing as how the attempt to do the same with soyuz seemed to go unpleasantly. [00:47] calc: (I was one of the developers who put a lot of time into the GTK installer in Debian, so I'm reasonably familiar with it ;-) ) [00:47] cjwatson: isn't the alternate cd d-i also? [00:48] though I didn't do the final stretch, just a good deal of the early underpinnings - there was a time when I was about the only person working on it [00:48] calc: yes [00:48] infinity: ok [00:48] cjwatson: so its different than the current desktop installer but a nicer looking version of what we currently use in the alternate cd [00:49] calc: it's just dropping in a different debconf frontend, basically [00:49] hmm we'd have to put that on the alternate cd as a gui installer then, since the desktop cd isn't done in a manner that d-i knows how to install from, aiui? [00:49] that's correct. More realistically, the server CD [00:49] ok [00:50] (since the server CD has quite a bit of spare space but the alternate is usually kind of tight) [00:51] ok [00:51] or just lzma a package or two and then we have space on the alternate cd again ;-) [00:51] also, the corporate services guys are actually pushing for this [00:52] * calc has to run buy groceries but will be back ~ 1hr [00:52] Some of the other UbuntuStudio devs have also expressed an interest in using the GTK d-i for the studio disks. [00:53] right [00:53] I did try to do it for jaunty but the aforementioned GTK problems shot it down [00:53] Yep [00:54] A11y wise, I don't care about GTK d-i, since the Ubuntu desktop CD's installer is accessible enough to be usable now. [01:01] I made a modification in the kernel source code. Now I wan to test this modification but before I need to compile the kernel [01:02] Can someone refer me so some documentation ? [01:03] cjwatson: I'm a believer in sticking to .gz for base packages. But what's your view on alternate compressors ("advancecomp" et al) that can get another 15% out of the deflate stream by trying harder [01:05] In fact, I thikn I only need to compile one part of the kernel : hid [01:07] sladen: IME we are at the point where other strategies for saving space, such as working on more efficient representations and shipping of translated documentation, represent much better bang for our buck [01:10] cjwatson: I say we just remove libx11 and anything that depends on it, and see how that works out. [01:10] infinity: wow, that could save me a lot of work [01:11] bryce: See? I'm a giver. [01:17] kees: do you know what happened to the sudo patch I'm sure we used to have that passed through all environment variables if sudoers says you can run whatever command you like? [01:22] kees: are we supposed to just use sudo -E now? (This comes up because libgksu never passes sudo -E.) === nhandler_ is now known as nhandler [01:50] I'm working on remastering a LiveCD, and I have a CD that works with i386 but when I generate the amd64 version it fails right as the graphical boot menu is displayed. Any ideas why this happens? [01:50] Or who might know? ;-) [01:54] slangasek: hmm.. I thought pitti fixed that already... will check. [01:57] ArneGoetje: ah; the new ones still show up in component-mismatches [02:27] hi [02:29] i am reading in debian-policy website that the latest version is 3.8.1.0.... but what should i use in my debian/control file 3.8.1.0 or 3.8.1 ?? [02:30] javito: you should read the part of the policy manual that covers the Standards-Version field [02:30] javito: section 5.6.11 [02:32] cjwatson: thanks [02:32] is very clear here [02:33] Heh, the Debian Policy manual is so complete that the answer to the question "I'm reading the manual and don't understand, what do I do?" is "Read the manual" :) [02:35] it's comprehensive enough that it's understandable for people to get lost until they've made enough use of it to have a general feel for where to find things [02:36] iirc one of the first things you are supposed to do before applying as a DD is to read the entire policy manual :) [02:37] and be able to recite the dfsg backwards ;-) [02:37] * tedg tried to be in AA but couldn't remember all 12 steps ;) [03:10] If "Sebastien Bacher" is present: Can you take a quick peek at ? It's a small bug with a simple fix, but would be great to get into one of the -ubuntuX patchsets for Nautilus before Jaunty final. Apologies in advance for the drive-by patch/question, but my connection time is limited. [03:10] Ubuntu bug 357438 in nautilus "Always show "backup" directories." [Undecided,New] [03:11] If you see this, thanks! [03:11] quick drive-by :) [03:12] when will be allow upload bugfix to karmic? [03:13] well, you can upload bugfixes to jaunty [03:13] including minor bugs? [03:13] but new versions, and new packages, and new features need to wait until karmic opens (~6 weeks) [03:14] javito: if it helps the quality of the release, yes [03:14] but a bugfix impleis increase the version [03:15] javito: increasing the -0ubuntuX package version, yes [03:15] ok then thanks [03:15] javito: to get a new upstream release in, it has to be a bugfix only release [03:16] sorry for my stupid questions i'm really newbie here :D [03:16] javito: yeah, if nobody was allowed to fix bugs and nobody allowed to upload, we might as well have released a month ago... :) [03:17] sladen: maybe only for critical bugs, like in stable debian [03:17] since frozen [03:17] it just why i asked [03:18] javito: post-release, you need a stable-release-update exception (eg. a security issue) [03:18] cjwatson: I think pitti removed most of it, but kept a whitelist for things like http_proxy, but beyond that I don't remember. [03:19] cjwatson: though, recently, I was pretty sure that environment was passed unfiltered if one was in sudoers with an "ALL". [03:25] hrm. How does one not use these new notifications for just pidgin... [03:25] Hobbsee: irssi :) [03:26] i do like it for things that are actual "doesn't matter if you miss them" notifications, but for pidgin, it means that i miss the notifications, and there's absolutely *no* indication that someone's sent me a reply to a message. And i'm not convinced that the notifications actually fire every time, at all. Fail. [03:26] sladen: i guess I could use bitlbee and konversation, yes. [03:26] sladen: irssi, if i use the notifier, is goign to have exactly the same problem, on a bigger scale? [03:27] Hobbsee: there was discussion about irssi and notifications here yesterday, I can't remember the result of that conversation however. [03:27] Hobbsee: somebody (look for the logs) was discussing something similar earlier. Ah, TheMuso has beaten me to it [03:28] s/look/lookings/ [03:28] TheMuso: i get the feeling that the more i send thru the notifier, the more messages i'll only see hours later when i happen to manually open up the window again, though [03:28] Hobbsee: I don't know how the notification works, since I don't use it. [03:28] TheMuso: ah [03:29] sladen: i'll do that, but i suspect that'd make my problem worse again [03:29] Hobbsee: IIRC, libnotify is "replaced" by the new back end so everything is gets diverted to /dev/null (well, a 3 second pop-up, then /dev/null) [03:30] sladen: damn. that's *exactly* what i'm trying to avoid. [03:30] sladen: and i'm really not convinced that the popups work all the time, either [03:31] Hobbsee: probably what will happen is that the pidgin authors will get pissed off, and switch to their own internal notification system, skipping the "standard" as it doesn't do what they want [03:31] sladen: heh. That sounds about right. [03:31] hrm, I guess i could switch to kopete or something [03:32] which should bypass the new notifier stuff [03:33] you can do other mad setups, like irssi-proxy (irssi, but it also listens on a port and passes everything through so that you can daisy chain to another IRC client) [03:33] that's true. I'm already doing funky things with proxies. [03:33] but that's not a bad idea [03:34] if daisy chaining worked /really well/ you could have a IRC "client" that did nothing but display flashing notifications [03:34] I don't really have a solution (as it's not a use-case I've hit and had to work around yet) [03:35] the irc isn't really the problem - konversation doesn't (and probably won't) use the new notifications for this release. It's only the other protocols. [03:35] perhaps the long term key is to modify the funky new libnotify backend, to be ware of state [03:36] mmm [03:36] eg. is the screen-saver running, is there keyboard/mouse activity, is there a full screen movie [03:37] what's that four-quardrant system for prioritising? There's two axis, one is urgency and the other is importance [03:38] so it could be less urgent than watching the full screen movie, but important enough to save until later [03:38] true [03:39] Hobbsee: anyway, we probably know who to blame for the new design if it gets panned post-release ;-) [03:39] sladen: hmmm? [03:45] Hobbsee: the same person who brought us nekkid desktop wallpaper! [03:46] I suspect the new notifier will end up covering 90% of use-cases better than it did before, but leave an empty space for the other 10% [03:46] ah, right [03:46] indeed. [03:47] although I personally think "give me some obvious notification that i've received a message over msn / jabber / etc via pidgin, after the fact" would be a fairly large use case for pidgin... [03:48] Ah well. I'm sure the notification team will decide how to do it best, and if they don't, debian will get a lot more people. [03:49] * Hobbsee goes off to research how to make pidgin behave more sensible [03:49] er, sensibly [03:49] Hobbsee: the messaging-indicator is where the missed notifications get queued up [03:49] jcastro: except that it doesn't work. [03:49] jcastro: they never show up there [03:49] that's a bug then [03:50] sorry, the ones about the fact that someone's signed in do, but not the ones that say i've received a message. [03:50] only the messages are supposed to queue up in there [03:50] not like, when people sign in and stuff [03:50] and speaking of that, i'm not so crash hot on having to have *two* icons on my panel about it, or having to keep pidgin open, because otherwise pidgin quits. [03:51] if only it worked...When's the ETA on getting a fix? [03:51] there should just be one icon, the little envelope for the messaging indicator [03:51] dunno, mine works, maybe ask in #dx? [03:51] hmmm. might have to ask later when they're awake [03:52] jcastro: so unread IM/email messages are hiding until the envelope, until you click it, and then they /are/ being shown? [03:53] right [03:53] sladen: that would make sense. [03:53] when there are new ones in there there's a green dot on the envelope [03:53] so basically, instead of a pidgin icon, a new evo mail icon, etc. you just have them all under one icon [03:54] which also btw brings the contact list up for pidgin [03:54] I thought empathy got ported as well [03:54] right [03:55] I never thought notifications would be the reason Ubuntu started maintaining a large number of deltas from upstream [03:55] alot are accepting them [03:55] though I don't have statistics on that [03:56] Sure, if it's just "cleanup your usage to check for features" and stuff like that [03:56] or "use append mode" [03:59] yeah, it missed the first /msg but saw the second and third [03:59] and gives me an exploding envelope [04:02] well, that'd be pidgin crashing [04:07] Hmm, the option is called "Shutdown..." but the action is immediate (so should not have the ...) [04:37] is there a reason that arrow keys don't work in insert mode in vim, i seem to recall this used to work at some point [04:37] but not i just inserts a new line with eg "B: [04:37] er "B" [04:44] jcastro: hrm. It appears that my existing config was making it not behave properly. Has the dx team tested with previous release configs, instead of clean slates? [04:44] Hobbsee: not sure, that is a good point though [04:44] ah, but the minimise to tray functionality isn't there. dammit. [04:45] or the minimise to anything, really [04:45] jcastro: did you find a way around that? [04:48] at least i get the right things in the envelope now! [04:57] calc: wfm. my fading memory suggests that tends to be caused by a zotted or simply wrong TERM - eg should be xtermc, is set to sun-console :-/ I think terminals can also get into a gaga state as well, and need resetting. [04:58] ah, that would be https://bugs.edge.launchpad.net/ubuntu/+source/pidgin/+bug/295713 [04:58] Ubuntu bug 295713 in pidgin "pidgin ends (instead of minimizing) when closing buddy list" [Undecided,Incomplete] [04:58] hrm. can one assign things to the dx team, i wonder.... [05:07] sladen: out of curiousity, which gnome theme are you using? [05:08] spm: hmm gnome-terminal sets it to xterm, if i change it to linux it seems to work [05:09] actually no i just tested in the wrong spot [05:10] it seems the place it is happening is set to TERM="screen" [05:10] for screen console program [05:13] calc: Under screen, it's "TERM=screen" for you? [05:13] calc: I get "TERM=screen.linux"... [05:30] infinity: yea [05:30] infinity: shows up as just TERM=screen for me [05:46] Any chance someone can commit the .debdiff attached to bug #311732? [05:46] Launchpad bug 311732 in linux "2.6.28-4 breaks thinkfinger" [Undecided,Fix released] https://launchpad.net/bugs/311732 [05:46] Thinkfinger is completely broken atm [05:57] Hobbsee: *boggle*. Human, I guess. I try and stick with the defaults (assuming the default theme hasn't changed) [05:57] sladen: hm, OK [05:58] Hobbsee: I did have some problem recently whereby all the fonts were too big, until you went System->Settings->Appearance. Didn't have to click anything but that was enough to fix the fonts to the correct size [05:59] * Hobbsee nods [05:59] gnome-settings-daemon not being started properly on login? [06:05] Keybuk, do you think you could have a look at the patch? === nysosym is now known as nysosym|away [08:06] sladen: I have noticed this immediate shutdown thing too - *intermittently* ! === nysosym|away is now known as nysosym === jcm is now known as Guest98047 [08:29] hi [08:30] i tried to create a deb package and it worked, but my program doesn't have any files in /bin for example and it created and packaged empty directories, how can i remove thatN [08:31] fix your package's .install file? [08:31] also, packages don't go into /bin they go into /usr/bin [08:32] .install file? [08:33] dade [08:33] i do not have any .install file [08:34] he probably meant debian/install file [08:34] emacsen-install? === azeem_ is now known as azeem [08:35] htrejh, in your debian/ folder in your source package, the install file (or pkgname.install for multi-package source packages) lists the actual files to include in the package [08:35] strange i do not have this file [08:35] you have to create it [08:35] k [08:35] and you have no files either. go figure [08:35] are there any sample files? [08:37] which packaging helper does your package use? [08:38] dh_make [08:38] but i found, there were plenty ex files generated :) [08:39] so plain dh. [08:39] which compat version? [08:40] 7 [08:41] are you using minimal-style rules? [08:42] hm [08:42] dunno [08:42] does your rules file contain a bunch of dh_this dh_that dh_other, or a couple of "dh $@"? [08:43] http://svn.debian.org/viewsvn/pkg-cli-libs/packages/gtk-sharp2/trunk/debian/rules?revision=3937&view=markup is old-style full rules, http://svn.debian.org/viewsvn/pkg-cli-libs/packages/smartirc4net/trunk/debian/rules?revision=4049&view=markup is dh7 minimal rules [08:43] a bunch of... [08:43] it's the first one [08:44] right. is dh_install called in your binary(arch|indep) rule? [08:46] hm no, it is commented [08:48] dh_install is the rule that actually, well, installs files generated by the build into a deb. it uses the debian/install file as a manifest [08:48] http://svn.debian.org/viewsvn/pkg-cli-libs/packages/smartirc4net/trunk/debian/libsmartirc4net0.4-cil.install?revision=3806&view=markup is a little example [08:49] euhm [08:50] so i must uncomment it right? [08:50] right. [08:50] ok, thanks [08:55] but i still have the empty dirs :s [09:01] hi, how could I change the mysql user ID correctly ? [09:01] actually I change it manually in /etc/passwd and /etc/group and now the server refuse to start [09:02] lyhana8: #ubuntu for support please. [09:03] Hobbsee: I come from there, they have no idea. Plus it's a high config issue [09:05] I'm trying to share a mysql DB among a gentoo and an ubunutu. User ID change and file permissions change work fine on gentoo, but not on ubuntu [09:05] lyhana8: no it isn't. and that doesn't make this channel a good forum for support. you could try #ubuntu-server or any of the local community channels please. [09:05] -server's not a support channel either for non-server stuff, actually. [09:05] lyhana8: a simple google search finds http://ubuntuforums.org/showthread.php?t=109752 which appears to be appropriate. [09:05] But please note, this really isn't the appropriate channel. [09:05] Hobbsee: well, I found running mysqld as a user a bit strange ;-) [09:06] Hobbsee: thanks [09:06] Nafallo: it does? [09:06] Hobbsee: I meant. mysqld on a non-server. sorry :-) [09:06] oh, right [09:06] Hobbsee: but yea. you're right. I was jumping to conclusions :-) [09:14] can someone explain why i get those empty dirs packaged? [09:24] htrejh: do you have a dirs file in your debian/ folder [09:24] yes, i just removed it, will try again :) [09:25] htrejh: that was most likely your problem. I ran into this once or twice myself :) [09:25] good luck! [09:25] ok thanks :) [09:42] persia: are you still working on bug #349458? [09:42] Launchpad bug 349458 in ffmpeg-debian "Please provide a VFP-optimised build for ffmpeg-debian" [Medium,In progress] https://launchpad.net/bugs/349458 [09:49] Riddell: is the disappearing dbus/mainloop/qt.so still a issue with current intrepid->jaunty upgrades? I just ran anohter test upgrade and its still there, I suspect its a particular package that triggers it (and that I do not have installed) === dpm_ is now known as dpm [10:21] damn :-/ [10:21] my X just crashed after a VT switch again [10:23] nafallo: which chipset? any stacktrace at the end of xorg.log.old? [10:23] /var/log/gdm/\:0.log.1 [10:23] will put it on a pastebin. [10:24] once my firefox stop crashing... [10:26] http://paste.ubuntu.com/146805/ <-- 00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 0c) [10:27] Nafallo: for firefox crashes, try to create a new user - this will definitely sort it if it's a local /home/ problem in the configuration :) [10:28] savvas: yeah. I know. I reckon it's one of the newly updated plugins. just hard to catch. [10:28] also, a new user won't have my stored passwords, so a no go. [10:29] there's a plugin for that [10:29] password exporter [10:30] * Nafallo shrugs. [10:30] still not the issue :-) [10:30] :P [10:30] I'm pretty sure it's a plugin considering it started happening right after I upgraded vimperator ;-) [10:30] safe mode [10:31] well, if you ever consider on recreating a new .mozilla directory: https://addons.mozilla.org/en-US/firefox/addon/2848 [10:31] anyway. my bigger issue is X ;-) === asac_ is now known as asac [10:57] mvo,seb128: does the libgksu patch I attached to bug 357101 look OK to you? [10:57] Launchpad bug 357101 in ubiquity "GUI does not exit on reboot after install on armel" [High,Triaged] https://launchpad.net/bugs/357101 [11:23] is anyone going through sync requests anymore? [11:27] tjaalton: yes, archive admins [11:34] seb128: ok, good [11:40] evand: it seems we need to work on the time zone map some more...two problems: 1) the pics are scaled up during a "real" install and 2) the text is black and the background too dark for that [11:40] kwwii: by being scaled up do you mean that they're scaled to be greater than their original resolution? If so I believe cjwatson fixed that. [11:41] and indeed on the text. Did you have a better color in mind? It's quite easy to change. [11:41] evand1: yes, exactly...the pics show up blurry because of it [11:41] evand1: yeah, I'm talking to others about it and I'll get back to you [11:41] evand1: thanks [11:42] kwwii: sure thing === evand1 is now known as evand [11:46] evand: I didn't do anything about that [11:47] evand,kwwii: the change I made (relatively recently) was to prevent the images being unnecessarily scaled down from their original resolution when the screen is big enough to comfortably accommodate the full sizes [11:47] the map is fiddly to use when scaled down; that's acceptable when we have to do it to make it fit on the screen, but IMO we should make use of bigger screen sizes when we have them [11:47] oh, my mistake. I misread the changelog entry then. [11:48] cjwatson, evand: the problem I am seeing is that the pics are being scaled up larger than the original size and therefore have blurry edges [11:48] kwwii: I'll take care of that [11:48] cool :-) [11:48] that's very odd. what size of screen? [11:48] evand: is this something you can reproduce? [11:49] cjwatson: 1920x1440 or such [11:49] the code just says "if screen big enough: use normal images; else: use half-size images [11:49] " [11:49] cjwatson: it seems to be normal sized when I install from the live CD but when I click on install ubuntu directly it scales them up it seems [11:49] aha [11:49] that'll be the full-screening [11:49] I bet that the map widget is set to expand [11:49] or fill, or some combination of those [11:49] cjwatson: ah actually, probably not given my low resolution, but I can roll debs for kwwii to test [11:49] right, that is how it looks [11:50] ah [11:50] (I can never remember the difference and always have to look it up ...) [11:51] cjwatson: Could we get a debian-installer upload (I can do one if you like) today? I'd like to test updated hd-media kernel built by d-i for https://bugs.launchpad.net/ubuntu/+bug/354226 [11:51] Ubuntu bug 354226 in ubuntu "Jaunty lpia installer fails with hd-media kernel on dell mini 9" [Undecided,Confirmed] [11:52] if a kernel config is missing on lpia, we might still have a change to fix that one [11:52] lool: ok, I can do one later; I'd like to get a bit further with diagnosing bug 341928 first though, if that's OK, since that may well require a d-i upload anyway [11:52] Launchpad bug 341928 in lvm2 "Consistently reproducible "device or resource busy" error on partitioning" [High,Triaged] https://launchpad.net/bugs/341928 [11:54] cjwatson: That's fine; thanks [11:56] if i do a while statement in shell like: while read moo;do echo $moo;done < $multiline_string ... is there any way to avoid using a fifo (read wants to open something, is there any replacement i can use to parse teh string directly ?) [11:57] evand: how much can we change the wy the text looks? ie can we add a background color or a line around the outside? Can we change the X to a dot or such? [11:57] s/wy/way [11:58] kwwii: it's done in cairo, so we can add a different color line around the text or a solid background quite easily [11:58] and yes, we can change the x to a dot [12:00] evand: cool, once I have an idea of what to change I will get back to you...thanks again [12:01] kwwii: great, thanks! [12:01] bah, i'm dense today [12:02] ogra: you mean you're trying to split up $multiline_string into lines? [12:03] cjwatson, no, i'm trying to change one line in a multiline string while keeping the order of the lines intact ... [12:03] ogra: why not just use sed then? [12:03] but i just noticed i'm silly and use: echo "$string"|while ... [12:04] foo | while read has some very odd semantics that are likely to bite you if you aren't careful [12:04] the while loop turns into a subshell so any variables set in there will be lost [12:04] hi [12:04] I'd recommend using sed if you can [12:05] ogra: can you join #nm for a moment? [12:05] hmm, sed is tricky here since the lines can vary and i dont know all possibilities [12:05] what status should be assigned to a bug if you are attached the debdiff and subcribed motu for get sponsoring? confirmed? in progress? [12:05] asac, i'm there [12:06] cjwatson: patch looks good === korn_ is now known as c_korn [13:11] hah [13:11] * cjwatson cracks bug 341928 [13:11] Launchpad bug 341928 in lvm2 "Consistently reproducible "device or resource busy" error on partitioning" [High,Triaged] https://launchpad.net/bugs/341928 [13:12] cjwatson: what was it? [13:12] cjwatson: \o/ [13:13] (oh, nm, it's in the bug) [13:13] elmo: lvm2 scans all devices in /dev and didn't know that it should prefer, well, just about any other name over /dev/block/ [13:13] I haven't actually tested the fix yet, that's in progress [13:13] fader_: thanks for those instructions, very helpful [13:13] cjwatson: No problem, glad to help :) [13:14] cjwatson: (especially with this bug as it's been hitting me on a daily basis ;) ) [13:15] fader_: sorry it took a while; lvm2 and devmapper are twisty as hell and it was tricky to find the exact bit at fault [13:16] I never knew devmapper included a regex library ... [13:16] cjwatson: I can imagine... I don't envy you the task of tracing through all that [13:16] Heh [13:17] Now you just need to add the ability for it to read email and it will be a complete program. [13:19] fader_: it does irc already? [13:20] Hobbsee: Heh, probably not, but I'm counting on there being an SMTP->IRC gateway out there somewhere ;) [13:20] haha :) [13:26] fader_: I think that can be done in about 5 lines of Perl [13:27] If it were a week ago I'd be sorely tempted to file a bug on this. === mariusb_ is now known as marius [14:04] anybody knows when pitti will be around? I have a GPM question [14:06] BUGabundo: he's at the linux foundation summit for the week [14:08] seb128: thanks [14:08] forgot about that [14:08] you can ask your question on the channel [14:08] anyone else in charge of GPM? [14:08] some other people might know [14:08] gonna file a wishbug against it [14:08] what do you call gpm exactly? [14:09] to save profiles [14:09] Gnome Power Manager [14:09] even Win95 did that [14:09] you should open wishlist bug upstream rather [14:09] and current version of GPM doesn't [14:09] I will [14:09]