[00:19] is there any reason to prefer dput over dupload or vice versa? [00:31] aquo: well they basically do the same thing but I personally prefer dput [00:31] aquo: amongst other things, dput will check signatures and checksums described in the changes file before uploading [00:31] preventing an inconvenient upload that just immediately gets rejected [00:32] ok, the documentation for ppa also mentions dput ... [00:33] i think i will stick to this, just didn't want to have two packages for same purpose on my system [00:33] and no need to ask simultaneously in two different channels, by the way [00:34] jdong: you are right. sorry [00:55] hi [00:55] just wanted to notice about this: https://bugs.gentoo.org/show_bug.cgi?id=209460 [00:56] bugs.gentoo.org bug 209460 in Kernel "kernel 2.6.17 - 2.6.24.1 vmsplice Local Root Exploit" [Critical,New] [00:56] http://mureakuha.com/paste/?20add195bc2206d3f38531b7e4d0d977 it works fine for me :) [00:58] teprrr, the proper way to report a security bug is by filing it as a security bug on launchpad.net [00:58] teprrr: Would you mind also filing that in launchpad? IRC bugs sometimes get lost, and that looks fairly important. [00:59] Seveas, wasn't me who reported about it.. just wanted to give a note on this channel, as there are devels around [01:01] teprrr: as said before, IRC is a very lossy medium and not every developer is familiar with or equipt with the ability to address kernel bugs [01:02] teprrr: filing it on Launchpad is the one and only way of doing so, but I bet the security team is already aware of this from their automated vulnerability aggregrators. [01:02] jdong, yup, well, I think it'll get fixed later on when new kernel comes from the upstream :P [01:02] yup, I see [01:02] but it's worth to note, if there are people who maintain their own comps also :P [01:03] again, this is not the right place to do that either ;-) [01:04] fwiw, hardy is vulnerable [01:05] Seveas, yup [01:05] jdong, ye, no problem.. I'll be gone soon, no worries :) [01:17] fixed [01:30] any main sponsor around, who could ack me a sync request (bug #190484)? [01:30] Launchpad bug 190484 in sdl-image1.2 "please sync sdl-image1.2 (1.2.6-3) from unstable/main to main, ubuntu override ok" [High,Confirmed] https://launchpad.net/bugs/190484 [01:30] (there's a CVE fixed with this one ;)) [01:36] * persia seeks a second advocate for http://revu.ubuntuwire.com/details.py?package=whysynth as step 3 in the plan to enable gstreamer-midi for hardy [01:36] Err. Sorry. wrong channel :) [01:37] heh [01:47] it appears that the atl2 module is not working for any kernel packages > 2.6.24-4 [02:01] * avb reading http://milw0rm.com/exploits/5092 [02:01] not good :( [02:05] sweet I don't have to type my password now ;) [02:08] :) [02:08] siti: a prefere not to type my pass editing sudoers :) [02:09] using this tool somebody can not type a pass as well on mine servers :) [02:09] yeah it's a pretty lethal flaw ;( [02:09] i'm wondering how this hole exist throw 5 releases [02:38] jdong: FYI, dupload has those same checks [02:39] cjwatson: thanks for letting me know. I guess my memory is from dated materials from when I first was introduced to dput/REVU :) [02:39] what's your take between the two tools, then? [02:40] I've been using dupload since 2001 and don't see a particular reason to change [02:40] I don't see a whole lot to choose between them if you're coming at it afresh [02:42] sounds reasonable [02:42] jdong: what checks can dupload/dput do? (/me always thought it was just a wrapper to do upload a bunch of files in a specific order using a specific protocol, e.g. ftp) [02:42] sistpoty: making sure the GPG sig on changes is valid, same with the md5sums inside the file [02:42] ah, sure... that sounds sane :) [02:42] indeed :) [02:43] dupload at least can do arbitrary checks in a pre-upload hook, defaulting to GPG checking; dput is probably similar [02:44] both come with examples or options on how to run lintian before upload [02:44] though I wouldn't do that; there are reasons to upload non-lintian-clean packages sometimes [02:44] (debuild runs lintian at the end of a build by default) [03:42] please tell me that the all-new-diskmanager-by-default doesn't have the typos in it [03:44] uhhh. why does that have a milestone, if it's not even set to approved? [03:45] * Hobbsee wonders how to check the activity log of a spec [03:48] Hobbsee: maybe you could ack a main sync for me? *g* (bug #190484) [03:48] Launchpad bug 190484 in sdl-image1.2 "please sync sdl-image1.2 (1.2.6-3) from unstable/main to main, ubuntu override ok" [High,Confirmed] https://launchpad.net/bugs/190484 [03:51] sistpoty: done, but i can't unsub u-m-s [03:51] Hobbsee: thanks! [03:52] heh, me also can't unsubscribe u-m-s *g+ [03:54] :P [04:06] Hobbsee: bug filed against malonge: bug #190608 ;) [04:06] Launchpad bug 190608 in malone "unsubscribe s.o. else" [Undecided,New] https://launchpad.net/bugs/190608 [04:06] -g [04:06] sistpoty: seems kinda logical - at least while you can subscribe others. [04:06] thanks1 [04:07] thanks for your ACK ;) [04:07] you're welcome === asac_ is now known as asac === Opik is now known as Opik^Sleeping [06:44] Good morning === imbrandon_ is now known as imbrandon === masACC is now known as maswan === _emgent is now known as emgent === \sh_away is now known as \sh === hile_ is now known as hile [11:22] siretart, you're not around I suppose? [11:23] hey guys, i've found a bug in hardy but i'm having some trouble triaging it [11:24] gucharmap has recently had a patch that it remembers which character, font, window size, etc when you close it, and next time you open it, it should have the same settings [11:25] however, it seems that in hardy this only works when run from the command line [11:25] when gucharmap is run from the applications menu, it doesn't remember any settings [11:25] cornflakepirate: You likely want to have this discussion in #ubuntu-bugs. [11:25] ok thanks, i'll ask there! [11:29] how long should it take for a new kernel to reach the daily-live cds? [11:33] ssam: Depends on when the daily build happens as compared to the kernel release from NEW. Between 2 and 24 hours usually. [11:34] the powerpc daily live still seems to have 2.6.24-5.9 when 2.6.24-7.12 came out 4 days ago [11:35] (assuming that uname-a at the initramfs prompt is a resonable way to tell) [11:38] (and assume that uname -a giving 2.6.24-5 means a package version of 2.6.24-5.x. kernel version numbers are complicated) [11:40] ssam: according to LP linux-ubuntu-modules failed to build on powerpc [11:41] linux-meta, linux-image, linux-backports-modules and linux-restricted-modules are ok though [11:42] Still, the failure of linux-ubutu-modules likes creates some dependency issues that block inclusion on the CD. [11:42] ok [11:42] is that worth filing a bug on? [11:43] No. The FTBFS is known, and it will be addressed before Alpha 5. The dailies are known unstable. [11:44] ssam: you can have a look at : http://cdimage.ubuntu.com/ports/daily/current/report.html to know what's broken [11:45] that URL does not list linux-ubuntu-modules [11:46] check for linux-meta [11:47] linux-meta is there [11:48] thanks for all the info === \sh is now known as \sh_away [12:15] howdy, hardy experts here ? I just upgraded to that and now the machien wont boot. The problem is that root = md0 and that's for somereason is not activated === ubuntu is now known as _Pete_ [12:16] <_Pete_> anyideas how to fix it ? [12:21] _Pete_: The short answer is that hardy is broken in lots of ways. You might find #ubuntu+1 useful. [13:44] hm, is it me or are launchpad and the wiki down? [13:50] Treenaks: it's you :), LP and wiki WFM [13:51] stgraber: trace ends at canonical-gw.datahop.net for me.. 99% packet loss behind that (gw0-0-gr.canonical.com and wiki.ubuntu) [14:02] what would be the correct package to file this bug against -- https://bugs.edge.launchpad.net/ubuntu/+bug/190677 [14:02] Launchpad bug 190677 in ubuntu "Backport acer-wmi to hardy 2.6.24 kernel" [Undecided,New] [14:03] rohan: kernel bug? 'linux' package, I guess [14:04] asac: will firefox 2.0.0.12 get into hardy or should I grab it from gutsy-security if I want an uptodate firefox? [14:04] wow, i never knew there was just a generic "linux" package [14:04] thanks Treenaks [14:05] rohan: there wasn't, before hardy :) [14:05] geser: i planned that for tomorrow ... in hardy we have now ffox 3 [14:06] Treenaks: doesn't seem that way -- http://packages.ubuntu.com/cgi-bin/search_packages.pl?searchon=names&version=all&exact=1&keywords=linux [14:06] asac: most of the addons I use aren't ported to ffox 3 yet, so I still use ffox 2 [14:06] rohan: hm.. ok :) [14:07] will hardy have ffox 3 as default ffox? [14:07] geser: there will be a session about ffox 3 extension packaging in ubuntu developer week [14:07] geser: maybe join so we can take a look ... often it should not be hard to preport those [14:09] asac: it is about packaging only? The addons I use are from addons.mozilla.org and installed in my profile only. [14:09] there is no change in hell that ubuntu hardy will have 2.6.25, right? [14:10] geser: well packaging also implies looking at changes needed to make it work on ffox 3 [14:11] asac: ok, I will see if I find time to join the session as I should prepare myself for exams that week [14:12] geser: its just 1 hour :) === ryu2 is now known as ryu === jdong_ is now known as jdong [18:19] Is there any way to get Croquet in a deb for AMD64 Hardy? === lando_ is now known as lando === blueyed_ is now known as blueyed [20:14] Hm... The server cd doesn't have all the kernel udebs. Why could that be? === awalton__dos is now known as awalton__ [20:14] d-i was updated to use 2.6.24-7, and it built succesfully, however, the cd-build-log says stuff like: "! Allowing d-i kernel versions: ['2.6.24-5-generic']" [20:15] http://people.ubuntu.com/~ubuntu-archive/cd-build-logs/ubuntu-server/hardy/daily-20080210.log [20:18] Hi === kylem_ is now known as kylem [20:37] Oh, what do you know... The seeds explicitly state which version of the kernel should go into the installer. [20:40] cjwatson: I haven't quite wrapped my head around the shifting about of the seeds, you're doing, so could you please update the appropriate seeds to point to 2.6.24-7 instead of me trying and messing up something? [20:40] * soren goes to make himself useful around the house [20:40] Is https://bugs.launchpad.net/ubuntu/+source/kdeutils/+bug/118723 fixed ? [20:40] Launchpad bug 118723 in kdeutils "KMilo/Volume Hotkeys regression" [Medium,Fix committed] [20:41] cause I also have this bug in gnome [20:41] when I run compiz [20:41] Hi all === hunger_t is now known as hunger [20:47] mateusz: what's the exact cause of that. Kmilo isn't used anymore [20:48] mateusz: any idea why only when you run compiz? [20:48] sladen: when I press volume button with metacity volume changes, while when I run compiz, OSD also apears with a bit diffrent look (transparent) and sound is unaffected. [20:50] Riddell: any idea what KDE/Kilo does when compositing is running? [20:51] sladen: oh sorry it doesnt work even with metacity === Traxer is now known as Traxer|on === Traxer|on is now known as Traxer [20:59] soren: whoops. done. [21:17] cjwatson: Cool, thanks.. Are there any situations in which the seeds should specify a version different from the one d-i references? [21:17] * soren thinks that even the question doesn't make much sense. [21:27] speaking of d-i, I have a question too: if I do echo "foo/bar string mystring" | debconf-set-selections, then install the corresponding package, the variables is used properly. In a preseed file used during installation, if I have d-i foo/bar string mystring and then d-i preseed/late_command string apt-install mypackage, then the variable doesn't seem to be used. [21:28] cr3: Right. [21:28] instead of late_command, perhaps I should be using: tasksel tasksel/first [21:28] cr3: AIUI there are two debconf databases. The one in the installer, and the one in the installed system. [21:29] soren: aha! that sheds quite a bit of light! [21:29] cr3: When you do the d-i thing it lands in the installer one... [21:29] yeah, you get the idea. [21:29] :) [21:30] And by "the d-i thing", I mean preseeding. [21:30] sorry, attempting to watch the telly while irc'ing. [21:30] Or more correctly: I'm attempting to irc while watching the telly. [21:31] soren: yeah, the d-i thing seem to be for the installer debconf database whereas late_command is for the installed system [21:32] cr3: Er... Sort of. [21:32] soren: by the way, there's a section in the example preseed files called "Preseeding other packages" but it doesn't give any sample code. is that section meant for setting preseed variables in the first or second database? [21:33] First. [21:34] soren: to preseed the second database, I essentially have to do the ugly echo thing | debconf-set-selections in a late_command, right? [21:34] cr3: late_command is just that. A late command. apt-install calls apt-get inside the target system (throug chroot and stuff) [21:35] late_command is run in the installer environment. [21:35] soren: err, are you sure about that? haven't I see late_command call in-target? [21:35] if you want to execute stuff in the target system, you use "in-target" [21:35] cr3: Precisely. If it wasn't executed in the installer environment, there'd be no need to specify in-target :) [21:35] ah, my misunderstanding. when you said "installer environment", I read "installed environment" as in "target system" [21:35] Ah. [21:36] cr3: Well, something akin to 'in-target "echo whatever | debconf-set-selections"' ought to work, but I don't know if there are clever shortcuts for that sort of thing. [21:36] all clear now, trying out my new found knowledge on a System76 Jackal system :) [21:37] ...or whehter it's the right way to do it at all. [21:37] soren: yeah, I'd be curious to know as well [21:37] cr3: #ubuntu-install is your friend. [21:37] Er.. [21:37] soren: yeah, but only during cjwatson hours :) [21:37] sladen: why would compositing have anything to do with kmilo? [21:37] cr3: #ubuntu-installer, that is. [21:38] cr3: Quite :) [21:38] soren: yeah, my irssi configuration autojoins that channel :) [21:38] Riddell: quite. But this /is/ KDE we're talking about here. And I hear you're an expert in that area [21:39] sladen: we're not talking about anything, I just saw you asking me earlier today what would happen. I can't imagine it having any effect [21:40] Riddell: mateusz raised it [21:42] guys, mahbe somebody knows the progress with #42926? [21:43] its a bug with too many fonts dependencies in ubuntu-desktop [21:44] they are in recommends but still... [21:46] bug 42926 [21:46] Launchpad bug 42926 in baltix "More granular font selection for the default install" [Wishlist,Invalid] https://launchpad.net/bugs/42926 [21:52] soren: do you think there's a difference between calling "apt-install foo" or "in-target apt-get install foo" in late_command? [21:52] cr3: Not at that point. [21:53] cr3: Er... not true. [21:53] cr3: apt-install makes sure that the debconf frontend of the installer is hooked up with the apt-get command in the chroot, so that you can see stuff happening. [21:54] soren: good to know, especially considering the importance of logging [21:54] cr3: apt-install calls ""debconf-apt-progress --logstderr --no-progress -- apt-get -o APT::Install-Recommends=false -y --no-remove install $packages" || ERRCODE=$?" inside the target. [21:55] soren: no wonder apt-install is provided as a facility, that command is hard on my stomach [21:57] soren: by the way, I tried to specify my own package in the list of "tasksel tasksel/first" packages. however, it never got installed and /var/log/installer/syslog didn't seem to show any errors regarding my package [21:58] is "tasksel tasksel/first" the wrong place to add my package? would it really be more appropriate to use late_command for my own packages? [21:59] cr3: You want pkgsel/include. [21:59] cr3: tasksel/first is for tasks. [22:02] soren: how can I preseed values for packages insatlled with pkgsel/include? [22:02] cr3: magic [22:03] aha! that's in the installer environment so I suspect d-i just like any other preseed in the file [22:07] cr3: Probably not. [22:10] soren: oh yeah! we'll just have to see about that :) trying it out [22:12] cr3: :) [22:13] cr3: Will you be in Boston two weeks from now, by the way? I forget.. [22:13] soren: seeds should always match the d-i initrd; it was my fault for not updating them [22:13] cjwatson: In that case I'm curious why it's necessary to specify them at all? [22:13] cjwatson: Why not just specify the meta package? [22:14] soren: there's no metapackage for udebs [22:14] cjwatson: Point. [22:15] cr3: the owner (where you keep putting "d-i") may only be "d-i" if you're preseeding an installer component; if you're preseeding a package to be installed on the target system, it must always be the name of the package [22:15] soren: that's not scheduled yet, so probably not [22:15] cjwatson: Hm... I still think there ought to be some way to make germinate just use whatever the installer already has. [22:15] soren: quite probably, there just isn't :) [22:15] cjwatson: Ah :) [22:16] I think the best answer would be to put virtual package names in the seeds and have it pick the version-wise newest [22:16] cjwatson: name of the package as in: package preseed/variable string value [22:16] cr3: yes [22:16] cjwatson: that's actually very clean, thanks! [22:16] cjwatson: Those get copied to the target system's debconf? [22:17] cr3: if you get it wrong it will either (a) fail to preseed the package entirely (b) fail to disappear from the debconf database on purge several years later and cause you weird headaches [22:17] soren: anything with an owner != d-i gets copied [22:17] soren: hey, have you forwarded your gtk-vnc ext_key_event patch upstream? (in case it makes sense, and it looks like it does to me) [22:17] cjwatson: Ah, clever. [22:17] pochu: It's *from* upstream. [22:17] cjwatson: heh, I've encountered (b) before. very fun to debug :) [22:19] when you purge a package that uses debconf, all questions that are owned by it are dropped, unless they're also owned by another package (debconf stores a list) [22:21] at some point i recalled coming across something indicating that prelinking was not done by default for some reason. does anyone recall what that would be? [22:21] soren: really? Because I needed to update it for the 0.3.3 release, and it wasn't there [22:22] pochu: It's not *in* upstream. [22:22] pochu: *from*. [22:22] :) [22:22] soren: oh :) and they don't want to apply it in their source tree? [22:22] pochu: I'll look at your debdiff tomorrow. [22:22] pochu: Not yet, no. [22:23] soren: thanks [22:25] * soren heads to bed [22:26] Does Ubuntu have ThinkFinger packages with GUI ? [22:30] mateusz: no, i dont think so [22:30] mateusz: most likely. I've never tried getting it to work; perhaps start here: https://wiki.ubuntu.com/ThinkFinger The only commandline you need is a single tf-tool --acquire per user [22:30] it's manual isn't it, not packaged [22:31] it have it working on my toshiba [22:31] popey: well also modprobe hdaps cause error device not found.. while it should work [22:35] mateusz: do you have a really new machine? perhaps there are new IDs that will need adding to the next release? [22:36] sladen: not so new its Lenovo T61 [22:38] heh, i am on a t61 too [22:38] popey: does it work for You ? [22:38] FATAL: Error inserting hdaps (/lib/modules/2.6.22-14-generic/kernel/drivers/hwmon/hdaps.ko): No such device [22:38] this one doesn't have a fingerprint reader [22:38] popey: ahoy! by the way, thanks for the screencast information you provided a couple weeks ago, much appreciated. [22:39] np cr3 [22:39] popey: mine has finger reader [22:39] mateusz: there are 3 varieties of t61, this one doesnt [22:39] popey: I've been exploring other ways to create a screencast which doesn't require two machines (vnc or rdesktop), I'll send you a howto in case you might find it interesting [22:39] popey: how to fix hdaps to work on mine ? [22:40] cr3: i dont use two machines most of the time, i record the local machine [22:40] mateusz: no idea, sorry, I dont have the same machine as you so can't test [22:40] mateusz: sounds like you have a new revision of a machine (even if "T61" has been around for a year) [22:41] sladen: there are 3 varieties, one nvidia based, one intel based iirc [22:41] soon lifprint should be fully ready [22:41] it supports practically all model of fingerscanners [22:41] popey: mine is intel [22:41] will be good point to integrate into ubuntu [22:42] How can i change the graphic driver in hardy? I can not select it anymore with dpkg-reconfigure xserver-xorg. [22:42] libfprint* [22:42] sladen: how to read this device ID ? [22:42] sladen: and how to recompile easly Ubuntu's kernel? [22:42] I'll try to add this ID to driver [22:44] mateusz: google for something like hdaps t61 update new ids [22:46] sladen: people on #hdaps says to try module tp_smapi [22:46] cjwatson: pkgsel + package preseed/variable worked on gutsy, thanks! [22:46] which is new implementation [22:47] mateusz: okay, that one is doing access via the "offical" BIOS API, rather than raw hardware access [22:47] mateusz: Ubuntu will have tp_smapi in the next version of the kernel (mostly) [22:48] probably next week [22:48] amitk: ok I'll write it down (tomboy) [22:51] popey: can your method record a video playing with gstreamer? [22:52] popey: I mean, a video within the screencast :) [22:52] i dont see why not [22:54] amitk: installing from debian package automates and compiling with module-assistant was easy go [22:55] good... [22:55] amitk: [23443.237067] hdaps: device successfully initialized [22:56] cool [23:39] mr_pouit: does XFCE have some kind of built-in help system that can show manual pages, or anything that's equivalent to yelp/konqueror's man: URLs? I'm trying to do a survey of everything that implements a man page viewer [23:48] I've been inspired to write a free programming language. [23:48] Here's the hello world program: http://theunixgeek.pastebin.com/d600ba359 [23:53] How to add a patch to ubuntu kernel ? [23:53] cjwatson: no [23:54] mateusz: meaning "how to apply one yourself" or "how to get one added to Ubuntu"? [23:55] crimsun: how to apply patch on the internet not included in ubuntu [23:55] crimsun: http://www.zen24593.zen.co.uk/hdaps/hdaps_protect.20060430.patch [23:55] crimsun: there are hdaps utils in ubuntu but I am not sure if kernel is patched against it ? [23:57] crimsun: I want this to work http://www.thinkwiki.org/wiki/How_to_protect_the_harddisk_through_APS [23:58] crimsun: so.. installing kernel src, make menuconfig; make-kpkg thing.. and dpkg ?