[12:09] Is there a way for me to convert my daily system to use all the default artwork? [12:16] hornbeck: right click on the desktop, and select the Ubuntu Default Desktop [12:16] it just did it by its self [12:17] crashed my x for some reason [12:17] than came back up with the new artwork [12:17] who are the people? [12:17] models [12:18] figured [12:18] the default desktop for new installs has the ubuntu logo [12:18] yeah I have it up [12:18] nice brown [12:18] I like it [12:18] but the preview had the calendar selected, so everyone who had the preview and even looked at the desktop options, gets the calendar now [12:19] man for some reason when it crashed it launched another x display [12:19] hmm [12:19] we need to tell them somewhere that more is available [12:20] more artwork, that is [12:20] wiki? [12:21] for instance :) [12:21] there needs be readme like doc on the wiki I suppose. to explain how to get over certain quirks that might get fixed or noticed on the release. [12:22] yeah [12:22] we need to just figure out a way to organize the wiki alittle better [12:22] need a sitemap on it [12:22] sabdfl: congratulats to the idea with the pics :) [12:22] yes, we do. firstly, the documentation areas needs to become the documentation develop area, [12:23] as it's more of a collectiong of docs and stuff. I am going to organize the /UDP page, [12:23] so it will look more as a starting point for people seeking the latest documetation which is not on *.ubuntu.com [12:23] I think we should have one main page for docs that links off to new docs [12:23] (yet) [12:24] well at least we agreed on that [12:24] ok cool. [12:24] amu: thank you [12:24] sivang do you want to start putting that together [12:25] sivang we need to do a site survay to see where all the docs are [12:25] I'm on to it \ [12:25] ok [12:25] yes [12:25] but let's wait for the full response [12:25] everybody, please let us know if this is annoying the RC stuff here..:) [12:26] we might do it on #ubuntu-docs [12:27] :) [12:27] I think we are fine right now [12:28] sivang: did you see the new start page for firefox? [12:28] hornbeck : yes I did, mark sent me the final. [12:29] cool [12:29] hornbeck : however, it doesn't include the gfx, what's on there gfx wise? [12:29] sivang: just the logo at the top [12:29] we need to get some how-tos for the site also [12:29] cool [12:30] or tutorials that is [12:30] agreed [12:30] however, we might want to save all this decisions for until the meeting? [12:30] I am making a list of things to bring up at the meeting [12:30] so a larger crowd would attend and comment, suggest, take up tasks etc? what do you think? [12:30] stuff I think needs to be done [12:30] great [12:31] I am just thinking outload :) [12:31] send over when done. [12:31] did you ever get tomboy working? [12:31] cause it is all in tomboy notes [12:31] I did on the laptop, [12:31] I'll install it here. [12:31] ok, so would you rather a ooo file [12:32] why not it's already installed here :) [12:32] ok [12:32] 2.49 eta for RC testing... [12:34] I am so happy people have keep with my format on the hardware support === ggi [~ggi@ggi.base.supporter.pdpc] has left #ubuntu-devel ["Leaving"] [12:51] can anybody tell if I might get better throughput using rsync instead of the torrent? [12:52] sivang: probably rather the other way around [12:52] sivang: notes on ideas on the way. [12:52] tomboy really needs a export function [12:52] hornbeck : thanks [12:52] no prob [12:53] Mithrandir : you mean rsync is better? Looks like it, the torrent only use 50% of my bandwidth [12:53] sivang: no, the torrent should be better. [12:54] Mithrandir : I've had better results with the rsync untill now. maybe the torrents have dropped (I've used to get those at 98Kb/s) due to heavy load. [12:55] the torrents should go up with load, not doewn [12:56] s/doewn/down/ [12:56] oh right, cause you have more uploaded.. [12:56] yup [12:58] upstream seeds [01:03] I added the info about the doc meeting to my blog [01:03] since I know there are some ubuntu users who read it [01:04] opps that was supposed to be pm to sivang [01:04] no prob :) === nasdaq4088 [sdfsd@tkp-ip-nas-1-p82.telkom-ipnet.co.za] has joined #ubuntu-devel === lamont_r [~lamont@168-103-92-87.dnvr.qwest.net] has joined #ubuntu-devel [02:30] moof [02:53] moof? === lamont_r is between a private lesson and the evening's jujutsu class [03:11] did we hear any more from alex yet? === lamont_r goes to class. on again in a couple hours. [03:12] what's the word on netgear wireless cards? [03:12] prolly 802.11g [03:16] "I'd be upset if my splash screen suddenly featured a tribute to Ronald Reagan, or a reminder that Jesus loves me." [03:16] ...ah, the memory of gimp splash screens past [03:46] hhaa. iso donwloaded. going to test RC on my new SCSI fileserver :-) === hazmat [~hazmat@c-24-15-10-12.client.comcast.net] has joined #ubuntu-devel [04:21] Keybuk: 'The Politically Incorrect Release' [04:29] got some errors on the SCSI AIC-7xxx driver on boot up, but post install seems to be running smoothly [04:30] also got "pciehp" module insertion error [04:30] "Operation not permitted" [04:31] ok, not to download live cd [04:31] *now [04:58] ooh, Python 2.4 [05:04] Keybuk : word has it you're the Python lead and expert for Ubuntu, how do you get accepted to the Python team? :) [05:06] Keybuk : I meant, how can someone can accepted to that fine high lever programming team? :) [05:07] to be honest, I haven't really put a huge amount of thought into it yet [05:07] I'm mostly involved with other Canonical projects at the moment, and spending only relatively little time on Ubuntu [05:08] Keybuk : Oh, I see. ARCH, rozetta and freinds? === azeem_ [~mbanck@socks-out.lrz-muenchen.de] has joined #ubuntu-devel [05:08] the general focus will be getting Python scripting support into every application [05:09] and brushing up and evangelising that which already exists [05:09] yeah, I'm working on the really kick-arse arch stuff :p [05:10] Keybuk : does it menifest on arch's CVS? or , coming to think of it, ARCH's arch archive? :-) [05:11] nah, there's no public manifestation of what I'm working on yet; soon, hopefully [05:12] Keybuk : canonical had spawned a private tree of the code, and would flush it back when done? [05:13] nah, it's not directly related to the arch code; more exciting things you could do with arch :o) [05:14] Keybuk : sounds very apptetiting :) === lamont_r tries to import a disk with sound juicer, and discovers that 'music player' doesn't find any of the .ogg's... [05:15] why is that, I wonder? [05:16] maybe I should finish my upgrade? [05:18] lamont_r : i try to upgrade my dpkg's failing on me...:) [05:18] ouch [05:19] 202 upgraded, 1 newly installed. [05:22] morning guys [05:23] mdz: you around? [05:24] morning fabbione [05:24] up so early? [05:25] feeling better? === hornbeck [~hornbeck@adsl-68-92-227-78.dsl.okcyok.swbell.net] has joined #ubuntu-devel [05:25] fabbione: yes [05:25] sivang: i was ok yesterday too :-) [05:25] fabbione: how is the move going? === lamont_r heads for home. [05:25] mdz: portmap. i can upload it, but didn't you have the packages already? [05:26] fabbione: yes, I have a .changes ready to upload [05:26] mdz: i decided to stay another day here at the old house and have the office on the floor [05:26] mdz: it will give my ISP another day to fix the adsl line at the new house [05:26] ah [05:26] mdz: do you want me to do another .changes or will you upload it directly? [05:27] mdz: in anycase tomorrow i must move out. either to the new house or to the other office [05:27] fabbione: I thought it might be good to wait some time after the RC before uploading new packages [05:27] mdz: make sence [05:28] mdz: but we need to get these new packages tested as well [05:28] fabbione: but on the other hand, I would like for it to get as much testing as possible before final [05:28] :-) [05:28] mdz: i think you should upload it [05:28] fabbione: I think we are both very confident that it does not break the desktop [05:28] mdz: i am confident [05:28] my only concern is users who are running NFS and such [05:28] desktop is working fine here [05:28] you tested client or server or both? [05:28] mdz: yes [05:28] client is 100% ok [05:28] I have been without sleep for a while, so excuse my poor recall :-) [05:28] server needs to be reconfigured in case [05:29] mdz: it's ok.. blame bugzilla that deleted all the nice report i did yesterday [05:29] I wrote some notes in bugzilla [05:29] fabbione: justdave fixed that, by the way [05:29] mdz: yup [05:29] fabbione: do you think there is any _simple_ way we can make the NFS server Just Work if the user installs it? [05:30] I guess the answer is, "wait for hoary" [05:30] when hopefully we can get rid of fam, and follow the plan I outlined in bugzilla [05:30] jdub: what are the odds on replacing fam for Hoary? [05:30] yes i read the comments [05:30] i agree with them [05:31] gamin seems to be active [05:31] so hopefully it will be a viable option [05:31] ok, I guess I will upload [05:31] i mean.. who implements a nfs server has at least a minimum knowledge of filysystems [05:31] and what to look for [05:31] mdz: so the GetModeLine isn't OK now? [05:31] it's not that simple for a newbie to write an exportfs [05:31] but I am not aware of any other unix which ships a portmap listening only on localhost by default [05:31] mdz: too bad for them :-))) [05:32] daniels: we have a list of serious issues to deal with; let's leave trivial items for Hoary [05:32] daniels: what GetModeLine? the 099l ?? [05:33] I don't object to the patch, but we need to change as little as possible between now and final [05:34] mdz: let's discuss X in a few minutes ok? [05:34] fabbione: ok [05:34] i am preparing an interdiff [05:34] but working from the floor doesn't make me any faster ;) [05:34] I would like to give gzip a good workout [05:35] but there is nothing to upgrade :-) === fabbione uploads X to make mdz happy [05:35] har har [05:35] mdz. you can simulate ;) [05:35] dpkg-getselection > log [05:35] apt-get --reinstall install < log [05:35] or something like that [05:36] daniels: do you realize that xscreensaver writes a line to stdout (and thus ~/.xsession-errors) whenever it activates anyway? [05:36] there is a lot of work to be done on eliminating unnecessary disk access, but it is a Hoary project [05:37] fabbione: yeah, 099l [05:37] mdz: ok [05:37] if you try to reinstall many packages at once, apt tends to get confused [05:37] mdz: well .. that's how i saved my server once [05:37] mdz: / was lost in lost+found [05:38] eek [05:38] found dpkg via md5sum [05:38] and the minimum pieces required to run apt [05:38] impressive [05:38] also via md5sum [05:38] that is a lot of pieces [05:38] glibc and nss and libapt and the methods [05:38] mdz: not if you have /var/lib/dpkg/info intact [05:38] script to check md5sums === azeem [~mbanck@socks-out.lrz-muenchen.de] has joined #ubuntu-devel [05:39] and to move files around according to them [05:39] fabbione: I noticed in my testing that when you upgrade portmap, it does not restart :-/ [05:39] once you get the minimum working [05:39] zack [05:39] I do not know if the bug exists in Debian [05:39] hmm it did here [05:39] I don't have any more Debian systems to test [05:39] wait a sec [05:39] every time in my test upgrades, it would still be listening after upgrade [05:39] because it was still the old process [05:40] dpkg - warning: downgrading portmap from 5-4ubuntu4 to 5-4ubuntu3. [05:40] * Stopping portmap daemon... [ ok ] [05:40] Installing new version of config file /etc/init.d/portmap ... [05:40] * Starting portmap daemon... [ ok ] [05:40] * Restoring old RPC service information... [ ok ] [05:40] same when upgrading [05:42] I will test in a moment when I upgrade my NFS server :-) [05:43] once the package is built [05:43] mdz: back in 5 minutes to discuss X === fabbione needs to visit the thinking room [05:43] Mithrandir: ping? [05:43] asleep I suppose === azeem [~mbanck@socks-out.lrz-muenchen.de] has joined #ubuntu-devel [05:52] only Mithrandir, elmo and Kamion seemed to have been able to reproduce the gzip bug [05:52] Mithrandir gave me an account on his system, but his IP seems to have changed [05:53] re [05:54] http://people.ubuntulinux.org/~fabbione/diff [05:54] mdz: ^^ [05:54] reading [05:54] ubuntu24 to ubuntu25 [05:54] fabbione: do you have tiff prepared? [05:54] there are a lot of noisy bits like changelog, CHANGESETS, TODO [05:55] mdz: i woke up when i said goodmorning here [05:55] ah [05:55] mdz: i am going to do it after we discuss the diff [05:55] it won't take too long since the patches are in bugzilla [05:55] don't worry about it [05:58] i only have one pending change to the FAQ [05:58] but i guess that's not an issue ;) [05:59] the doc updates of course I have no problem with [05:59] we need to talk about xkb, v4l and #2328 [05:59] I read #2328, but from only the bug, it is impossible to tell either the problem or the fix :-) [05:59] ok [06:00] mdz: 2328 is a simple fix [06:00] it's at the end of the diff [06:00] it is hard to see with such a long line; what is the change? [06:00] - db_get xserver-xfree86/config/monitor/mode-list [06:00] + if [ -n "MAXRES" ] ; then [06:00] + RET="MAXRES" [06:00] + else [06:00] + db_get xserver-xfree86/config/monitor/mode-list [06:01] mdz: is in xserver-xfree86.postinst.in and .templates [06:01] the end of the diff is a template change [06:01] yes and the bit before it [06:01] there are 2 problems basically [06:01] fabbione: shouldn't that be $MAXRES? [06:01] daniels: correct [06:01] because "MAXRES" will always be strlen() != 0 [06:02] fixed [06:02] mdz: Carlos specific problem is that his resolution [06:02] was not known at @60Hz [06:03] so basically debconf was returning 640x480 @ 60Hz [06:03] as default value [06:03] but this is a specific fix [06:03] reviewing the logic [06:03] i noticed that can happen for all the unknonw resolutions [06:03] now if MAXRES exists it means that we are either probing or guessing [06:04] and it contains the same value as the template (theoretically) [06:04] this change is very uncomfortable to me [06:04] mdz: it's simple [06:04] it's difficult to explain [06:04] it had one bug already [06:05] mdz: i didn't finish all the tests yet [06:05] it's not like i wouldn't have spot it [06:05] the problem with changing anything in the detection logic at this point is that we can never re-test all of the hardware that has been tested [06:05] we have had hundreds or thousands of tests at this point [06:06] mdz: it doesn't really change the logic [06:06] it fixes it [06:06] mdz: if you see the script carefully: [06:06] MAXRES = something [06:06] db_set template $MAXRES [06:06] later: [06:06] db_get template [06:06] yes... [06:06] and than we compare what's in there with the list [06:06] now [06:07] limitation: [06:07] the template might not be aware of that MAXRES [06:07] debconf returns the default of 640x480 @ 60Hz [06:07] the frequencies detection goes banana [06:07] the if then else fi [06:08] simply avoids a limitation to the debconf template [06:08] and it is not intrusive. [06:08] under what circumstances is MAXRES set? [06:08] i am not "growing" the template like i did for the other resolution list [06:08] MAXRES is set only when we cannot detect the frequencies via xresprobe [06:09] # if we are here we did not detect the frequencies, but we have the resolution. [06:09] # get the highest resolution. [06:09] so MAXRES is the answer to the mode question [06:09] so this changes the behaviour for (potentially) every circumstance where the user is asked the mode question [06:10] nope [06:10] because the mode list is all known and mapped to mode-list [06:11] so if the user is promped for the resolution it can only enter well known values of MAXRES [06:11] that will match later on the values into the frequencies case [06:11] i wonder if fixing this would allow high quality displays to do above 100hz freqs :) [06:11] sivang: not in your case [06:12] it's another bug here we are talking about [06:13] mdz: what you need to look at is this: [06:13] # set a sane default for mode-list [06:13] MAXRES="$MAXRES @ 60Hz" [06:13] db_set xserver-xfree86/config/monitor/mode-list "$MAXRES" [06:13] yes [06:14] the problem happens only if MAXRES is not known to the template [06:14] currently, that falls back to the default in the template if $MAXRES is not in the choices [06:14] the default is 1024x768 @ 60 Hz [06:15] mdz: apparently no, if you try to set the template and than db_get [06:15] you get the first value of the list [06:15] that might be a frontend bug [06:15] I agree [06:15] that is a bad debconf behaviour [06:15] frontend ;) [06:17] does it look too dangerous to you? [06:17] otherwise there is another simple way to implement it [06:17] if that can make you feel better [06:17] i can db_get the template [06:18] if the MAXRES is != null and the 2 are the same than ok [06:18] otherwise prefer MAXRES to db_get [06:18] same effect as you see it in the patch [06:18] kust written in a different way [06:18] s/kust/just/ [06:19] here is my concern [06:20] well [06:20] is there any possibility of this affecting any case other than when 640x480 @ 60Hz would be the result? [06:21] how can we test this? [06:21] nope [06:22] mdz: simply asking carlos to test the new package [06:22] he has the testcase ;) [06:22] I am not so concerned about the bug [06:22] and just try to reprobe on well known machines [06:22] it seems to affect very few people [06:22] it is the regression testing [06:22] mdz: well not really a few... [06:22] which cases do we need to test for regressions? [06:23] there were a bunch of people having the 640x480 problem [06:23] where xresprobe fails? [06:23] mdz: where xresprobe returns a resolution but not a frequency set [06:23] you can even wrap xresprobe to simulate that for you [06:23] just write a bogus sh script to return: [06:24] rabidbt: blabla [06:24] ops [06:24] identifier: blabla [06:24] freq: [06:24] res: 1024x768 [06:24] or something like that [06:24] that will trigger the MAXRES [06:24] but you need to use a not known resolution [06:24] no discard my last comment [06:25] it can be also on well known resolution [06:25] it doesn't really matter [06:25] mdz: looks like the preview set calendar as the default, so anyone upgrading a preview to current bits gets people. [06:26] lamont: if that is happening to _anyone_ who upgrades from preview, there is a new and interesting bug somewhere [06:26] it is known to happen to anyone who even breathed on the background preferences dialog [06:27] fabbione: it is all a question of how much testing we can get for it [06:27] mdz: I didn't change background preferences prior to today, and when I pull that up, it says 'ubuntu monthly calendar' [06:27] lamont: you don't have to change them; you only have to look [06:27] I see. [06:27] anyone who has a ~/.gnome2/backgrounds.xml or whatever [06:27] will get a surprise artwork change [06:28] I see [06:28] ok. [06:28] and does removing the file make things better, or worse, I wonder? [06:28] removing the file should give you the new default [06:28] which is the gradient+logo background [06:29] fabbione: thanks for the explanation, I understand the change much better now [06:29] fabbione: can we talk about xkb? [06:29] mdz: so which all prism drivers do we have now? [06:29] mdz: of course. [06:29] lamont: I don't think we add any prism drivers, but we do add the firmware for prism54 [06:29] mdz: we have 2 problems with XKB [06:29] so we have whatever is in the 2.6 tree [06:30] mdz: so prims54 should "just work", or should be "trivial to enable"? [06:30] lamont: should 'just work' post-installation [06:30] mdz: for the testing i can simulate all the cases replacing xresprobe. that's not a problem [06:30] may or may not work in the installer, not sure of the status of that one [06:30] mdz: it was already in my TODO. it's not like i tend to push untested changes [06:31] mdz: let's go with xkb when you are ready [06:31] mdz: what about liveCD? [06:32] lamont: the live CD has a completely different kernel [06:32] fabbione: I need 10 minutes for workrave [06:32] mdz: ok [06:32] wrists are killing me [06:32] mdz: i need to help my gf a few minutes so it fits perfectly ;) [06:32] we publish a livecd RC? [06:34] looks like the answer is 'no' === lamont checks and discovers that his warty mirror will fit on a dvd. Just barely [06:48] well, about 20MB to spare.. :-) [06:50] lamont: I thought you weren't able to build one which booted [06:50] lamont: did you and alex find a solution? [06:50] mdz: haven't heard from alex since he headed for home. [06:51] lamont: please keep on him [06:51] but he had uploaded one, and it looks like we held off on calling that an RC. [06:51] he's on the top of my list. [06:52] the one that he built was based on day-old stuff, rather than the RC packages [06:53] right [06:54] have there been any post-rc uploads yet? [06:54] fabbione: back [06:56]