| broder | huh, i apparently missed the exciting discussion | 00:03 |
|---|---|---|
| broder | Laney, maco: if you want people to make core-dev look less intimidating, you two should apply. right now i feel a little awkward applying for core-dev when my app is going to be reviewed by two people who were motu before me and who i know do more than i have recently | 00:04 |
| micahg | broder: I don't think most people are thinking about that | 00:05 |
| * ajmitch agrees that they should apply though :) | 00:06 | |
| slangasek | bryceh: installation should always recurse down the recommends as well as the depends; so I don't know why it didn't get pulled in on install. Is it *available* at install time? (I.e., is it on the CD where it should be?) | 00:27 |
| slangasek | cjwatson: bugpattern - all seems reasonable to me | 00:27 |
| bryceh | slangasek, cjwatson sorted it. I thought pitti had seeded it but it hadn't. | 00:29 |
| slangasek | bryceh: got it | 00:29 |
| bryceh | er, wasn't | 00:29 |
| lamont | if I go lucid->maverick->natty, do i need to bother rebooting into maverick? | 00:31 |
| * penguin42 would | 00:33 | |
| * charlie-tca would, just to run all the updates | 00:34 | |
| slangasek | lamont: let us know! :) | 00:34 |
| lamont | slangasek: heh | 00:34 |
| slangasek | lamont: btw, has kees pestered you yet about the bind9 package not shipping the dnssec root cert? :) | 00:35 |
| lamont | yeah | 00:35 |
| slangasek | that sounds inconclusive. :-) | 00:35 |
| lamont | he even convinced me that having installs of bind9 a year down the road fail horribly to run out-of-the-box, until manually corrected, makes sense | 00:36 |
| slangasek | ah :-) | 00:36 |
| lamont | /bin/sh: $python tests/test_all.py -q; \: not found | 00:36 |
| lamont | neat. | 00:36 |
| maco | broder: i kinda doubt ive done more than you recently :P also my next aspiration is kubuntu-dev :) | 02:03 |
| maco | broder: all ive done lately is sorta-emergency ubiquity poking | 02:04 |
| micahg | maco: there's no rush :), kubuntu-dev is very cool | 02:09 |
| broder | maco: i may be giving you guys crap as much/more than i'm being serious :-P | 02:12 |
| maco | and now that alpha 2 is out and kubuntu still doesnt have a working ubiquity, the emergenciness of my maybe figuring out how to fix it goes up *sigh* i think slangasek was a lot closer to figuring out what its doing than i am | 02:13 |
| slangasek | maco: oh ugh, is that bug still there? I figured it'd gotten fixed somewhere along the line, since I hadn't heard any more about it :/ | 03:01 |
| slangasek | maco: I understand what it's doing *wrong*, I just have no proposed solution for how to make it DTRT again | 03:02 |
| maco | slangasek: i havent seen an email from lp saying i can stop worrying about it... | 03:02 |
| maco | (i did see 9 emails from lp saying harald had decided i get 9 more bugs...) | 03:02 |
| maco | slangasek: ooh though given i see some bug reports about it now crashing on the partitioner, maybe it is getting past that part. will need to attempt in a VM | 03:05 |
| slangasek | we're talking about the keyboard map bug, right? which should have been language-specific? | 03:08 |
| ScottK | slangasek: Could you promote grantlee and it's binaries back to Main. kde4libs is now depwait for it. | 03:27 |
| slangasek | ScottK: done | 03:36 |
| micahg | ScottK: so, anything left to do with bug 601662 | 03:38 |
| ubottu | Launchpad bug 601662 in grantlee (Ubuntu) "[MIR] libgrantlee-dev" [Critical,Triaged] https://launchpad.net/bugs/601662 | 03:38 |
| ScottK | slangasek: Thanks. | 03:38 |
| ScottK | micahg: Yes. It needs some work, but let's not block everything for a MIR that's been undealt with through a whole dev cycle. | 03:38 |
| micahg | k | 03:39 |
| slangasek | the package was already in main last cycle, so I think it's reasonable to repromote it on that basis anyway | 03:39 |
| pitti | Good morning | 04:50 |
| pitti | bryceh: no, I didn't check main promotion for xdiagnose, I just wanted the thing to work and use GTK 3 :) | 04:51 |
| pitti | bryceh: it doesn't need seeding, it already has a reverse dependency | 04:51 |
| pitti | i. e. it wants to go into main already, just needs an MIR and actually promoting it | 04:51 |
| dupondje | Only Gnome Classic seems to work atm :( | 05:01 |
| dupondje | :p | 05:01 |
| pitti | not unity 2d? I heard talks that even 3d works on some platforms now | 05:02 |
| dupondje | nope, Xorg going 100% cpu and doing nothing :p | 05:02 |
| dupondje | gnome3 starts, but no menu's | 05:03 |
| dupondje | unity 3d also 100% cpu on Xorg | 05:03 |
| dupondje | :) | 05:03 |
| dupondje | something in last upgrades broke Gnome3 | 05:04 |
| pitti | oh, somehow I read that as "on arm" | 05:04 |
| pitti | dupondje: you might want to talk with RAOF then | 05:04 |
| RAOF | dupondje: Define ‘last upgrade’. Also, nvidia binary driver? | 05:05 |
| dupondje | nope | 05:05 |
| dupondje | broken Zeitgeist couldn't cause that ? | 05:06 |
| dupondje | else I upgraded compiz yesterday | 05:06 |
| RAOF | Oh, yeah. Jason was reporting that ubuntuone was eating all his CPU. | 05:07 |
| RAOF | Possibly a similar thing? | 05:07 |
| dupondje | Yea zeitgeist causes that | 05:07 |
| dupondje | but that could also cause Gnome3 doesn't start completely ? | 05:07 |
| dupondje | thats https://bugs.launchpad.net/ubuntu/+source/zeitgeist/+bug/807203 btw | 05:10 |
| ubottu | Ubuntu bug 807203 in zeitgeist (Ubuntu) "ubuntuone-syncdaemon crashed with AttributeError in __getattr__(): 'Symbol' object has no attribute 'PAGINATED_TEXT_DOCUMENT'" [High,Confirmed] | 05:10 |
| dupondje | Any idea's on how I can debug the real cause ? Could be usefull :) | 05:16 |
| didrocks | good morning | 06:42 |
| evfool | morning didrocks | 06:50 |
| didrocks | hey evfool | 06:50 |
| dholbach | good morning | 07:56 |
| mvo | cjwatson: the changelog stuff should be fixed now and the backlog it has caught up on now over night | 08:15 |
| === hunger_ is now known as hunger | ||
| jml | good morning | 08:29 |
| RAOF | Ra raw! What's broken? :) | 08:30 |
| jml | RAOF: M-d in my virtualbox install of Ubuntu | 08:34 |
| RAOF | Oh. Again, or still? :) | 08:35 |
| jml | RAOF: More "still" than "again". | 08:35 |
| jml | RAOF: am trying out the proprietary alternatives to see if things are better there | 08:37 |
| RAOF | Ah. | 08:37 |
| RAOF | It's not unity annoyingly eating the super key? | 08:37 |
| jml | RAOF: no. At first I thought it was (and I think unity eating super complicates things), but after much xmodmap hackery and also, you know, seeing what happened without unity, I ruled it out. | 08:38 |
| RAOF | Heh. It's my great fear that at some point someone will really want me to learn about X keyboard handling in detail. | 08:39 |
| jml | RAOF: I guess as a fallback I could get used to having meta & super switched around. But that's not the way that technology is supposed to work. | 08:40 |
| RAOF | Right | 08:40 |
| cjwatson | maco: I'm pretty sure I fixed the keyboard bug. I tested my fix and I could no longer reproduce the bug. | 08:45 |
| cjwatson | mvo: great, thanks a lot | 08:45 |
| pitti | infinity, lamont: would it be possible to kill the glib build on allspice and gnome-keyring build on yellow? they both eternally hang in the test suite | 10:37 |
| pitti | (and block amd64 builds) | 10:37 |
| pitti | the i386 build of keyring on vernadsky has the same problem apparently | 10:38 |
| pitti | infinity, lamont: bah, today is a good day for a buildd to die :( pygobject on crested is hanging as well, can you please kill, too? | 10:59 |
| pitti | I understand the latter, and have a fix; for keyring/glib we are currently testing a fix | 11:00 |
| === dholbach_ is now known as dholbach | ||
| pitti | elmo: ^ can you kill builds? lamont will probably still sleep for a couple of hours | 11:04 |
| === chrisccoulson_ is now known as chrisccoulson | ||
| lamont | pitti: infinity can't | 11:23 |
| lamont | on it | 11:23 |
| lamont | pitti: if I'm hearing you correctly, that's glib on allspice, pygobject on crested (how about actinidiaceae??) and gnome-keyring on yellow? | 11:26 |
| pitti | lamont: actinidiaceae can be killed as well | 11:26 |
| pitti | lamont: thanks! | 11:26 |
| pitti | need to disappear for a few hours, bbl | 11:27 |
| === MacSlow is now known as MacSlow|lunch | ||
| dupondje | All right :) Gnome is working again :) | 12:17 |
| dupondje | what was broken now ? | 12:18 |
| dholbach | if somebody could approve my post to u-d-a I'd appreciate it :) | 12:24 |
| dholbach | supi | 12:25 |
| cjwatson | dholbach: done | 12:30 |
| dholbach | cjwatson, thanks muchly | 12:32 |
| nigelb | james_w: Do you want to talk about packageme for 5 minutes at UDS lightning talks? | 12:52 |
| james_w | UDS? | 12:52 |
| nigelb | er, UDW | 12:53 |
| james_w | not this time I don't think | 12:53 |
| nigelb | ok, np :) | 12:53 |
| mterry | apachelogger, looking at the MIR now | 12:54 |
| jbernard | serge_: no objections here | 12:59 |
| lool | doko: Not sure whether you saw this, but latest gcc-4.6 has a libgcc1 breaking older gcc-4.4 and 4.5, which seem to need a merge from Debian | 13:12 |
| doko | lool: yes, seen | 13:13 |
| abhinav- | dholbach: kudos for bringing the long descriptions to harvest but I think something is wrong with harvest at the moment. can't list bugs in any package | 13:18 |
| sabdfl | can IRC tell me last-seen, or do i need to ask a bot?> | 13:18 |
| lool | sabdfl: nickserv might tell you | 13:22 |
| lool | sabdfl: /m nickserv info sabdfl | 13:23 |
| sabdfl | thanks lool | 13:23 |
| === MacSlow|lunch is now known as MacSlow| | ||
| === MacSlow| is now known as MacSlow | ||
| serge_ | jbernard: thanks, i'll push the pkg | 13:41 |
| Amoz | dholbach, aha! found ya | 14:01 |
| Amoz | dholbach, maybe you already know this but harvest isn't showing any opportunities | 14:15 |
| maco | cjwatson: thanks! that one was scaring me :) | 14:26 |
| dholbach | Amoz, it's known - I hope we get it fixed soon | 14:28 |
| dholbach | abhinav-, yes, it's broken - I hope we get it fixed soon | 14:28 |
| === dendro-afk is now known as dendrobates | ||
| bdmurray | mvo: bug 801059 had a corrupt alternatives file added to it | 14:55 |
| ubottu | Launchpad bug 801059 in firefox (Ubuntu) "package firefox 5.0+build1+nobinonly-0ubuntu0.11.04.2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 2" [Undecided,Incomplete] https://launchpad.net/bugs/801059 | 14:55 |
| === dendrobates is now known as dendro-afk | ||
| === dendro-afk is now known as dendrobates | ||
| === ximion1 is now known as ximion | ||
| === and`_ is now known as and` | ||
| Amoz | dholbach, cool | 15:11 |
| jibel | users are reporting non working keyboard and mouse on Oneiric, bug 807306 , bug 807291, bug 807538 | 15:31 |
| ubottu | Launchpad bug 807306 in xorg (Ubuntu) "[oneiric] Keyboard & mouse not working in X" [Undecided,New] https://launchpad.net/bugs/807306 | 15:31 |
| ubottu | Launchpad bug 807291 in lightdm (Ubuntu) "Mouse and keyboard unresponsive until re-plugged in" [Undecided,New] https://launchpad.net/bugs/807291 | 15:31 |
| ubottu | Launchpad bug 807538 in linux (Ubuntu) "Keyboard and touchpad problems" [Undecided,Incomplete] https://launchpad.net/bugs/807538 | 15:31 |
| jibel | it started today, could someone have a look at it | 15:31 |
| seb128 | jibel: does downgrading the new xorg-server make it work? | 15:32 |
| seb128 | jibel: do you have the issue and can test? | 15:32 |
| jibel | seb128, no I tried, but can't reproduce with usb keyboard and mouse | 15:33 |
| seb128 | hum | 15:33 |
| seb128 | jibel: would be useful to have a list of the packages they upgraded before getting the update as well? | 15:36 |
| seb128 | update->issue | 15:36 |
| jibel | seb128, I requested more info and the history log. | 15:39 |
| seb128 | jibel: kenvandine has the issue, he says it's happening on a real box and with startx without lightdm | 15:40 |
| seb128 | cjwatson, did you do an autosync run around the time you did the manual sync round you did yesterday? | 15:41 |
| cjwatson | no | 15:42 |
| seb128 | cjwatson, I'm trying to figure if -changes has a list of all uploads around that time or not | 15:42 |
| cjwatson | we're past Debian import freeze, nobody should be doing autosyncs | 15:42 |
| seb128 | ok, so -changes probably has all what changed | 15:42 |
| cjwatson | it should do, yeah | 15:42 |
| tgardner | is it a known problem that oneiric x86en chroots cannot be installed on lucid/maverick/natty ? | 15:49 |
| dholbach | Amoz, abhinav-: fixed | 15:52 |
| abhinav- | dholbach: cool. thanks :-) | 15:52 |
| cjwatson | tgardner: the other way round is known. I hadn't heard of that way round | 16:04 |
| tgardner | cjwatson, I just tested using an oneiric host. it doesn't work there either | 16:04 |
| cjwatson | oh, you can't build an oneiric chroot at the moment anyway, the /run transition is incomplete and breaking stuff | 16:05 |
| cjwatson | mentioned it in the release meeting just now | 16:05 |
| tgardner | cjwatson, oh, is that the root of my problem ? | 16:05 |
| cjwatson | could well be | 16:05 |
| tgardner | I got busy watching the shuttle launch and forgot to pay attention to the release meeting | 16:06 |
| === vish is now known as jcastro_away | ||
| === jcastro_away is now known as vish | ||
| === jono is now known as jcastro2 | ||
| === jcastro2 is now known as jono | ||
| === beuno is now known as beuno-lunch | ||
| doko | slangasek: for the eglibc build failure: either you try to build with gcc-4.6, or you wait until gcc-4.5 is uploaded | 16:37 |
| slangasek | doko: I can wait for gcc-4.5 to be uploaded, certainly | 16:37 |
| slangasek | doko: should eglibc not be switched to use gcc-multilib, instead of gcc-4.5-multilib? | 16:38 |
| * doko didn't want to hear this ;p | 16:38 | |
| slangasek | heh | 16:38 |
| doko | no, a first eglibc build with a new major version should be checked first | 16:39 |
| slangasek | ok | 16:39 |
| === micahg_ is now known as micahg | ||
| Laney | can someone please look at sparkleshare in NEW/oneiric? It's been there some time | 16:47 |
| Laney | (and people are starting to ask me about it) | 16:47 |
| === beuno-lunch is now known as beuno | ||
| infinity | Laney: Done. | 17:16 |
| tkamppeter | pitti, hi | 17:20 |
| cr3 | is there a nice way to check if a package is installed programmatically, to be used in a shell script or a makefile? | 18:28 |
| cr3 | dpkg -l package_name | tail -n 1 | grep '^ii' doesn't seem absolutely reliable | 18:28 |
| brendand | cr3: is dpgk --get-selections not reliable? | 18:30 |
| cjwatson | I would go for something involving dpkg-query -W | 18:31 |
| cjwatson | brendand: it's the wrong question | 18:31 |
| cjwatson | --get-selections asks for the intended state, not the actual state | 18:31 |
| cr3 | cjwatson: so, if the package is installed, I get two columns, otherwise I get either one column if the package is known or the string "No packages found..." which might be localized. what about dpkg --status? | 18:34 |
| cjwatson | you can customise the output with the --showformat option - you don't need to parse out columns. read its man page | 18:35 |
| cjwatson | dpkg-query will be easier than using dpkg --status and parsing the output | 18:35 |
| cr3 | cjwatson: awesome, I will have a closer look. thanks! | 18:35 |
| cjwatson | the string "No packages found ..." goes to stderr, not stdout | 18:35 |
| cjwatson | 2>/dev/null and only look at stdout | 18:35 |
| cr3 | dpkg-query -W -f='${Status}' and what you said about stderr, I should be set | 18:36 |
| cjwatson | ${Status} is a bit fiddly because it has both desired state and actual state; I think I'd go for ${Version} being non-empty | 18:38 |
| cjwatson | hm, maybe not, that fails on packages that are removed but still have config files present | 18:38 |
| cjwatson | I guess ${Status} is the best you've got | 18:39 |
| cr3 | success: ifeq "$(shell dpkg-query -W -f='$${Status}' $package 2>/dev/null)" "install ok installed" | 18:44 |
| === daker is now known as daker_ | ||
| bambee | any idea about this strange behaviour http://paste.ubuntu.com/640306/ ? | 19:07 |
| penguin42 | echo $PATH | 19:09 |
| bambee | penguin42: already done, http://paste.ubuntu.com/640313/ <-- /usr/bin is there | 19:18 |
| bambee | apparently the symlink is dead | 19:21 |
| bambee | and not correct http://paste.ubuntu.com/640316/ (I am an amd64) | 19:22 |
| === Quintasan_ is now known as Quintasan | ||
| === yofel_ is now known as yofel | ||
| slangasek | Keybuk: hey there | 19:40 |
| slangasek | Keybuk: it looks like the base-files /run migration is incomplete compared to the Debian approach; /var/run isn't provided as a symlink. Is this deliberate for any reason, or can I patch up base-files to do the same as in Debian? | 19:42 |
| ion | Any major breakage in oneiric? Thinking of upgrading my laptop from natty. | 19:44 |
| slangasek | I'm sure there is some, though I'm typing this from it and have no specific complaints :) | 19:45 |
| slangasek | well, that's not true, I have many specific complaints but none that rise to the level of "major breakage" | 19:45 |
| slangasek | or even "I've gotten around to filing a bug" ;) | 19:46 |
| bdmurray | slangasek: I just uploaded a base-files debdiff to fix another issue | 19:49 |
| bdmurray | slangasek: and when I say uploaded I mean added to bug 744253 | 19:51 |
| ubottu | Launchpad bug 744253 in base-files (Ubuntu) "documentation deep-links in motd" [High,In progress] https://launchpad.net/bugs/744253 | 19:51 |
| === dendrobates is now known as dendro-afk | ||
| Keybuk | slangasek: umm, what base-files /run ? | 20:04 |
| slangasek | Keybuk: the one uploaded to Ubuntu yesterday with your name in the changelog? :) | 20:04 |
| bdmurray | @pilot in | 20:06 |
| === udevbot_ changed the topic of #ubuntu-devel to: Archive: open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: bdmurray | ||
| slangasek | bdmurray: do you think that change is required if we get the website fixed (which it ought to be, given that this URL is in existing releases)? | 20:06 |
| bdmurray | slangasek: yes, I think it is still a nice addition | 20:07 |
| slangasek | fair enough | 20:07 |
| bdmurray | slangasek: however I don't recall what used to be at the original url | 20:07 |
| Keybuk | slangasek: I didn't upload anything | 20:08 |
| Keybuk | in fact, I'm pretty damned sure I still can't upload | 20:08 |
| micahg | slangasek: barry uploaded that | 20:08 |
| slangasek | oh | 20:08 |
| Keybuk | due to the whole Launchpad hates my GPG key issue | 20:08 |
| slangasek | why did barry put your name on it? :) | 20:09 |
| slangasek | I saw the bzr log listed him, but thought that was some manual archive/UDD reconciliation | 20:09 |
| bdmurray | didn't he say something about it in his pilot report? | 20:09 |
| * slangasek looks | 20:09 | |
| stgraber | http://tinyurl.com/622frtw | 20:10 |
| Keybuk | I don't know, I'm not barry ;-) | 20:10 |
| stgraber | - merged base-files patch to resync archive with sjr's working branch. archive and udd branch should now be in sync. | 20:10 |
| slangasek | Keybuk: so in conclusion, I should do whatever I think is right? :) | 20:10 |
| stgraber | (from barry's e-mail to ubuntu-devel) | 20:10 |
| slangasek | stgraber: yep | 20:10 |
| Keybuk | slangasek: I think in general that is a good plan for life | 20:11 |
| slangasek | heh | 20:11 |
| Keybuk | for you, especially | 20:11 |
| stgraber | I remember that last time we talked about it in #ubuntu-devel, the discussion was that the changes barry uploaded were done before Debian started migrating to /run and that now that we're late for that change we should merge whatever they did | 20:11 |
| micahg | Keybuk: BTW, you could create a throwaway key just for uploading | 20:15 |
| slangasek | stgraber: yeah, that seems approximately correct | 20:16 |
| Keybuk | micahg: I'd have to install gpg on my Mac then ;-) | 20:16 |
| micahg | heh | 20:16 |
| Keybuk | natty really doesn't like VMs :-( | 20:21 |
| micahg | Keybuk: unity-2d works with one of the kvm drivers | 20:25 |
| Keybuk | micahg: all I see is an old-style gnome desktop with half the bits missing or not working | 20:25 |
| micahg | oh right,, classic was fallback | 20:26 |
| micahg | well, if you want unity, unity-2d it might work | 20:26 |
| * highvoltage wants to see what goes wrong in gnome 3 fallback this weekend. it works fine in debian | 20:27 | |
| === dannf is now known as dannf-lunch | ||
| Sarvatt | ion: lots of people reporting no mouse/keyboard after the post alpha-2 updates, might be something to watch out for | 20:35 |
| === _LibertyZero is now known as LibertyZero | ||
| ion | sarvatt: Thanks. https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/807306 apparently | 20:38 |
| ubottu | Ubuntu bug 807306 in xorg (Ubuntu) "[oneiric] Keyboard & mouse not working in X" [Medium,Incomplete] | 20:38 |
| ion | Oh, that’s about a VM. | 20:38 |
| Sarvatt | yep indeed, people are hitting in on real installs too though | 20:38 |
| Keybuk | micahg: tbh, I just wanted something vaguely usable for doing Ubuntu work on | 20:39 |
| ion | Sounds like something i might be able to debug. | 20:40 |
| ion | “After this operation, 8,765 MB disk space will be freed.” (Deleting a bunch of games i meant to try out but haven’t got around to. :-P) | 20:40 |
| ion | A.k.a. spring cleaning before release upgrade. | 20:41 |
| bdmurray | slangasek: speaking of base-files bug 285734 is rather interesting too | 20:43 |
| ubottu | Launchpad bug 285734 in base-files (Ubuntu) "[PATCH] In /etc/issue, tell users how to undo accidental Ctrl+Alt+F1 presses" [Undecided,New] https://launchpad.net/bugs/285734 | 20:43 |
| bdmurray | I'm particularly worried about that guy's sister | 20:43 |
| bdmurray | "OMIgawds!" | 20:43 |
| slangasek | bdmurray: who presses Ctrl+Alt+F1 by accident? :) | 20:45 |
| ScottK | People like that aren't rare. | 20:45 |
| ScottK | The same ones that hit Ctel+Alt+Backspace | 20:45 |
| ScottK | Apparently it was a lot. | 20:46 |
| bdmurray | slangasek: I seem to recall mpt mentioning it in a presentation at UDS | 20:46 |
| slangasek | hmm | 20:47 |
| slangasek | well, at the moment I'm trying to unbreak debootstrap... I don't think I'm in the right headspace to properly consider edits to /etc/issue | 20:48 |
| infinity | I'm as puzzled as others by the idea that you can ctrl-alt-f1 by accident. | 20:49 |
| * tumbleweed used to press ctrl+alt+backspace quite a lot when I was using ctrl+alt to change workspaces | 20:49 | |
| infinity | You could also end up with people, upon crashing their DM (okay, rare, but probably less rare than accidental ctrl-alt-f1) then wondering why the "text-mode's" instructions are a lie. | 20:50 |
| slangasek | crashing the dm is supposed to launch the failsafe X interface | 20:52 |
| slangasek | though I see now that /etc/init/failsafe-x.conf only handles gdm, not lightdm, hah | 20:53 |
| Sarvatt | slangasek: it's trying to exec /etc/gdm/failsafeXServer which doesn't exist too | 21:02 |
| infinity | slangasek: I like the theory and all, but there are still times when X is just so hideously hosed that you're ultimately going to end up at a terminal. Or calling tech support. | 21:02 |
| Sarvatt | looks like that's now in xdiagnose installed to /usr/share/xdiagnose/failsafeXServer | 21:03 |
| slangasek | Sarvatt: ugh | 21:03 |
| bryceh | lightdm doesn't call it anyway | 21:03 |
| infinity | slangasek: I dunno. I have no real objections to handholding in issue and motd, I just have old skool UNIX user grumpiness about issue being needlessly verbose, I suspect. ;) | 21:03 |
| bryceh | slangasek, lightdm just exits to console right now if X terminates during startup | 21:04 |
| infinity | slangasek: And one could keep it the same classic UNIX terseness with something like: "Ubuntu 10.04.2 LTS \n \l (Alt-F7 to return to GUI)" | 21:05 |
| infinity | slangasek: (And of course, even that's a blatant lie if you don't actually have X and/or something graphical on VT7) | 21:06 |
| bryceh | or if it is on vt8 | 21:06 |
| infinity | Yeah, or that. | 21:06 |
| bryceh | pidof /usr/bin/X would give you whether X is running | 21:07 |
| bryceh | ps a | grep X shows it on tty7, so seems like would be straightforward to figure out where it is too | 21:07 |
| infinity | Yeah, but /etc/issue is just a text file. | 21:08 |
| broder | ConsoleKit also knows | 21:08 |
| infinity | So, this goes from a wishlist "update a text file" bug to something nasty involving hacking getty. | 21:08 |
| infinity | Which still doesn't work, since getty outputs to terminals when it spawns, not when you switch to them. :P | 21:09 |
| broder | infinity: one option would be something along the lines of "press enter to switch back to your session" or whatever | 21:09 |
| broder | instead of outputting a specific tty to switch to | 21:09 |
| infinity | You'd have to find a key sequence to trap that doesn't break the brains of every console user in the world. | 21:11 |
| infinity | (for instance, the first thing I do when I connect a serial terminal is hammer enter a few times to get a prompt) | 21:11 |
| infinity | Though, obviously, our slightly-odd getty codepath would only be running on VTs. | 21:12 |
| infinity | I dunno. I think we've just proven the bug's a bit deeper than a text file addition, regardless. | 21:12 |
| bryceh | finally, a use for CAPSLOCK | 21:12 |
| broder | sure, i do the same with enter. the potentially simpler solution would be to just use enter anyway, but include in the "please press enter" printout instructions on how to disable that behavior | 21:13 |
| infinity | "To return to your graphical session, hit left, right, left, right, up, down, A, B, CAPS" | 21:13 |
| james_w | just make changing VT a sysreq combo | 21:13 |
| bryceh | if no user is logged in and it's just a login prompt, there's probably a set of keys which have no purpose which could be used. E.g. Home or Esc maybe | 21:13 |
| infinity | james_w: And how long until people complain that they can hit sysrq by accident? | 21:14 |
| infinity | james_w: Seems about as likely as ctrl-alt-f1, to be fair. | 21:14 |
| slangasek | not only is /etc/issue just a text file, it's output to the console in parallel to X startup | 21:14 |
| slangasek | right, infinity said that already | 21:14 |
| infinity | slangasek: Before, generally. | 21:14 |
| james_w | infinity, that's what was done for ctrl-alt-backspace | 21:14 |
| james_w | I haven't heard complaints | 21:14 |
| bryceh | I have ;-) | 21:14 |
| james_w | it was only a semi-serious suggestion anyway | 21:14 |
| infinity | james_w: I switch terminals a lot more than I need to kill X. | 21:15 |
| james_w | heh | 21:15 |
| james_w | complaints from people others that X developers :-) | 21:15 |
| infinity | (And, sometimes I switch terminals TO kill X, since ctrl-alt-bkspc no longer works out of the box) | 21:15 |
| bryceh | ;-) | 21:15 |
| slangasek | infinity: usually after :) | 21:15 |
| slangasek | infinity: we try to bring the X server up ASAP; we wait until rc is done before spawning getties | 21:16 |
| infinity | slangasek: Oh, right. Start on stopped rc, I forgot about that change. | 21:16 |
| slangasek | at least for tty1 | 21:16 |
| bryceh | c-a-b angst recently spotted @ http://linuxlock.blogspot.com/2011/06/canonical-alienates-their-major-asset.html | 21:16 |
| infinity | slangasek: Stil, close enough to parallel to not really allow you to do silly things like this. | 21:16 |
| infinity | bryceh: I can't be bothered to read to article, but I love that the URI implies that our "major asset" is people who like to kill X servers. | 21:17 |
| infinity | s/read to/read the/ | 21:17 |
| infinity | Besides, if he wants the old functionality back, all he has to do is install unity. | 21:18 |
| infinity | It's less predictable, but just as reliable. | 21:18 |
| bryceh | infinity, 'The first threads came apart when decisions were made to remove certain keyboard shortcuts, "for the good of the new user". Ctrl/alt/delete and ctrl/alt/backspace were removed.' | 21:18 |
| infinity | Ctrl-Alt-Del isn't removed... | 21:19 |
| bryceh | plus it was upstream that removed the shortcut. But why spoil a rant with messy facts. | 21:20 |
| infinity | That sort of proves his Ubuntu = Linux side-rant. | 21:20 |
| micahg | bryceh: it's in the comments | 21:20 |
| bryceh | micahg, yep | 21:20 |
| slangasek | ah, this is the HeliOS guy | 21:21 |
| bryceh | but in terms of actual ubuntu users, james_w is right, we haven't gotten bug reports about it. Seems a total non-issue with our users. | 21:25 |
| james_w | I was commenting more about the lack of complaints that sysreq was too easy to hit | 21:25 |
| james_w | I realise that there were complaints about taking away ctrl-alt-backspace | 21:25 |
| bryceh | james_w, ah, also true | 21:25 |
| infinity | james_w: I'll file a bug. | 21:25 |
| james_w | or rather hiding it in the keyboard configuration | 21:26 |
| infinity | james_w: My 1 Sysrq bug will match the 1 Ctrl-Alt-F1 bug, and we can have a showdown. ;) | 21:26 |
| slangasek | bryceh: I don't think bug reports is an accurate measure of whether people are bothered by the Ctrl+Alt+BkSp change | 21:26 |
| maco | bryceh: did the dude JUST realise c-a-b is non-default *now*? | 21:26 |
| slangasek | I expect people who use Ctrl+Alt+BkSp to also be more likely to research issues before filing new bug reports | 21:27 |
| bryceh | maco, I assume he's been holding it in until now ;-) | 21:27 |
| bryceh | slangasek, I like the world you live in, with sane bug reporters, can I join? :-) | 21:28 |
| slangasek | bryceh: well see, that just goes to show that the people you're getting bug reports aren't the ones who use c-a-b! :) | 21:29 |
| infinity | bryceh: It's easy, you just filter your bugmail and only read incoming mail from the 4 people you like. | 21:29 |
| bryceh | anyway, from what I understand of how c-a-b works, it likely isn't going to give the relief that it used to, since the freezy bits have moved from X to the kernel, and c-a-b won't help there | 21:29 |
| bryceh | it will help if a client goes seriously out of control and locks up X that way, but that seems to be thankfully rare in practice | 21:31 |
| bdmurray | slangasek: I seem to recall a guideline for patch numbering like 33-zyx.patch | 21:48 |
| bdmurray | slangasek: is there one? | 21:48 |
| === dendro-afk is now known as dendrobates | ||
| infinity | bdmurray: Patch numbering made more sense with systems like simple-patchsys where they were applied in filesystem order. | 21:49 |
| === Ursinha is now known as Ursinha-afk | ||
| infinity | bdmurray: But some large/complex source packages still like to codify meaning in numbers (like 000->500 = to/from upstream, 900+ = local only, never being forwarded, etc) | 21:50 |
| infinity | bdmurray: With people (in some cases, slowly) moving to source-format 3.0 and quilt, though, the filesystem ordering is meaningless, and even the coded meaning is only marginally useful, since they're all meant to have headers that don't suck. | 21:51 |
| bdmurray | ah okay thanks | 21:52 |
| slangasek | yeah, I don't know of any guideline for patch numbering | 21:56 |
| slangasek | there's a guideline for headers within the patches | 21:57 |
| infinity | Does dpkg-dev, devtools, or quilt, have a utility to list patches with consice header info? I haven't looked. | 21:58 |
| infinity | concise too. | 21:58 |
| infinity | devscripts, even. It's clearly Friday afternoon. | 21:59 |
| bdrung_ | infinity: it's Saturday. :p | 22:12 |
| === dannf-lunch is now known as dannf | ||
| micahg | bdrung_: the debian package doesn't include the entire multiarch diff as specified in the 1.4.6-5 changelog, is that not a problem? | 22:36 |
| micahg | bdrung_: sorry, libgcrypt11 ^^ | 22:37 |
| bdrung_ | micahg: multiarching it was done differently | 22:38 |
| bdrung_ | micahg: the result should be the same | 22:38 |
| bdmurray | @pilot out | 22:51 |
| === udevbot_ changed the topic of #ubuntu-devel to: Archive: open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: | ||
| === dendrobates is now known as dendro-afk | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!