[00:07] <bryce> LaserJock: not that I know of
[00:07] <bryce> LaserJock: patches welcomed tho
[00:11] <wgrant> Shouldn't xorg.conf be replaced if it hasn't been modified?
[00:34] <bryce> wgrant: well LaserJock's talking about people who've customized their xorg.conf's to set up input devices
[00:35] <wgrant> bryce: Ah. Maybe we need some XKit magic to remove the default device entries?
[00:35] <wgrant> Actually, why doesn't it work?
[00:36] <bryce> wgrant: i-h/-evdev steals all the devices and ignores xorg.conf settings
[00:36] <wgrant> Exactly.
[00:36] <wgrant> But having an entry for the touchpad breaks things.
[00:36] <bryce> https://wiki.ubuntu.com/IntrepidReleaseNotes - see X.org Input Devices
[00:36] <wgrant> So it's obviously not entirely ignored.
[00:40] <jcristau> the synaptics from xorg.conf should steal the device before the hotplugged one
[00:43] <tseliot> bryce, wgrant: maybe we could make a backup of xorg.conf files and remove inputdevice sections and serverlayout sections with XKit.
[00:43] <wgrant> tseliot: That's what I was thinking.
[00:43] <wgrant> Hmmm.
[00:43] <wgrant> Shouldn't we just remove the default inputdevices and remove their references from serverlayout?
[00:44] <wgrant> Removing custom inputdevices and custom serverlayouts sounds like a recipe for irate users.
[00:45] <bryce> commenting them out would be preferable to removing them
[00:45] <wgrant> jcristau: If it's stealing the device first, why doesn't the entry in https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-input-synaptics/+bug/214252/comments/16 work?
[00:45] <tseliot> wgrant: yes, I agree with you. We'll need to plan (and then test) this carefully
[00:46] <wgrant> We haven't got long, though...
[00:47] <tseliot> bryce: currently XKit doesn't allow to do comment things out but having a backup file, say, xorg.conf_dist-upgrade, should be enough
[00:47] <bryce> hmm
[00:48] <tseliot> wgrant: it's 01:47 AM here and by carefully I mean when I'm not half asleep ;)
[00:48]  * wgrant is glad this can be done in the dist-upgrader, and doesn't have to be a postinst somewhere.
[00:48] <nxvl> slangasek: did you accepted lrm alreadu?
[00:48] <nxvl> already*
[00:48] <wgrant> tseliot: Heh.
[00:49] <tseliot> bryce, wgrant: ok, we can talk about this maybe via email, and notify mvo
[00:49] <bryce> sounds good
[00:49] <slangasek> nxvl: yes
[00:50] <jcristau> wgrant: that bug doesn't have a recent log, so, dunno.
[00:50] <tseliot> ok then, good night
[00:50] <nxvl> slangasek: \o/
[00:50] <nxvl> slangasek: then i can update in the morning?
[00:50] <slangasek> nxvl: that depends on when linux-meta is uploaded, you'd have to ask the kernel team on that
[00:51] <wgrant> jcristau: The recent comments at the end seem to be unrelated to the rest of the bug.
[00:51] <slangasek> nxvl: well, or you can upgrade by hand, yes
[00:51] <nxvl> slangasek: yes, but there is still no linux-restricted-modules-2.6.27-5
[00:54] <slangasek> nxvl: there is; though it may not be on your mirror yet
[00:55] <nxvl> right, i'm using pe.archive
[00:55] <nxvl> that might be
[01:09] <nxvl> slangasek: Bug #256459
[01:10] <nxvl> slangasek: can you confirm that it is on debian-installer actually?
[01:14] <slangasek> nxvl: 'partman' is a more exact starting point, but 'debian-installer' is also ok
[01:15] <nxvl> slangasek: ok, i will look at partman, thank you
[01:23] <LaserJock> wgrant: sorry, was afk
[01:23] <LaserJock> wgrant: my problem is that I had my touchpad set up in xorg.conf and that got overridden by HAL so I wrote an fdi file
[01:23] <LaserJock> *now* xinput overrides HAL so my fdi file doesn't work
[01:24] <LaserJock> so in the end I've moved through 3 different configuration mechanism to configure my touchpad
[01:24] <wgrant> LaserJock: What do you mean "xinput overrides HAL"?
[01:24] <LaserJock> wgrant: meaning, xinput settings override my HAL settings
[01:24] <LaserJock> so I had my fdi file all set up but it didn't have any effect
[01:25] <LaserJock> so their seems to be a hierarchy, xinput > HAL > xorg.conf
[01:25] <wgrant> LaserJock: You mean you were trying to override some of the ~4 XI properties that gnome-settings-daemon now sets?
[01:25] <wgrant> Right.
[01:26] <LaserJock> wgrant: I'm not sure what's responsible
[01:26] <wgrant> Which were you trying to set?
[01:26] <LaserJock> I just know what works and then it stops working
[01:26] <LaserJock> turn off the tap-to-click and vertical scrolling
[01:26] <wgrant> Right, gnome-settings-daemon overrides those.
[01:26] <wgrant> You want System->Preferences->Mouse->Touchpad.
[01:27] <LaserJock> sure, I figured that out eventually
[01:27] <LaserJock> but I had a fdi file that worked, and previously I had an xorg.conf that worked
[01:28] <LaserJock> I think users are gonna have a bit of a shock if they've, like me, done any configuration of their touchpad
[01:28] <wgrant> If people are delving into their xorg.confs and fdi files, I think we can probably expect them to be able to work this out...
[01:28] <LaserJock> why?
[01:28] <wgrant> fdi files only just started existing, so this doesn't affect upgrades.
[01:28] <LaserJock> no, but it's in the wiki
[01:28] <LaserJock> people are going to find it and file bugs as to why it doesn't work right
[01:28] <wgrant> LaserJock: Where in the wiki?
[01:28] <LaserJock> and existing xorg.conf files not working is a bigger concern
[01:28] <LaserJock> X/Config
[01:29] <wgrant> Right, the xorg.conf issue needs to be worked out.
[01:29] <wgrant> But I don't see fdi files as a particularly significant issue.
[01:32]  * wgrant fixes the wiki page to mention that gnome-settings-daemon overrides a few things.l
[01:39] <LaserJock> well, I'm glad to not need the fdi file anyway :-)
[01:40] <LaserJock> but there are a lot of good documentation on help.ubuntu.com that will need updating
[01:40] <wgrant> We certainly do need better docs.
[01:40] <LaserJock> has anybody talked to the Doc Team about this change?
[01:40] <mrooney> TheMuso: ping?
[01:43] <slangasek> superm1: ping
[01:44] <mrooney> wgrant: well, I'll boot back in Hardy so I can get some things done, and try later, thanks for your help!
[01:45] <wgrant> mrooney: np
[02:03] <slangasek> superm1: contentful ping: if you have a chance, I'd appreciate a review of https://wiki.ubuntu.com/IntrepidReleaseNotes#nVidia%20%22legacy%22%20video%20support for accuracy
[02:03] <wgrant> slangasek: Who grants UI freeze exceptions?
[02:04] <slangasek> wgrant: the release team who grants other freeze exceptions for the component in question
[02:04] <wgrant> slangasek: OK, so you count for main?
[02:05] <slangasek> yes :)
[02:06] <wgrant> slangasek: Given that we're breaking everybody's custom Synaptics touchpad configuration, I'd like to ensure that the most customised options are exposed through the GUI - the two I'd like to add are horizontal and vertical two-finger scrolling. That'd just mean two extra checkboxes in the Touchpad tab of gnome-mouse-properties, and a few extra lines of code there and in g-s-d... what are my chances?
[02:07] <slangasek> wgrant: chances are excellent; the UI freeze is mostly to prevent gratuitous changes that piss off the doc folks, not to get in the way of fixing major UI blunders
[02:10] <wgrant> slangasek: Not really a UI blunder, but sounds good...
[02:12] <wgrant> Well, I guess forcing people to config things through xorg.conf is a UI blunder.
[02:12] <slangasek> that's my view :-)
[02:16] <wgrant> slangasek: What's the process for officially getting an exception?
[02:22] <wgrant> Oh. I didn't see that it was on FreezeExceptionProcess. Sorry.
[03:03] <bryce> slangasek: ok, -ati git snapshot packaged, tested locally, and uploaded.
[03:42] <superm1> slangasek, Isn't the migration to "nv" going to be automatic for those users too though?
[03:42] <superm1> like how you mentioned in the -ati one
[03:45] <superm1> wgrant, is it possible too for a checkbox for circular scrolling for touchpads that support it?
[03:46] <superm1> i'm not sure how much is in the synaptics driver - but i know i've got some hardware that supports it at least
[03:49] <wgrant> superm1: It's entirely in the driver.
[03:49] <wgrant> superm1: I'd like to add it, and I'm working on that code now, but the hard bit is taking input for the start position.
[03:49] <wgrant> A guess a set of radio buttons would work.
[03:50] <wgrant> The touchpad device is really pretty stupid - it just reports touch position, pressure, sometimes number of fingers, and sometimes width.
[03:50] <superm1> wgrant, ah i wasn't aware it was all software supported
[03:51] <wgrant> superm1: Any other options that you'd like exposed?
[03:51] <superm1> wgrant, what else is there?  I just knew about circular scrolling, horiz and vert scrolling
[03:52] <wgrant> superm1: xinput list-props "SynPS/2 Synaptics TouchPad"
[03:52] <wgrant> There's a lot.
[03:52] <wgrant> Much of it probably shouldn't be exposed, but more of it should be eventually.
[03:53] <superm1> hum, for me unable to find device SynPS/2 Synaptics TouchPad
[03:53] <wgrant> xinput list
[03:53] <superm1> oh on this machine it's ALPS
[03:53] <wgrant> Some types have different names.
[03:53] <wgrant> Ah.
[03:54] <superm1> wgrant, two finger scrolling would be nice
[03:54] <superm1> i've seen some mac folks who have had that
[03:54] <wgrant> superm1: That's what I'm implementing now.
[03:54] <wgrant> A lot of people request it.
[03:54] <superm1> wgrant, cool :)
[03:55] <bcurtiswx> anyone here understand Oct  3 15:03:52 weather kernel: Inspecting /boot/System.map-2.6.27-4-generic
[03:55] <bcurtiswx> Oct  3 15:03:52 weather kernel: Cannot find map file.
[03:55] <bcurtiswx> is that a bad thing?
[07:08] <slangasek> superm1: I don't know yet; I've asked that question in the bug, but I'm not writing that in the release notes until it's implemented
[07:18] <slangasek> superm1: did it look accurate, otherwise, in terms of the chipsets claimed to be affected?
[12:25] <saurabh> i think the global menu patch is a bad idea, normally we access the menu using the ALT key, so why not just hide the menu and then show it after the user presses the ALT key?, this saves vertical screen space as well
[12:27] <ion_> What patch?
[12:27] <Hobbsee> i expect you end up with teh ie7 situation then
[12:27] <Hobbsee> a whole bunch of users going "what the hell have you done with my menu?!?!?!"
[12:28] <saurabh> Hobbsee: we can make it optional then
[12:28] <Hobbsee> saurabh: even so.
[12:29] <wgrant> "global menu patch"?
[12:29] <saurabh> wgrant: yes
[12:30] <wgrant> Not quite the response I was wanting.
[12:30] <ion_> Ditto
[12:30] <saurabh> i just need the ability to hide the menu for every application
[12:31] <wgrant> What is the global menu patch, and why is it a bad idea?
[12:32] <saurabh> wgrant: you get mac like global menu functionality
[12:33] <wgrant> saurabh: Ah.
[12:33] <digistyl3> hi TheMuso, are you online?
[12:33]  * wgrant doesn't normally access menus using Alt.
[12:37] <digistyl3> hey everyone, i seem to have a strange problem, and i think it's related to the intel driver (i use an intel x3100)
[12:37] <digistyl3> when i use virtualbox
[12:37] <digistyl3> the screen contents sometimes freeze
[12:37] <digistyl3> they only refresh when i move the window
[12:37] <digistyl3> it happens with me in dosbox too
[12:37] <digistyl3> i don't have the fancy desktop effects enabled
[12:38] <digistyl3> how can i debug this?
[12:40] <digistyl3> it also happens with flash
[12:40] <digistyl3> if i watch youtube videos
[12:41] <digistyl3> the image in the video freezes, i have to scroll for it to continue updating the contents
[12:54] <digistyl3> oh well, i just found two bugs about this and joined them, people thought it was a flashplugin-nonfree bug https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/275285
[12:58] <ogra> thats more likely a driver prob
[12:58] <ogra> digistyl3, you should ask for X driver info
[12:58] <digistyl3> ogra: someone said that he's using nvidia graphics
[12:59] <ogra> yeah, i see two people mentionint intel though
[12:59] <ogra> *mentioning
[13:05] <digistyl3> ogra: i posted my xserver-xorg and xserver-xorg-video-intel version numbers
[13:06] <ogra> good
[13:06] <digistyl3> you don't think the bug is in the kernel, do you?
[13:06] <ogra> no idea, but it sounds like an xorg driver or compiz bug
[13:07] <digistyl3> ogra: i have visual effects set to "None", so i think compiz is out of the picture here
[13:24]  * mcasadevall pokes ScottK 
[14:10] <siretart> what does '夠荒唐' mean? (keybuk's quit message)
[14:11] <Spads> siretart: it's a chinese phrase allegedly meaning "to hell with it" (or equivalent) that was used in the dialogue in Firefly
[14:11] <realist> I'm not sure, my terminal only displayed one of those glyphs
[14:12]  * realist wonders how to fix that
[14:12] <Spads> realist: I installed ttf fonts until everything appeared: http://pastebin.ubuntu.com/23131/ (as of July)
[14:13] <Spads> \〠/ <-- eventually you'll be able to see the postman there
[14:14] <realist> I'm laking some fonts, it appears
[14:16] <Spads> realist: well to be fair, I doubt most of those fonts are relevant to your locale, so it makes sense that they aren't installed by default.  Still, I'd love a ttf-lots-and-lots-please metapackage of some sort
[14:16] <StevenK> ttf-every-font-and-then-some
[14:19] <Spads> yeah precisely
[14:22] <realist> I'm about to try something like; apt-get install `apt-search ttf |awk '/^ttf/{print $1}'`
[14:23] <Spads> yeah I've done something similar
[14:23] <realist> ^search^cache search^
[14:23] <Spads> you have to grep out conflicting ones
[14:23]  * realist nods
[14:25] <realist> 359MB, maybe not
[14:40] <lfaraone> How would I go forward with a bug like bug 277302 that has a bzr branch attached and works (IMHO), where ubuntu is upstream?
[14:41] <lfaraone> (correction, it's filed against networkmanager and is triaged as medium)
[14:50] <siretart> Spads: ah, thanks
[16:44] <exco> to chroot into a broken 8.10 - do I need an 8.10 live cd or should it also work with an 8.04 one?
[16:50] <siretart> exco: please use #ubuntu for end-user support
[16:51] <exco> ok thanks, siretart
[16:59] <hspaans> g'day all
[17:00] <hspaans> now Ubuntu 8.10 is in beta, all packages have been landed? this because X sometimes freezes up
[19:07] <EvanCarroll> Is there a unannounced plan for regression in Xorg to 7.3 in the event ATI doesn't update fglrx in time for release?
[19:07] <EvanCarroll> or is 8.10 really at risk for being pushed onto the shelf without fglrx?
[19:08] <cjwatson> the latter; reverting X is pretty much off the table at this point
[19:08] <EvanCarroll> wow, that amazing.
[19:08] <cjwatson> we may be forced to do a post-release update of fglrx, although I don't like that option
[19:09] <tseliot> currently we are migrating installations to the open source driver with Update Manager + X-Kit
[19:09] <EvanCarroll> the failure of ATI to get fglrx out is going to have massive blowback for the adoption of the distro if it goes out like this.
[19:10] <Mirv> EvanCarroll: the open source driver should nowadays support good 2D resolutions out of the box on even the X3000/X4000 series, so ordinary people don't notice until they start to want some fancy 3D (at which point there might be an update available already)
[19:10] <Mirv> the main problem before was that they couldn't even start Ubuntu without fglrx, nowadays luckily that's not a problem
[19:11] <EvanCarroll> even the opensource driver (radeonhd i presume) don't support HDMI, and the ATI driver doesn't support sound on HDMI, and the conversion from fglrx to ati is actually apearing to draw outside of the size on my screen.. But I could be wrong - I'm looking into that now to file a bug
[19:11] <Mirv> (since ATI has failed to release drivers supporting latest X.org before, too)
[19:12] <EvanCarroll> yea, I'll definitly speak with my money on my future purchases I'm afraid new people aren't smart enough to point a blame finger and the buck will stop at the distro
[19:12] <tseliot> EvanCarroll: we use the "ati" driver (not "radeonhd"). Please file a bug report
[19:14] <Mirv> EvanCarroll: well NVIDIA has equally lacked support for latest X.org releases, so currently with external gfx cards you choose wrong anyway :) ATI's going to better though, while NVIDIA probably won't.
[19:14] <Mirv> EvanCarroll: anyway, the point is that if users install intrepid in eg. November, they might have fglrx install option as soon as they boot the installed system
[19:24] <EvanCarroll> oh well, at least we can see the bumps in the road up ahead. I just wonder if a system that warns early adopters that this will probably not live up their expecatations is called for prior to the update and download of packages. IE some sort of outstanding "critical" issues syndication that integrates with the libapt in a post-calculating-ugrade process
[19:27] <EvanCarroll> since this will certaintly at the very least break all of the new HTPCs running on the ati chipset
[19:32] <slangasek> EvanCarroll: the issue will be documented in the release notes, which will be displayed in update-manager when you click 'upgrade'; except that it looks like we're currently showing the wrong document, so I'll dig into that
[19:33] <EvanCarroll> Is that a feature of update-manager of an extention to libapt?
[19:33] <slangasek> update-manager, which is the supported upgrade path between Ubuntu releases
[19:34] <EvanCarroll> interesting, I've always just used apt-get dist-upgrade and figured they had the same functionality
[19:35] <Mirv> EvanCarroll: read the documentation ;)
[19:36] <slangasek> they don't; in addition to displaying the release notes, update-manager implements a number of quirks that can't be implemented sanely within apt, to handle specific upgrade issues between releases
[19:36] <sebner> slangasek: aloha. did you notice my nfdump debdiff (since you are subscribed)?
[19:37] <slangasek> sebner: hmm, apparently not; sorry, it may have been lost in the wash of beta-related mail
[19:38] <sebner> slangasek: nvm :) Still interested or should I subscribe u-u-s?
[19:38] <EvanCarroll> slangasek: thanks for the info, Mirv not mentioned in man pages of update-manager, in fact it is called a "front-end into the apt package management system" -- which almost explicitly says, "I do nothing special"
[19:38] <sebner> sladen: debdiff is reallllllllllllllllly small though  ^^⁺
[19:38] <EvanCarroll> my mental model just lead me to believe update-manager was aptitude or dselect or somtehing ported to GTK
[19:39] <EvanCarroll> and that they were all front-ends to libapt
[19:41]  * EvanCarroll reads the code to update-manager
[19:41] <slangasek> cjwatson: seems to be some memory fuzziness on my part; I thought update-manager would link directly to the release notes, but it seems to only link to dists/$distseries/main/dist-upgrader-all/current/ReleaseAnnouncement - am I thinking of ubiquity?
[19:45] <IntuitiveNipple> What's the policy when an existing patch is incomplete? correct the patch, remove it and add a full patch, or add a patch that does the 'diff' between existing and 'full' ?
[19:56] <slangasek> sebner: uploaded, thanks
[19:57] <sebner> slangasek: np, thx to you :)
[19:59] <EvanCarroll> Mirv: yea, there are some things it does, like checks symlinks to /usr/bin/python
[19:59] <EvanCarroll> and that you don't upgrade more than one leap at a time
[20:01] <geser> IntuitiveNipple: I guess there is no policy but commonly the patch gets fixed (e.g. a new version gets attached to the bug)
[20:01] <EvanCarroll> it seems to be like it would be advantagous to move this stuff into a libapt extention
[20:04] <IntuitiveNipple> geser: I'm repackaging it so I'll just regen the patch then
[20:16] <EvanCarroll> Mirv: what is the supported upgrade method if you don't have X? like with ubuntu-server?
[20:16] <ion_> do-dist-upgrade
[20:16] <ion_> do-release-upgrade, that is.
[20:17] <EvanCarroll> interesting. =)
[20:27] <Mirv> EvanCarroll: the documentation is http://www.ubuntu.com/getubuntu/upgrading - of course man pages could be good too, but if one knows man pages etc. one should possibly be able to fix any small breakages apg-get dist-upgrade might cause
[20:33] <EvanCarroll> I tend to agree, the stuff it does is fairly minor but it does a lot of fairly minor stuff, even reseting themes if they are being used and known to cause problems
[20:34] <EvanCarroll> some of the stuff it does seems kind of fishy: like executing random scripts in the local directory, and making the assumption that it is being run under sudo
[20:41] <slangasek> pitti, tjaalton, bryce: in https://wiki.ubuntu.com/IntrepidReleaseNotes we say that some keys may misbehave in X if you have the wrong keyboard model set.  Is this really going to be a problem on upgrade?  If it is, shouldn't we be doing something better to detect the keyboard model on upgrade?
[20:43] <cjwatson> slangasek: ubiquity links to the online one; I've never looked at what update-manager does
[20:44] <slangasek> ok
[20:44] <cjwatson> EvanCarroll: update-manager was conceived as executable release notes
[20:44] <cjwatson> at least that was how I put it and I liked the phrase, I don't know if it caught on :)
[20:44] <cjwatson> EvanCarroll: release notes often include all sorts of random little stuff, and update-manager is no different really
[20:53] <tseliot> slangasek: as regards the problems caused by InputDevice entries for the mouse and keyboard in the xorg.conf, we were discussing with bryce and wgrant about letting Update Manager remove these entries with X-Kit
[20:55] <tseliot> slangasek: if you were referring to this problem in your last question.
[20:56] <slangasek> tseliot: I don't know that this has any relation to what is shown as the keyboard type in the GNOME interface
[20:56] <slangasek> tseliot: btw, I had questions regarding your follow-up to bug #251107
[20:57] <tseliot> slangasek: ok, I guess I misunderstood your question then ;)
[20:57] <slangasek> tseliot: aren't these nvidia drivers shipped in binary-only form?  If so, isn't it a *bug* if the ABI declaration is not hard-coded?
[20:58] <NCommander> slangasek, no, the NVIDIA drivers have a small LGPL glue code and then the blob
[20:58] <tseliot> slangasek: only a part is a binary
[20:58] <NCommander> (its how they run around releasing closed blob drivers)
[20:59] <tseliot> slangasek: and of course the part which is affected by the problem described in the bug report is the binary one
[21:02] <slangasek> hum - so the X driver includes an LGPL shim?
[21:03] <slangasek> still seems buggy to me to auto-increment the ABI, since clearly the binary bit can also break it
[21:06] <NCommander> Anyone here a firefox guru?
[21:07] <IntuitiveNipple> NCommander: I became a bit of one recently, fixing up xulrunner1.9, but I don't guarantee anything :)
[21:09] <NCommander> IntuitiveNipple, I think I found a problem with the firefox alternatives file
[21:10] <IntuitiveNipple> NCommander: ah ok. I was more focused on internal xulrunner core code, although I did do some stuff with incorrect flashplugin paths a while back.
[21:10] <NCommander> IntuitiveNipple, can you tell me if firefox-3.0 works for you
[21:10] <NCommander> (specifically that command)
[21:11] <IntuitiveNipple> Not for intrepid, no. I'd have to run up a VM and since the daily on Friday it fails to run the CD installer for me
[21:13] <tseliot> slangasek: I see your point however, if that kind of breakage happens, I assure you that I will notice it (as Xorg won't start). It is more likely that I forget to bump the ABI.
[21:14] <slangasek> tseliot: well, that seems to assume no one else ever does a package rebuild for other reasons without having the hardware to test with
[21:14] <slangasek> NCommander: er, check the bug list?
[21:15] <slangasek> bug #275410
[21:15] <tseliot> slangasek: well, in my opinion you shouldn't package new upstream versions which you can't test
[21:16] <tseliot> it's a bit risky
[21:16] <NCommander> Hrm, this was a bug with x-www-browser
[21:16] <slangasek> tseliot: I'm not talking about packaging new upstream versions
[21:16] <tseliot> slangasek: what's the use case then?
[21:17] <NCommander> slangasek, should a release-blocker bug be considered critical or high
[21:17] <slangasek> tseliot: anyone doing archive-wide fixing of packaging bugs
[21:17]  * NCommander can't remember :-/
[21:17] <slangasek> NCommander: release-blocker status is unrelated to the distinction between critical and high
[21:17] <NCommander> ok
[21:17] <slangasek> https://wiki.ubuntu.com/RCBugTargetting
[21:17] <NCommander> slangasek, this bug completely breaks the Xubuntu Xfce help system
[21:18] <NCommander> Thank you
[21:18] <slangasek> well, and?  I've already marked the bug as release-critical...
[21:19] <tseliot> slangasek: I hope that such a thing doesn't happen without talking to the X-team first
[21:19] <slangasek> Maintainer: Ubuntu Core Developers <ubuntu-devel-discuss@lists.ubuntu.com>
[21:20] <slangasek> tseliot: ^^ that's the maintainer field on nvidia-graphics-drivers-71, which implies any core-dev can expect to reasonably upload
[21:20] <slangasek> NCommander: you're right that the bug title is incomplete; fixing, thanks
[21:20] <NCommander> I'm adding my two cents
[21:20] <NCommander> (a second script /usr/bin/firefox-3.0 is also broken)
[21:21] <slangasek> that's not a second script
[21:21] <tseliot> slangasek: that doesn't look good to me. Just my 2 cents
[21:21] <perlsyntax> Does anyone know how to do raw socket in unbuntu?
[21:21] <slangasek> tseliot: then apparently someone should change it? :)
[21:22] <perlsyntax> how do i programming in main root?
[21:22] <slangasek> tseliot: my understanding of current practice is: if the maintainer field points to a specific person or team, talk to them first before uploading.  If not, you can upload without coordination.
[21:22] <slangasek> perlsyntax: this is not a help channel
[21:23] <perlsyntax> do i sudo bash to do that?
[21:23] <perlsyntax> ?
[21:23] <tseliot> slangasek: good point. Maybe we could use something like this team: https://launchpad.net/~ubuntu-x-swat
[21:23] <perlsyntax> hello?
[21:23] <trenton_> Hello all, trying to build Amarok2 from trunk but getting mysql error on kubuntu hardy. needs libmysqld.a which package do I need please?
[21:24] <tseliot> slangasek: suggestions are always welcome ;)
[21:24] <perlsyntax> slangasek, any ideas?
[21:24] <slangasek> perlsyntax: I said, this is not a help channel.  Try #ubuntu, or a help channel specific to your programming language.
[21:24] <perlsyntax> i did they told me to come here
[21:24] <slangasek> they told you wrong
[21:24] <NCommander> slangasek, have you done any work as cooking a patch for this bug, or shall I try to do one myself
[21:27] <tjaalton> slangasek: it shouldn't matter anymore, so that part can be dropped from the release notes
[21:27] <slangasek> tjaalton: great, thanks
[21:28] <slangasek> tseliot: ~ubuntu-x-swat looks like a pretty good fit for that, yeah
[21:28] <EvanCarroll> I think I was just affected by a keybinding bug my caps lock just randomly got remapped to caps lock
[21:28] <NCommander> slangasek, I figured out what I need to patch to in the firefox-3.0 source package. I'm going to have it checked for the appname (i.e., if it tries to call gnome-sensible-browser, it will try that first, then fall back to $LIBDIR if not found)
[21:28] <NCommander> slangasek, any objections?
[21:29] <tseliot> slangasek: ok, thanks for spotting the problem and for your suggestion
[21:29] <slangasek> NCommander: yes, I object because I outlined the correct fix in my follow-up to the bug
[21:29]  * NCommander rereads steve's bug
[21:30] <NCommander> slangasek, isn't that in effect what I just suggested? Check for the name it was called by, and if it doesn't exist, call the always known to be there firefox binary?
[21:31] <NCommander> or I said my fix would be
[21:31] <slangasek> NCommander: no, you said something about "falling back to $LIBDIR"...
[21:32] <NCommander> In the script, it tries to call $LIBDIR/$APPNAME
[21:32] <NCommander> Which works out to be /usr/lib/firefox-3.0.3/$APPNAME
[21:32] <NCommander> So if I call it as x-www-browser, it should try /usr/lib/firefox-3.0.3/x-www-browser before falling back to firefox, unless I misunderstood you
[21:35] <nxvl> cjwatson: can you please process Bug #274173?
[21:35] <hansin> Anyone here heard on any install issues with the beta alternate CD for i386?
[21:35] <slangasek> it should try /usr/lib/firefox-3.0.3/x-www-browser; then it should do NAME=$(readlink $0); APPNAME=$(basename $NAME); and check for that; and it should iterate until either readlink says it's not a link, or NAME = /usr/lib/firefox-3.0.3/firefox.sh, and /then/ fallback
[21:36] <slangasek> NCommander: ^^
[21:37] <NCommander> slangasek, ah. I wasn't quite familar with readlink's behavior, but that makes perfect sense.
[21:37]  * NCommander attempts to recover the tomes of shell script programming in my head
[21:58] <NCommander> slangasek, maybe I'm missing something, but is there a reason that you can't simply call the symlink directly?
[21:59] <NCommander> Since you can do -L $LINK to see if its valid, and if not, drop to firefox
[22:05] <NCommander> ^^- slangasek?
[22:18] <NCommander> slangasek, http://pastebin.ca/1219253 - this is what my current tweak to the firefox.sh script looks like
[22:18] <cjwatson> nxvl: it's a weekend and I have to get up early tomorrow; I'm off to bed
[22:19] <cjwatson> nxvl: I assume ubuntu-archive is subscribed and thus it is in the queue
[22:26] <NCommander> nxvl, how's your shell scrpiting?
[22:31] <nxvl> cjwatson: yes, sleep tight!
[22:32] <nxvl> NCommander: shoot
[22:58] <slangasek> NCommander: fails for the case of gnome-www-browser -> abrowser.
[22:58] <NCommander> nxvl, http://pastebin.ca/1219253
[22:58] <slangasek> NCommander: I really did have a reason for specifying the fix that I did.
[23:03]  * NCommander feels lost
[23:03] <NCommander> Wait
[23:03] <NCommander> hrm
[23:04]  * NCommander gets it!
[23:20] <NCommander> slangasek, one problem with your idea. x-www-browser -> firefox-3.0/abrowser -> /usr/lib/firefox-$VERSION/firefox.sh
[23:20] <NCommander> (well, x-www-browser -> /etc/alternates/x-www-browser -> *)
[23:20] <slangasek> I don't think you've understood my idea yet
[23:20] <NCommander> I think your right
[23:21] <NCommander> Oh, I see
[23:21] <NCommander> I missed the "a matching name is found"
[23:25] <NCommander> Hrm
[23:25] <NCommander> abrowser still has Mozilla Firefox branding
[23:25] <slangasek> I have a patch here for it; I'll push it to the bug
[23:26] <NCommander> Oh, you do?
[23:26]  * NCommander just finished fixing his patch
[23:26] <NCommander> oh well, it was a good experience in shell scripting
[23:26] <slangasek> I do now, yes
[23:26] <NCommander> Sorry for the duplication of effort, but at least the bug gets resolved, and that's what counts
[23:29] <slangasek> NCommander: patch posted; please test
[23:31] <NCommander> slangasek, probably a stupid question, but /usr/lib/firefox-$VER/firefox will exist even if only abrowser is installed, right?
[23:32] <slangasek> that file appears to belong to the firefox-3.0 package, so probably not?
[23:33] <NCommander> Your call of last resort should do a sanity check to see that $LIBDIR/firefox exists, and then try $LIBDIR/abrowser before bombing out
[23:34] <cody-somerville> oh noes
[23:34]  * NCommander is test building your package as we speak
[23:34] <cody-somerville> It should try icemonkey or whatever it is called too
[23:34] <slangasek> it won't fail for the case where the called link is, or points to, abrowser
[23:34] <slangasek> the last-resort is only ever reached if the caller has done something very wrong
[23:35] <NCommander> I'm just noting a case that for whatever reason someone links to firefox.sh directly, so ...
[23:35] <NCommander> Yeah, pretty much
[23:35]  * NCommander notes that abrowser REALLY needs a better icon
[23:36] <slangasek> and actually, it appears that abrowser Depends: firefox-3.0, so even that isn't really a concern currently
[23:37] <wgrant> firefox depends upon firefox-3.0-branding, which is the one with the nasty bits.
[23:37]  * NCommander nods
[23:38] <NCommander> Just out of idle curosity, slangasek, do you use abrowser or firefox?
[23:38]  * ogra sighs ... debian is such a pain if you are used to ubuntu
[23:38] <slangasek> firefox
[23:39] <NCommander> slangasek, your fix appears to work fine
[23:40] <NCommander> (I can use x-www-browser and sensible-browser without any issues and abrowser pops up)
[23:40] <NCommander> wgrant, what about you?
[23:41] <NCommander> slangasek, I commented on the bug
[23:48]  * slangasek nods
[23:48] <NCommander> slangasek, I envy your shell scrpiting abiltiies; any tips on where I can learn better scripting?
[23:49] <slangasek> hmm, absorb the entire archive of debian-devel through your skin, perhaps :)
[23:49] <wgrant> NCommander: I use abrowser
[23:49] <slangasek> but probably applying a joeyh pattern filter first
[23:49] <wgrant> Hah.
[23:50] <NCommander> slangasek, as soon as we get an open source time machine, I shall ;-)
[23:50] <slangasek> time machine?
[23:50] <NCommander> wgrant, are you agreed with me that abrowser needs better icons?
[23:50] <NCommander> slangasek, I'd like to go back intime and sit in on #debian-devel :-)
[23:50] <wgrant> NCommander: I don't think it's too bad, actually.
[23:51] <NCommander> I can't believe I'm the one saying this, but its too blue :-P
[23:51] <slangasek> NCommander: I was referring to the mailing list
[23:51] <NCommander> slangasek, ah. Maybe I'll sit aside and read the entire archive one day
[23:52] <slangasek> that's certainly a long way to go about it, though.  Alternatively, if Ian Jackson has ever written a shell scripting guide, that would be a good starting point too.
[23:52] <slangasek> and probably only a third as long ;-)
[23:52] <NCommander> lol
[23:52] <NCommander> Hrm
[23:52] <NCommander> Things I find frustating
[23:52] <NCommander> Its impossible to describe a song over the internet if you don't know the name
[23:53] <NCommander> er
[23:53] <NCommander> wrong channel
[23:53] <NCommander> But true
[23:53] <NCommander> slangasek, the longest thing I ever read was the entire hurd-devel archive
[23:56] <NCommander> ogra, why is Debian a pain compared to Ubuntu?
[23:57] <ogra> well, its starts with /sbin not being in any path
[23:57]  * ogra is playing with a debian port on his n800, where typing is massively painful and they didnt manage to give *any* function to the keypad
[23:58] <NCommander> ogra, sounds like an Ubuntu port to ARM is in order
[23:58] <ogra> its only fou silly keycodes to make it work as joystick, sigh
[23:58] <ogra> *four
[23:58] <ogra> not yet
[23:58] <ogra> there is mojo.handelds.rog though
[23:58] <ogra> *org
[23:59] <slangasek> ogra: what are you missing being able to run from /sbin?
[23:59] <slangasek> I think the /sbin in the default path only makes sense because of sudo-by-default
[23:59] <ogra> well, ifup and ifdown
[23:59] <slangasek> which you can't run as a non-root user anyway?
[23:59] <ogra> if you type with a pen giving the full path is just a pain