[00:43] <kees> say, when including X11/extensions/XTest.h, it now includes X11/extensions/XInput.h, but libxi-dev isn't a depend of x11proto-ext-dev
[00:45] <bryce> that's probably a packaging oversight
[00:46] <bryce> kees: how did you run into that?
[00:47] <kees> bryce: I was recompiling synergy to test some small changes, and hit a configure failure.  tracking it to config.log, I saw the "XText.h: cannot find XInput.h" error
[00:48] <kees> a trivial fix would be to have x11proto-ext-dev depend on libxi-dev, but I have no idea if that's the "right" fix.
[00:49] <bryce> it probably is; xinput is new and gradually is getting folded in to various things.  
[01:02] <bryce> x11proto-xext-dev, right?
[01:04] <bryce> hmmm
[01:04] <bryce> kees: weird, the version we have of that is 7.0.2 which was released in like Dec 2005
[01:06] <kees> bryce: maybe XInput.h moved from something that was a long-time Depend of x11proto-ext-dev into a separate libxi-dev package?
[01:08] <bryce> yeah... checking the debian patches first tho
[01:08] <bryce> no patches :-)
[01:08] <bryce> hmm
[01:12] <bryce> well, x11proto-xext depends only on x11-common
[01:13] <kees> I was poking at x11proto-xext-dev
[01:15] <bryce> right, I meant in general x11proto-xext* only has a single dependency
[01:15] <kees> ah
[01:15] <bryce> (well, plus debhelper)
[01:15] <bryce> crap no I'm wrong
[01:16] <bryce> Depends: ${shlibs:Depends}, ${misc:Depends}, x11proto-input-dev, libxau-dev
[01:16] <bryce> Pre-Depends: x11-common (>= 1:7.0.0)
[01:16] <bryce> was looking only at the latter line
[01:16] <kees> oh, I bet libxi-dev should be a depend of x11proto-input-dev
[01:17] <bryce> http://packages.ubuntu.com/search?searchon=contents&keywords=XInput.h&mode=exactfilename&suite=hardy&arch=any
[01:17] <bryce> that says that XInput.h is provided by x11proto-input-dev
[01:17] <bryce> aha
[01:17] <kees> http://packages.ubuntu.com/search?searchon=contents&keywords=XInput.h&mode=exactfilename&suite=intrepid&arch=any
[01:17] <bryce> so in intrepid this has changed
[01:17] <kees> and in intrepid it moved.  :P
[01:17] <bryce> http://packages.ubuntu.com/search?suite=intrepid&arch=any&mode=exactfilename&searchon=contents&keywords=XInput.h
[01:17] <bryce> mm
[01:18] <kees> this is pretty recent, too.
[01:18] <bryce> okay so guess we just need to update the Depends 
[01:18] <kees> it wasn't like that a few weeks ago
[01:20] <bryce> kees, mind filing an LP on this?  I'll forward it up from there
[01:20] <kees> sure
[01:20] <bryce> and I'll have the upload ready in a jiff
[01:21] <bryce> oh and please indicate if installing libxi-dev is sufficient for fixing synergy's compilation issue
[01:23] <kees> bryce: https://bugs.edge.launchpad.net/ubuntu/+source/x11proto-xext/+bug/273386
[01:25] <bryce> thx
[01:26] <bryce> http://bryceharrington.org/ubuntu/x11proto-xext_7.0.2-6build2.debdiff
[01:29] <kees> typo'd: libx1-dev vs libxi-dev
[01:29] <bryce> http://www.pbs.org/cgi-registry/poll/poll.pl
[01:29] <kees> also, maybe the dep should go into x11proto-input-dev ?
[01:30] <bryce> good catch; strange that pbuilder accepted and built it ok
[01:31] <bryce> hmm, not sure whether it should go into x11proto-input-dev.  But I'll mention that when forwarding to debian.
[01:41] <bryce> kees: ok filed with debian.
[01:42] <kees> cool, thx
[01:48] <bryce> deb bug #499858
[02:30] <bryce> jesse says he'll try to be more careful in the future
[07:17] <tjaalton> bryce: the XInput.h change was made by me, so it doesn't affect debian
[07:17] <tjaalton> not yet anyway
[07:17] <tjaalton> that was needed for the input properties
[07:17] <soren> Does anyone know why I don't have any dead keys anymore?
[07:19] <tjaalton> soren: look at /etc/default/console-setup
[07:19] <tjaalton> are the XKB* same as in your old xorg.conf?
[07:20] <soren> tjaalton: Looks like it, yes.
[07:20] <tjaalton> ok, what's in there?
[07:21] <soren> XKBMODEL="pc105"
[07:21] <soren> XKBLAYOUT="dk"
[07:21] <soren> XKBOPTIONS="lv3:ralt_switch"
[07:21] <soren> XKBVARIANT=""
[07:21] <tjaalton> no nodeadkeys there either
[07:21] <soren> No "nodeadkeys".
[07:21] <soren> Nope.
[07:21] <soren> Not in setxkbmap -print either.
[07:22] <tjaalton> hmm ok, so what did you want again?-)
[07:24] <tjaalton> dead keys work here though
[07:24] <tjaalton> what does setxkbmap -print output?
[07:24] <tjaalton> put out, whaeva
[07:25] <tjaalton> bryce: btw, don't use -Xbuild* versions if the package has real changes, since they would be lost when packages are being autosynced
[07:34] <soren> tjaalton: Sorry, had a phone call. I'll pastebin the setxkbmap -print output.. Just a sec.
[07:34] <soren> tjaalton: http://pastebin.com/m69f31be7
[07:35] <soren> "setxkbmap dk nodeadkeys" doesn't help either (in case it somehow got inverted or something)
[07:36] <soren> What I want is dead keys. ´a, ~n, and  ¨a just don't look right :)
[07:41] <tjaalton> soren: ok, so that's identical to mine except I have fi instead of dk.. I'll have a look at xkb-data
[07:47] <soren> tjaalton: "setxkbmap fi" doesn't give me back my dead keys either.
[07:48] <tjaalton> m
[07:48] <tjaalton> hm'
[07:48] <tjaalton> uh
[07:48] <tjaalton> try the guest-user
[07:49] <soren> Can I log in from GDM as "him"?
[07:49]  * soren doesn't run GNOME
[07:49] <tjaalton> ok, dunno
[07:49] <soren> ...so I don't have the user-switch applet
[07:49] <tjaalton> create a dummy user then
[07:50] <tjaalton> do they work on the console?
[07:52] <soren> Yes. Console is fine.
[07:53] <soren> The new dummy account doesn't work either.
[07:53] <tjaalton> hrm
[07:54] <soren> To make things even more fun, my original X session is now just a black screen with my mouse cursor on it. It happened after I switched to the dummy user's session and back. :/
[07:54] <soren> Er.. Ok, scratch the "console is fine" line.
[07:54] <tjaalton> intel?
[07:54] <soren> Yes.
[07:55] <tjaalton> hm, works here (965)
[07:55] <tjaalton> but I've seen that
[07:55] <tjaalton> if console is broken as well, I'm out of ideas :)
[07:56] <soren> No, the console seems to be fine again.
[07:56] <soren> It was just acting really strangely for a bit there. Instead of deleting characters when I pressed backspace, it just put two diamond shaped characters where the cursor was.
[07:56] <soren> That was at the login prompt. WHen I'm actually logged in, it works fine.
[08:12] <tjaalton> soren: and dead keys with dk work here
[08:12] <soren> I'm not sure where to begin to look. I don't have an xorg.conf.
[08:13] <tjaalton> do you have gnome installed?
[08:13] <soren> tjaalton: Yes, the dummy user logged into GNOME. No luck.
[08:13] <tjaalton> bah
[08:33]  * tseliot starts bugging tjaalton
[08:33] <tseliot> tjaalton: can you upload my patch for touchpads, please?
[08:34] <tseliot> https://bugs.launchpad.net/ubuntu/+source/xfree86-driver-synaptics/+bug/271823
[08:50] <tjaalton> tseliot: done
[08:50] <tseliot> tjaalton: thanks a lot. Where shall I send that crate of beer I promised you? :-P
[08:51] <tjaalton> tseliot: Ristihaantie 8 C 34, 02750 Espoo, Finland. Thanks! :)
[08:52] <tseliot> ;)
[10:17] <tjaalton> sigh.. https://bugs.edge.launchpad.net/ubuntu/+source/xkeyboard-config/+bug/261573/comments/45
[12:48] <Kano> hi
[12:48] <Kano> why has ubuntu 8.10 a different (1 line) xorg.conf than lenny
[12:49] <Kano> Monitor section
[12:49] <Kano> Device          "Configured Video Device"
[12:49] <Kano> is there but not in lenny or sid
[12:49] <Kano> wouldnt it be nice to write the _same_ config file?
[12:49] <tjaalton> look at /usr/bin/dexconf, they are not the same
[12:50] <Kano> i would prefer to write both times the same, because otherwise i need a workaround to remove or add that line...
[12:51] <tjaalton> then you edit dexconf to suit your needs
[12:51] <Kano> thats not possible
[12:52] <Kano> but nice to know that file...
[12:53] <tjaalton> btw, that line is there in sid
[12:53] <tjaalton> in the sid version of dexconf
[12:53] <Kano> but not in lenny
[12:53] <Kano> would you add or remove that line
[12:55] <tjaalton> sure is in lenny too
[12:56] <Kano> well it is in 2 positions in ubuntu
[12:56] <Kano> i mean the MONITOR section
[12:56] <Kano> the one with device in front
[12:56] <Kano> the one with indentifier is in both
[12:57] <tjaalton> those sections are identical in all three versions
[12:57] <tjaalton> lenny/sid/intrepid
[12:57] <Kano> sorry screen section
[12:58] <Kano>         Identifier      "Default Screen"
[12:58] <Kano>         Monitor         "Configured Monitor"
[12:58] <Kano>         Device          "Configured Video Device"
[12:58] <Kano> thats written in intrepid
[12:59] <Kano> the device line is not present in lenny, just installed yesterday
[12:59] <Kano> very curious that you say it is in all of em...
[13:00] <tjaalton> because you weren't talking about the Screen section before
[13:00] <tjaalton> it is there alright
[13:01]  * wgrant wonders why that would need to be worked around.
[13:01] <Kano> wgrant: well the "new" style xorg.conf is really small
[13:01] <Kano> therefore i added a patch code to my scripts
[13:02] <Kano> that makes am slightly larger to be used to change to nvidia/fglrx later
[13:02] <tjaalton> bryce: btw, better use "-I".git" next time for xorg ;)
[13:02] <tjaalton> um, '-I".git"'
[13:02] <Kano> and what i dislike is when that patch does not apply when i set it for one of those
[13:03] <Kano> as i need to change the screen section...
[13:04] <Kano> i prefer 1 code for all,but it seems i need a workaround
[13:04] <Kano> removeing a line is usually more easy than adding
[13:04] <tseliot> Kano: what's the problem with having a reference to the Device section in the Screen section?
[13:04] <tseliot> it's perfectly legal to do so
[13:05] <Kano> there is no problem, just differnt to debian
[13:05] <Kano> and that difference makes my script to not apply for one of em without mod
[13:05] <jcristau> and why does that matter?
[13:05] <Kano> therefore i wanted to know WHY it is different
[13:06]  * wgrant suggests that it is because they're not the same.
[13:06] <tseliot> Kano: why don't you use X-Kit to modify the xorg.conf? It would take only few lines of Python code
[13:06] <tjaalton> oh right; "dexconf: Bring Device back to Screen-section"
[13:06] <Kano> i dont know phython
[13:06] <tjaalton> "nvidia-settings and aticonfig need that. (LP: #181405)"
[13:06]  * jcristau uses sed to modify xorg.conf
[13:07] <tjaalton> Kano: so there, fix nvidia-settings and aticonfig first ;)
[13:07] <Kano> tjaalton: ok, then i will add that line for debian
[13:07] <tjaalton> although I wouldn't be pissed if those broke again
[13:07] <jcristau> (and i get scared i'll break something every time i have to do that)
[13:07] <tjaalton> since that's all they do anyway
[13:08] <Kano> aticonfig is really unimportant for 8.10 as fglrx does not work anyway *g*
[13:08] <jcristau> tjaalton: how did you handle the i810 -> intel and via -> openchrome stuff, btw?
[13:08] <jcristau> (if at all)
[13:08] <tjaalton> jcristau: with cruelty
[13:08] <Kano> i810 -> intel is trivial
[13:09] <tjaalton> jcristau: ie. no mangling like you've done, although we might want to merge to get the code it
[13:09] <tjaalton> *in
[13:09] <Kano> debian just uses a wrapper module
[13:09] <jcristau> tjaalton: okay
[13:09] <jcristau> Kano: well, no.
[13:09] <Kano> jcristau: the i810 package has it
[13:10] <Kano> which would be there on update
[13:10] <tjaalton> there is no i810 package
[13:10] <jcristau> the i810 package is empty in sid, and doesn't exist in experimental. so no.
[13:10] <Kano> well it was there before..
[13:10] <jcristau> and that's besides the point anyway
[13:10] <tseliot> tjaalton: BTW I had to change a handler in Jockey because amdcccle crashes if you put something like Disable "dri2" in the Module section of the xorg.conf.
[13:11] <jcristau> so fix amdcccle?
[13:11] <tseliot> jcristau: if it were open source I would fix it myself
[13:11] <tjaalton> it's silly to break because of that
[13:12] <tseliot> :_/
[13:12] <jcristau> it seems there's all kinds of silly xorg.conf parsers all over the place
[13:12] <tseliot> hehe right
[13:12] <jcristau> Kano: the point is, 'Driver "i810"' in xorg.conf doesn't work with xserver 1.5, so it has to be changed to intel.
[13:13] <Kano> jcristau: most easy way: remove the line
[13:13] <jcristau> Kano: you're useless, thanks.
[13:13] <Kano> ;)
[13:13] <Kano> xorg can find the driver automatically
[13:14] <Kano> just not nvidia or fglrx
[13:14] <jcristau> tell me something i don't know
[13:14] <Kano> well you know everything it seems, then you surely have got fglrx working with xserver 1.5
[13:15] <tjaalton> Kano: for the record, jcristau is the head of Debian XSF ;)
[13:16] <Kano> well then he should write the same configfile for all ;)
[13:16] <tjaalton> but ubuntu messes things up, so no
[13:16] <Kano> bbl
[13:17] <jcristau> we should probably not write an xorg.conf at all in squeeze.. at least in the common case.
[13:17] <tjaalton> yeah
[13:19] <Ng> no xorg.conf ftw \o/
[13:19] <jcristau> i'll write to debian-boot, to see what their plans are re: console-setup
[13:19] <Ng> except the emulatescrollwheel bit ;)
[13:19] <tjaalton> jcristau: \o/
[13:19] <tjaalton> Ng: just use xinput :)
[13:20] <Ng> tjaalton: hmm? all I saw on it was mvo's blog post about building his own evdev or something and using xinput properties - is that properly in intrepid now?
[13:21] <tjaalton> Ng: sure is
[13:21] <Ng> suh-weet
[13:21] <Ng> should I just pinch the config bits from said blog post?
[13:22] <tjaalton> xinput list, xinput list-props $id etc
[13:22] <tjaalton> man xinput too
[13:22] <tjaalton> but the blog post might help too
[13:44] <Ng> win
[13:45] <Ng> two xinput commands seems to do the trick
[13:46] <tjaalton> nice
[13:46] <tjaalton> in time there'll be gui support for it
[13:47] <tjaalton> now you need to run those on every session start..
[13:47] <Ng> yeah, for now a session startup script will do :)
[13:48] <Ng> thanks very much :)
[13:48] <tjaalton> my pleasure
[13:49] <Ng> the pleasure is mine, I have my scrolling back ;)
[13:54] <Ng> I added a comment to mvo's post about it with the commands I ran
[13:54] <mvo> thanks Ng
[13:59]  * jcristau also pesters the xcb folks about the libxcb buffer size
[14:38] <Kano> btw. what could be the problem with ubuntu 8.10 that my nvidia 8800 gts 512 only runs with binary driver? the driver detected does not show any errors in the log but there is no pic (vga via dvi adapter)
[19:06] <bryce> tjaalton: is there a way we can set things to use that switch automatically?
[19:08] <jcristau> alias dpkg-buildpackage='dpkg-buildpackage -i -I'
[19:14] <tjaalton> something like that ;)
[19:15] <superm1> bryce, make a file named ~/.devscripts
[19:15] <superm1> that contains this:
[19:15] <superm1> DEBUILD_DPKG_BUILDPACKAGE_OPTS="-i'(?:^|/).*~$|(?:^|/)\..*\.swp|DEADJOE|(?:/CVS|/RCS|/\.svn|/\.deps|\{arch\}|\.arch-ids|\.arch-inventory|\.bzr|\.bzrignore|\.shelf)(?:$|/)' -ICVS -I.svn -I\{arch\} -I.arch-ids -I.arch-inventory -I.bzr -I.bzrignore -I.shelf"
[19:15] <bryce> thanks
[19:15] <superm1> I'm not sure where, how, or from whom i got it, but it works
[19:15] <jcristau> you don't need all that
[19:16] <superm1> if you use debuild it works at least
[19:16] <jcristau> dpkg-buildpackage has default regexps for -i and -I
[19:16] <superm1> they weren't catching bzr stuff for me ever until I started to use  that
[19:16] <bryce> maybe DEBUILD_DPKG_BUILDPACKAGE_OPTS="-i -I.bzr -I.svn -I.git" is enough?
[19:19] <jcristau> bryce: those should be taken care of by -I afaict
[19:19] <jcristau> superm1: dpkg's changelog says it added bzr stuff in 1.13.14
[19:20] <jcristau> hmm. only for -i. although tar_ignore_default_pattern contains .bzr and friends here.
[19:21] <bryce> jcristau: I'd assumed so too but timo said at least -I".git" is needed
[19:22] <jcristau> oh.
[19:22] <jcristau>   * Allow dpkg-source -I without a pattern which will load a default
[19:22] <jcristau>     list of pattern similar to -i without regexp. Patch by
[19:22] <jcristau>     Jari Aalto. Closes: #440972
[19:22] <jcristau> that's 1.14.7
[19:23] <jcristau> a year ago
[20:20] <tjaalton> I've seen that changelog entry, but for some reason it doesn't seem to work..
[20:21] <bryce> tjaalton: btw, Intel says that Q3 final release of -intel will be Oct 14th
[20:21] <bryce> tjaalton: they asked if we could include it at that date.  I said we'd need to see the list of changes
[20:22] <tjaalton> bryce: does it work without GEM?
[20:22] <bryce> tjaalton: yeah, they're not shipping this one with GEM enabled yet
[20:22] <tjaalton> oh
[20:22] <bryce> they found too many serious stability issues
[20:22] <tjaalton> hehe
[20:41] <mnemo> hey guys, I just bought a new machine with intel G45 chipset (integrated GMA X4500GD card).. Xorg goes bonkers after gdm, flashing for a while and then shows me unreadable dialog (due to corrupted graphics, kernel still live though)
[20:41] <mnemo> im getting this X.org stacktrace from the intel driver:
[20:41] <mnemo> http://launchpadlibrarian.net/17890192/Xorg.0.log
[20:41] <mnemo> any ideas on how I can work around it? already tried disabling compiz, still crashes though
[20:44] <tjaalton> mnemo: use the vesa driver?
[20:46] <mnemo> how can I configure it to do that?
[20:46] <tjaalton> man xorg.conf
[20:46] <tjaalton> Device section
[20:47] <mnemo> is that the same thing as saying AccelMethod XAA ?
[20:47] <tjaalton> no
[20:47] <mnemo> which one would be better intel in XAA mode or VESA?
[20:47] <mnemo> because I just noticed that it doesnt crash in XAA mode
[20:48] <tjaalton> oh
[20:49] <tjaalton> sorry I misread, yes it's the same section where you put that option
[20:49] <tjaalton> intel probably is better
[20:50] <tjaalton> bryce: we probably need to put the beta version in a PPA
[20:52] <mnemo> i'd be happy to help you out with testing the intel driver
[20:52] <jcristau> 2.4.2 has known problems on the new platform
[20:52] <tjaalton> yep
[20:54] <bryce> mnemo: if you actually want to help with testing, join the ubuntu-x@ mailing list; that's where we announce testing opportunities generally 
[20:55] <mnemo> im on it ;>
[21:48] <tjaalton> intel 2.5-branch still needs libdrm >=2.4.0, so won't build
[21:50] <tjaalton> bryce: ^^
[22:18] <pwnguin> there's an ubuntu-x mailing list?
[23:10] <bryce> pwnguin: yup