| tormod | so a few settings for a standard xorg-edgers cd, then just a few changes for a kms cd | 00:00 |
|---|---|---|
| bryce | btw does this iso include apw's kernel now? | 00:00 |
| tormod | or a xserver 1.7 cd which Sarvatt is customizing | 00:00 |
| tormod | yes, it has to, if kms should work | 00:01 |
| tormod | the build.radeon-kms.conf will d/l everything from the radeon-kms ppa so that will include the kms kernel there | 00:01 |
| tormod | basically I turned the build.sh into a generic "make a cd from a ppa" tool | 00:02 |
| bryce | heya jbarnes_PDX | 00:29 |
| bryce | jbarnes_PDX: a few weeks ago you did something to update priority on a bunch of intel crash/freeze bugs - by chance did you do this with a script? I'd love to see it, as I'm hoping to do up a launchpad/bugzilla report tool | 00:30 |
| jbarnes_PDX | bryce: no, just by hand | 00:51 |
| jbarnes_PDX | I'm a pretty manual guy :) | 00:51 |
| bryce | jbarnes_PDX: heh | 00:58 |
| Ng | how important is the fullscreen window unredirecting thing? Is it actually a massive performance win? I seem to have nothing but problems when I'm doing fullscreen things | 13:59 |
| tseliot | Ng: when playing videos, it should fix tearing problems, at least with nvidia. | 14:11 |
| Ng | tseliot: ah. the problem I have is that the fullscreen windows rarely seem to stay fullscreen&ontop, other things want to draw over them, so they have to flicker while they redirect | 14:12 |
| Ng | e.g. I can never get flash video to go fullscreen | 14:12 |
| tseliot | yes, that can be a pain | 14:13 |
| tseliot | e.g. notifications can break full screen unredirecting | 14:14 |
| Ng | yeah | 14:19 |
| hyperair | hmm it seems hal is being deprecated? | 17:51 |
| hyperair | in that case how does one go about configuring synaptics? | 17:51 |
| jcristau | synclient works | 17:52 |
| jcristau | or gpointing-device-settings, or whatever it's called this week | 17:52 |
| hyperair | gsynaptics i believe | 17:52 |
| hyperair | synclient isn't persistent | 17:52 |
| hyperair | and gsynaptics.... | 17:52 |
| hyperair | er.. | 17:52 |
| jcristau | gsynaptics is gone | 17:52 |
| hyperair | eh? | 17:53 |
| hyperair | yeah gpointingwhatever | 17:53 |
| hyperair | hmm | 17:54 |
| hyperair | let's see if it's any good | 17:54 |
| hyperair | but isn't there some desktop agnostic way of configuring it? | 17:54 |
| hyperair | xorg.conf doesn't exactly work now due to the whole autodetection thing which causes strange issues | 17:55 |
| hyperair | and if hal is deprecated, there goes another means. gpointingwhatever looks more like a hack than anything (setting the options when it's run upon desktop startup i suppose?) | 17:55 |
| jcristau | hal still works | 17:56 |
| bryce | for now | 17:57 |
| jcristau | it may be deprecated, but there's no replacement.. | 17:57 |
| bryce | jcristau: udev? | 17:57 |
| jcristau | meh. | 17:57 |
| bryce | *shrug* well that's what I've heard bandied about | 17:57 |
| jcristau | even if it's udev, the code to use that in X doesn't exist. | 17:59 |
| * bryce nods | 18:05 | |
| hyperair | and udev is a pain to configure | 18:11 |
| hyperair | even harder than hal | 18:11 |
| jcristau | oh, that's possible? | 18:12 |
| hyperair | er. i don't think it's possible for synaptics | 18:12 |
| hyperair | but doesn't the touchpad have some sort of kernel module? | 18:12 |
| tjaalton | possible being harder than hal | 18:12 |
| hyperair | it is possible then? | 18:12 |
| hyperair | doesn't every damn thing revolve around hal? why do we want to deprecate it then? | 18:13 |
| tjaalton | I think the goal is for the end user to never need to poke udev rules | 18:13 |
| tjaalton | upstream deprecated it | 18:13 |
| hyperair | seriously? O_o | 18:14 |
| hyperair | in favour of what? | 18:14 |
| tjaalton | udev | 18:14 |
| jcristau | hyperair: job security. when you have something that works, deprecate it, and start again. | 18:14 |
| tjaalton | hehe | 18:14 |
| hyperair | .. | 18:14 |
| hyperair | ....~. | 18:14 |
| hyperair | shit. ssh session acting up | 18:14 |
| hyperair | jcristau: good point. | 18:15 |
| hyperair | Sarvatt: is there something wrong with Gallium? | 20:47 |
| Sarvatt | it defaults to building a few gallium things, we dont ship it so its just wasting build time, and its basically not working right outside of softpipe right now? if you're asking why its disabled in a hook on edgers.. | 21:17 |
| bryce | heya Sarvatt | 21:18 |
| Sarvatt | heyo! | 21:18 |
| bryce | Sarvatt: btw how is the xserver snapshot in your ppa? Something we could pull into xorg-edgers, or the xi2 stuff still too unstable? | 21:18 |
| Sarvatt | works for me™ but as you can see alot of stuff needs to be disabled that I dont know about reenabling, and every input driver rebuilt against it (i only did the few there), only problem I have seen is gimp crashing due to XI2 | 21:22 |
| Sarvatt | switching away from edgers would be *fun* if people just blindly updated to the new stuff :D | 21:22 |
| Sarvatt | accidentally built that livecd on the ppa page for it with the radeon KMS kernel when it didnt have the supporting drm/mesa/ddx in there, have to fix that up | 21:24 |
| bryce | oh yeah driver rebuilds, hmm | 21:27 |
| bryce | Sarvatt: think there's going to be any more ABI changes now that xi2 is in? I didn't spot any obvious discussion either way on the xorg list. | 21:28 |
| bryce | with alpha-2 coming up, I'm thinking after that is released would be a good spot to push in the xserver snapshot | 21:29 |
| bryce | that'd maximize the amount of time we have for testing it before alpha-3 | 21:29 |
| bryce | I'd *sort of* like to see it tested in xorg-edgers prior to that, although rebuilding all the drivers for the ppa sounds like a hassle | 21:30 |
| bryce | tjaalton: do you have any comments on rolling out xserver? | 21:30 |
| Sarvatt | oh nice, the gimp crashes were fixed in the lastest one i uploaded, didnt notice | 21:33 |
| hyperair | hmm xi2? | 21:34 |
| hyperair | Sarvatt: xorg-edgers's snapshots have xi2 support? | 21:34 |
| Sarvatt | no i've been doing it in a seperate PPA | 21:35 |
| hyperair | aah i see | 21:35 |
| hyperair | how does one activate the xi2 bits anyway? | 21:35 |
| hyperair | or would i automatically have separate cursors for my touchpad and mouse? | 21:35 |
| tjaalton | bryce: not really, I haven't tried master at all. but I think that the video driver abi hasn't changed, so it's probably just the input drivers that need a rebuild? | 22:11 |
| bryce | sounds like it | 22:12 |
| tjaalton | which would mean that <gasp> fglrx would continue to work | 22:12 |
| bryce | can't be | 22:12 |
| bryce | certainly at least the kernel will have changed enough to break it... :-) | 22:12 |
| tjaalton | hehe :) | 22:13 |
| tjaalton | #define ABI_VIDEODRV_VERSIONSET_ABI_VERSION(5, 0) | 22:15 |
| tjaalton | so it's still at 5 | 22:15 |
| tjaalton | but that doesn't mean it won't change before the release ;) | 22:16 |
| tjaalton | bryce: also, in case you missed, xorg in unstable has stopped shipping the xorg.conf. it's halfway merged already, probably something to upload along the rest when it's done | 22:19 |
| Sarvatt | its at 6 now | 22:20 |
| tjaalton | INPUTDRV is | 22:20 |
| tjaalton | sorry, XINPUT_VERSION | 22:20 |
| Sarvatt | oh video I'm sorry | 22:20 |
| bryce | tjaalton: how do you feel about not shipping xorg.conf? | 22:22 |
| Sarvatt | funny thing is I havent had any luck using the drivers built against 1.6.1.901 on 1.6.99.1 (but I only tested intel), no idea if its due to the dri2proto change or what | 22:22 |
| bryce | I sort of feel there's more value to keeping at least a skeleton version of it | 22:22 |
| bryce | I wonder if there's any boot time benefit to running without xorg.conf though | 22:22 |
| tjaalton | bryce: I'm ok with it. there are some quirks in dexconf which might need another look | 22:22 |
| Ng | interesting situation atm, I just resumed, I have working kb and the mouse moves, but xev reports no motion events, and I can't click on anything (lack ofmotion events is slightly confirmed by a lack of mouse-following focus) | 22:23 |
| Ng | no obvious log entries | 22:24 |
| virtuald | where do i follow development of the radeon-kms ppa? where's the mailing list? | 22:25 |
| Sarvatt | do you guys --enable-config-dbus in xserver for a reason? master builds were broken for awhile after the xi2 merge with that and everyone i talked to said it shouldnt be enabled anyway.. couldnt find anything broken disabling it (it's default to disabled and i cant find any distros enabling it except ubuntu and gentoo), jcristau was saying it was for wacom before that was fixed to work with hal but wacom works fine with it disabled now | 22:25 |
| Sarvatt | http://sarvatt.com/downloads/Xorg.0.log.txt | 22:25 |
| tjaalton | Sarvatt: probably no reason to do that anymore. it was only enabled to allow some sort of wacom hotplug at the time | 22:26 |
| virtuald | i get no mouse pointer on my second screen :> | 22:26 |
| bryce | virtuald: most discussion is here on this channel, but there is also ubuntu-x@lists.ubuntu.com | 22:27 |
| virtuald | thanks | 22:28 |
| bryce | Sarvatt: maybe tseliot uses the dbus interface for his synaptics configuration stuff? (But his dbus daemon isn't in Ubuntu yet) | 22:28 |
| tjaalton | I think we'll notice ;) | 22:30 |
| Sarvatt | i should have gotten bootcharts of it when i was disabling it, it felt like it booted faster disabled | 22:31 |
| tjaalton | there should be no need for it | 22:32 |
| tjaalton | I hope tseliot uses Xinput in his tool, just like gpointing-blah | 22:33 |
| * Ng files a bug :) | 22:33 | |
| tjaalton | Ng: some app grabbed the focus? | 22:34 |
| Ng | tjaalton: hmm, I did have vinagre running across the suspend, which may do something like that, but I killed it from a console | 22:34 |
| bryce | tjaalton: perhaps not for configuration, but what about for troubleshooting? | 22:35 |
| Ng | man I love vimperator. my browser is just as useful without a mouse as with :D | 22:36 |
| tjaalton | bryce: xorg.conf? | 22:36 |
| tjaalton | bryce: you can still run dexconf | 22:36 |
| Ng | tjaalton: is there any way to reset that kind of thing? | 22:37 |
| Ng | (I'm guessing not ;) | 22:37 |
| tjaalton | Ng: I'm not sure.. | 22:37 |
| tjaalton | killing the app probably should release it | 22:38 |
| bryce | tjaalton: hmm | 22:40 |
| Ng | tjaalton: not so much ;( | 22:41 |
| === halfline is now known as halfline_ | ||
| === halfline_ is now known as halfline | ||
| tjaalton | gotta go, bye | 22:47 |
| Ng | ooh, it was definitely a grabbing issue, locking the screen and unlocking has restored normality \o/ | 23:00 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!