[00:20] B-) - cool [00:29] Anybody knows if some hardy repos are down atm (can't reinstall compiz)? [00:36] Hmm, seems the UK repos are down (some at least), Main works fine though. [00:48] Arand: please ask hardy questions in #ubuntu [00:51] Can't connect to there through mibbit, and my uni blocks IRC otherwise... [00:51] Arand: Follow the directions when you are forwarded to #ubuntu-proxy-users and it will let you in. [00:54] does anyone know anything about ubuntu mobile? is it out in beta yet? [00:58] Sounds like one weak uni === crd1b is now known as crdlb === ajmorris|AFK is now known as ajmorris [04:33] I want to reboot my intrepid test laptop, but then I'll lose xchat on it [04:33] wish there was something I could do to update the xchat package === cyclonut is now known as cyclo|sleep [04:56] I'd like to make a request to have consistent timezones in 8.10. Right now on my 8.04 I have some as stuff like "America/Chicago" but some as stuff like "America/Kentucky/Monticello" and still others like "America/Argentina/Buenos_Aires" [04:57] pick one and go with it === Sebast1an is now known as Sebastian [11:14] Hi all, I've to add repos in a wiki page for an ufficial sources.list to get start with Intrepid [11:14] Do you know where I can find urls repos ? [11:18] l3on: I'm not sure what you're asking, but the official Intrepid repositories are exactly the same as the official Hardy (and Gutsy, etc) repositories. [11:19] RAOF: yep, I'm wrong on ask... I would know if for intrepid there will be a commercial repo. [11:19] Yes, probably. [11:19] But not yet. [11:20] RAOF: ah ok... [11:20] these are right -> http://wiki.ubuntu-it.org/Repository/SourcesList#intre [11:20] ? [11:22] sorry, wrong link -> http://wiki.ubuntu-it.org/Repository/SourcesList#intrepid [11:31] bai. === ajmorris is now known as ajmorris|AFK [13:45] !java-plugin [13:45] Factoid java-plugin not found [13:45] !java [13:45] To install a Java runtime/interpreter on Ubuntu, look at https://help.ubuntu.com/community/Java - For the Sun Java runtime install sun-java6-jre from the !Multiverse repository (in !Backports for !Edgy) === fdd-0 is now known as fdd === Webspot_ is now known as Webspot === cyclo|sleep is now known as cyclonut [16:18] where can i find the pre alpha to download [16:19] there's not iso images yet [16:19] also if you ask that question you really dont wanna try it now [16:20] well i do i have an machine just for testing things so it would be ok [16:20] do you know when they will have the iso for that pre alpha [16:24] alpha 1 i guess [16:31] thanks [16:33] geez, a bit chomping at the bit [16:33] ? [16:33] asking for dailies/pre alpha [16:50] Why has alsa dmix been disabled in Hardy? [16:52] pulseaudio is intended to be used (and it sounds terrible) [16:52] can intrepid have consistent timezones? [16:53] humbolto: ...it hasn't. [16:54] right now I have them as varied as Europe/London, America/Argentina/Buenos Aires, America/Kentucky/Monticello, etc... consistency would be nice [16:54] also this is not the hardy channel [16:54] just don't use pulseaudio, e.g., remove that package or disable the esd gconf session option [16:54] crimsun: So why can I not play a youtube video when playing something in Rhythmbox? (fresh install) [16:54] humbolto: because of bug 192888. [16:54] Launchpad bug 192888 in pulseaudio "firefox crashes on flash contents when using libflashsupport" [High,In progress] https://launchpad.net/bugs/192888 [16:54] humbolto: so, install libflashsupport, or just disable pulseaudio altogether. [16:55] humbolto: and, rsk is on point. (These are #ubuntu questions.) [16:55] Firefox does not crash! [16:56] And what I am experiencing right now it the opposite. I can play flash, as I did that first, but Rhythmbox hangs. [16:56] that's all due to 192888. [16:56] pa is configured to grab hw:, which bypasses dmix. [16:56] No libflashsupport installed. And 192888 states that firefox crashes, when libflashsupport is installed. [16:57] (un)fortunately, pa also is configured to relinquish hw: after an idle period [16:57] the change to flashplugin-nonfree removed the dependency on libflashsupport, which means: [16:57] crimsun: For what purpose is it configured to grab hardware? [16:57] I'll explain that in a sec. [16:58] the change to flashplugin-nonfree removed the dependency on libflashsupport, which means: [16:58] 1) hw: is nondeterministically available to native ALSA apps [16:58] well, that's pretty much the major issue [16:59] so, if you're playing music in RB, you can't hear a Flash movie [16:59] likewise, if you play a Flash movie (the plugin grabs dmix vice hw:), RB won't be audible - in fact, RB won't even attempt to play [16:59] ok, the quick fix would be to make pa not grab the hw so dmix is still active. [17:00] Can I tweak that or is that mandatory by pa? [17:00] that's precisely what I proposed, but it's definitely suboptimal. [17:00] see the bzr branch in the comments [17:00] the proper fix for 8.04.1 would be to enable nspluginwrapper on i386 and to readd the libflashsupport dependency to flashplugin-nonfree [17:01] Flash is just one app that might not comply. What about the rest? [17:02] humbolto: "the rest" is limited to non-Free ones, fortunately. [17:02] there's always pasuspender [17:02] aka, pasuspender -- audacity [17:02] like aoss for alsa? [17:02] you would use padsp instead of aoss. [17:03] By the way, there seems to be another problem hidden somewhere! [17:03] Might have something to do with sampling frequencies. [17:03] you'd have to be more specific [17:03] (I can think of a couple symptoms that are not necessarily related to sampling freq) [17:04] My brother has some drum loop of two heartbeats overlapping from time to time. Nice story, he says he remembers hearing his heartbeat and our mothers overlap before he was born. [17:04] Now he has this loop on his website and he says it is not played in the right speed on my linux machine. [17:05] So I figured there might be some problem with sampling frequency conversion somewhere. [17:05] quite possible, though the symptom has multiple problem layers [17:06] I have to track down this a little further before I can file a bug report. [17:06] first, how is the sound provided? [17:06] sound clip* [17:06] straight wav? via Flash applet? [17:06] i.e., save yourself some time by providing these stack details up front, since I'm likely to be triaging the bug [17:07] But if this really is Linux's fault, as you seem to be more deeply involved with ubuntu's audio arch, would you want me to notify you of the bug report? [17:07] sure [17:08] I don't believe it will be linux's default (as in the kernel's fault) [17:09] shit, I have been disconnected [17:09] crimsun: what was the last line you received? [17:10] 12:08 crimsun > I don't believe it will be linux's default (as in the kernel's fault) [17:10] 12:08 = humbolto [n=elias@213-147-185-150.adsl.ycn.com] quit (Remote closed the connection) [17:10] crimsun: Ok, seems you received everything. [17:11] crimsun: Well more likely pulse's or alsa's fault. [17:11] crimsun: Are you involved more deeply with the audio arch? [17:12] humbolto: unfortunately, yes. [17:12] humbolto: and it's rather unlikely to be alsa's fault, but I certainly won't rule it out currently [17:13] crimsun: Didn't alsa (dmix) default to 48khz? Now I see pulse using 44.1 (if what is commented out in the config is the default ; default-sample-rate = 44100) [17:14] so is libflashsupport save now? it is not marked as supported by ubuntu in synaptic! [17:16] humbolto: no, dmix does not necessarily default to 48KHz [17:17] humbolto: (it depends on whether you're referring to alsa-lib <= 1.0.16 or > 1.0.16) [17:17] Gutsy [17:17] humbolto: I don't understand your question regarding libflashsupport being "save" [17:18] No, actually last time i looked was feisty [17:18] i think save == safe [17:18] It used to crash [17:18] humbolto: libflashsupport is in universe, so it is correctly not supported by Canonical. [17:18] humbolto: the crashing issues are known, but they need to be resolved on Adobe's end [17:19] But if it is not supported and therefore not installed by default, Hardy is shipped as a broken system! [17:19] humbolto: it would be broken even if it were shipped by default. [17:20] As everybody will definetely run into the two problematic scenarios I described above. [17:20] I understand with Alexander's intent, which was to prevent FF from being so crash-happy [17:20] s/with// [17:20] How to disable hardware grabbing of pulse. How to keep dmix working? [17:20] Or why exactly is leaving dmix enabled a bad idea? [17:21] I am sure there was some reasoning prior to that decision. Can I read up on it somewhere? [17:21] IRC meeing log, ... [17:21] humbolto: yes, it's on the pulseaudio user mailing list [17:21] I don't have a URL handy [17:22] Summary? [17:22] essentially: 1) higher latency; 2) awful resampling algorithm [17:23] good enough an explanation for me. [17:23] I don't really think (1) is that much of a problem unless you're using PA and dmix on a 300 MHz machine [17:23] (at least that's the baseline of my tests) [17:23] (2) is quite understandable [17:23] How to reenable dmix? [17:23] alsa-lib defaults to linear resampling; PA provides algorithms from speex [17:24] you'll need to see my bzr branch changes [17:24] Audio arch was always a bit of a headache in linux, wasn't it?! Hope pa will finally resolve this. [17:25] heh, the stack goes much deeper than PA [17:25] and even OSSv4 vice ALSA [17:25] it's a problem from the hardware itself upward [17:25] humbolto: cf. http://bazaar.launchpad.net/~ubuntu-core-dev/pulseaudio/hardy/annotate/crimsun%40ubuntu.com-20080424222750-jpm40x8ljxzbduod?file_id=0052use_alsa_dmix_ds-20080420175647-jra1ekhbq4xbialk-1 [17:26] Pulseaudio seems to use alot more cpu [17:26] h3sp4wn: it definitely does. [17:36] crimsun: The link you pasted shows how to edit the source to get it working with dmix. Now it seems I can easily apply the dmix section to /etc/pulse/daemon.conf, but how to prevent the unloading on idle? Is there an unload directive as well? [17:38] humbolto: why prevent unloading? [17:38] humbolto: (and it would be /etc/pulse/default.pa) [17:39] Because I get a click every time it needs to load something (don't know what exactly) again. [17:40] that's an alsa-driver issue; we're obviously missing twiddling a register [17:40] All in all, without any bad feelings I have to state, that I had better realtime audio performance on some 300mhz 2.4-ck machine than I have nowadays on my 1.8 ghz hardy. [17:40] just comment out the module-suspend-on-idle line [17:40] ok [17:42] I get another 20/30 frames a second with ut2004 using oss4 instead of pulse [17:42] there's a lot of optimisation remaining for pulse, true [17:43] and even more for linux kernel io scheduling! [17:44] well, yes, but there's no perfect scheduler [17:44] [that fits all workloads] [17:44] As said, I had a 300mhz 2.4-ck box, which played perfectly under any load I put at it, but hardy stutters as soon as I start copying some files. [17:45] are you using -17- from hardy-proposed? [17:45] (I use -17-server for a couple reasons) [17:45] crimsun: what reasons? still on -generic [17:46] 16-generic or -17-generic? [17:46] 16 [17:46] eh, yeah, that's the kernel config issue. [17:47] you need to dist-upgrade to -17-generic in hardy-proposed [17:47] what kind of issue [17:47] anyhow, I need the deterministic timers in -server [17:47] humbolto: bug 188226 [17:47] Launchpad bug 188226 in linux "Kernel should use CONFIG_FAIR_CGROUP_SCHED" [Low,Fix committed] https://launchpad.net/bugs/188226 [17:50] 188226, all clear. But what are deterministic timers? [17:51] humbolto: -generic uses dyntick [17:51] (you can search lwn.net or kerneltrap.org if you need a good explanation of it) [17:51] Is there anyone in here running Cadence stuff on Ubuntu (I am using a centos 3 xen at the moment) [17:51] dyntick I remember. [17:58] what is the reason for not daemonizing pa? LTSP? [18:03] The stuff 4front are doing for the sunray thin clients sounds interesting. [18:04] cannot imagine pulse is usable sensibly for a large number of clients [18:06] What exactly are they doing? === gnomefre2k is now known as gnomefreak [18:09] They want a version of oss that can scale to 100k clients http://mail.opensolaris.org/pipermail/opensolaris-arc/2008-January/005880.html [18:10] and run over the network [18:11] reading [18:15] h3sp4wn: sounds somewhat interesting. But I believe using the uid for path determination is a design flaw. What if you connect with the same uid form more than one location at a time. that should be supported too. [18:21] I am only bothered about listening to one thing at once 99% of the time [18:22] humbolto: I don't see how your test case hasn't already been accounted for [18:22] s/test/use/ [18:23] h3sp4wn: It is not so much about listening to music. But imagine having more than one Thinclient terminal for different tasks and you have simple accustic email notification on the one while you want to watch a flash video on the other. [18:23] crimsun: It has been in pulse! [18:24] crimsun: not so much in the oss4 sunray document pasted by h3sp4wn. [18:25] I still have this one to many clicks in pulse even after enabling the system daemon and realtime support. Have to check if it works differently with the -17 kernel [18:26] don't have this with direct alsa access. (tested from the gnome sound config dialog) [18:31] humbolto: enabling RT access doesn't directly affect the symptom [18:31] humbolto: because it doesn't directly affect the _scheduling_ grouping [18:32] it's a subtle difference, and many people don't realise it [18:32] crimsun: I know. Will try in a sec with -17 [18:32] crimsun: I do, read the bug report. [18:32] which bug report? [18:33] crimsun: The one you pointed me to. I did read it. [18:33] which of the two? [18:34] crimsun: bug 188226 has a very good bug description [18:34] Launchpad bug 188226 in linux "Kernel should use CONFIG_FAIR_CGROUP_SCHED" [Low,Fix committed] https://launchpad.net/bugs/188226 [18:35] crimsun: while I am not completely sure if this actually has an effect on my system. We will see. [18:37] it will [18:37] get back to you in a minute on this [18:44] crimsun: sorry that I have to report, -17-generic does not change a thing. minimal if at all. [18:45] crimsun: the 440hz sample starts to play with a click (that is normal on my soundcard). then plays for about 200ms, then there is the next click. and one when I stop the sample (that again is the same with direct alsa) [18:48] humbolto: 1. make sure the user does /not/ have RT privileges [18:49] humbolto: 2. confirm by looking at /proc/version_signature that you're using -17-generic [18:49] humbolto: 3. what changes to default.pa have you made? [18:49] uname -r 2.6.24-17-generic [18:50] humbolto: ideally, I need this test done on a stock default.pa [18:51] crimsun: Set everything back to default. [18:51] crimsun: same same [18:51] I try to record the sound [18:52] Except for the unloading part, this one I disabled [18:52] no, please use the stock default.pa [18:52] Does not make no difference. but I will [18:53] it does make a functional difference, and we need to eliminate all those. [18:53] it may not make a symptomatic difference, but that's irrelevant here. [18:53] ok [18:56] http://pastebin.org/36635 [18:57] right, and the sound clip? [18:57] one moment [18:58] (meaning of your recording the sound) [18:59] what is a pulse supported audio recording software? since pulse grabs all input as well, gnome-sound-recorder does not seem to work [19:00] parec [19:00] ok, seems to work now, just did not like wav [19:00] I will be back in several hours; just leave me a URL with the clip [19:40] crimsun: http://www.file-upload.net/download-856883/pulse.flac.html [19:40] crimsun: http://www.file-upload.net/download-856883/pulse.flac.html [19:40] crimsun: http://www.file-upload.net/download-856880/alsa.flac.html === marko_ is now known as marko-_- === ktogias_ is now known as ktogias [20:38] my usb hub is missing. how can I restart the usb subsystem and force it to redetect the hardware ? [20:53] hello, does anybody know some alternative to app "plan" or "tomboy", somewhere where i can manage tasks, in calendar,when to do what? [21:23] http://ars.userfriendly.org/cartoons/?id=20071216 === Luckrider_ is now known as Luckrider [21:29] hy [21:29] does anyone know how to bind space key in nautilus to launch my own script? [22:13] hi room :-) [22:15] i was wondering if anyone on here could be really kind and provide me with some guidance on how to use irc, i am a dyslexic linux newbie :-) [22:17] you're using it right now \o/