[00:04] It also occurs under the nv driver [00:42] nullack: ok [00:42] Jazzva, do you see that too ? http://ubuntuforums.org/showthread.php?t=909641 [00:42] nullack: try to downgrade nspluginwrapper [00:42] to the hardy version [00:42] Jazzva, (i do) [00:43] nullack: and sudo apt-get install --reinstall nspluginwrapper [00:43] asac: will test and report result [00:43] asac: before or after downgrade? [00:43] nullack: sorry ... last line should reed: [00:43] flashplugin-nonfree [00:43] ;) [00:43] asac right [00:43] after downgrading nspluginwrapper [00:43] :) [00:44] fta: x86? [00:44] fta: goes away by downgrading nspluginwrapper? [00:44] (and reinstalling flashplugin) [00:46] i don't have nspluginwrapper [00:46] fta: could you please try it? [00:46] and see how much worse it is ;) [00:46] on my laptop (amd64) i have it (obviously), it's sucking far too much cpu [00:47] fta: yeah. still i would be interested in x86 experiences ;) [01:00] fta, yep... from time to time [01:01] I thought it was related to nspluginwrapper, but someone reported that it happens without nspluginwrapper too [01:02] i see it a lot in prism (xul 1.9), not that much in ff3.1 [01:02] maybe it's related to fx? [01:02] I have 3.0.1 [01:03] fta, bug 250769 [01:03] Launchpad bug 250769 in xulrunner-1.9 "Firefox opens new popup window when leaving any page with swfdec content" [Undecided,Confirmed] https://launchpad.net/bugs/250769 [01:04] ok so dropping windowless support fixes nspluginwrapper quite well here [01:04] That's what I'm also seeing, beside flickering, and the rest [01:04] good to know that there is at least that option if w dont find why its so damn slow [01:04] so damn slow? npw/ [01:04] ? [01:05] Jazzva: flashplugin ;) [01:05] nonfree [01:05] Jazzva, i'm not seeing this with flash,, i do see a window quickly appear when i close a tab containing an embedded totem [01:05] its slow as hell here with nspluginwrapper 1.1.0 ;) [01:06] well, the video is working here nicely with 1.1.0... [01:06] though, only when I watch directly on youtube :) [01:06] http://wetter.rtl.de/ ? [01:06] the card on the left should animate when you hover the mouse [01:06] err map not card ;) [01:07] it does here. [01:07] same here... a bit slower, but not painfully slow [01:07] it's fine, fluid [01:07] fta: do you have nspluginwrapper? [01:07] (though I'm still on celeron 1.7) [01:07] Jazzva: which driver? [01:07] no, i'll try but not tonight [01:07] hmm... let me see :) [01:08] for nspluginwrapper with windowless + ati driver kills it [01:08] on amd64 [01:09] it puts my cpu to the roof [01:09] I think it's open-source driver for ati... forgot it's name [01:09] Jazzva: radeonhd? [01:09] i had that before ... now i have ati as someone said that 3d works there [01:09] umm... that doesn't sound familiar :) [01:09] Jazzva: then its just "ati" most likely [01:10] like what i have [01:10] Could I get some help on downgrading the package please. Ive edited my preferences: [01:10] nullack@PPP:/etc/apt$ cat preferences [01:10] Package: nspluginwrapper [01:10] Pin: version 0.9.91.5* [01:10] Pin-Priority: 1001 [01:11] And added deb http://archive.ubuntu.com/ubuntu hardy main restricted universe multiverse to sources.list [01:11] nullack: just download the deb from launchpad [01:11] and then [01:12] sudo dpkg -i nsplugin*.deb [01:12] ;) [01:12] Ok, so we dont pin and fiddle with preferences [01:12] i wouldnt do that [01:12] roger [01:12] its just for tesing ;) [01:12] testing [01:22] asac Im pleased to report that fixes it [01:22] nullack@PPP:~/Desktop$ apt-cache policy nspluginwrapper [01:22] nspluginwrapper: [01:22] Installed: 0.9.91.5-2ubuntu2 [01:22] Candidate: 1.1.0-0ubuntu2 [01:22] Version table: [01:22] 1.1.0-0ubuntu2 0 [01:22] 500 http://archive.ubuntu.com intrepid/multiverse Packages [01:22] *** 0.9.91.5-2ubuntu2 0 [01:22] 100 /var/lib/dpkg/status [01:23] nullack: so what are your exact symptoms? [01:23] On the intrepid revision the symptoms are [01:23] 1. When browsing flash sites such as tv.com the browser starts reading the page [01:23] 2. It will then vanish [01:24] 3. Sometimes I get a segfault message into the sys log but its rare [01:24] stop ;) [01:24] what vanishes? the browser? [01:24] Yes, FF [01:24] If I restart FF, FF knows it crashed and presents the restore or start new session GUI [01:24] ok. that sounds like you had 1.1.0-0ubuntu1 installed when you installed flashplugin-nonfree [01:25] did you try to --reinstall flashplugin-nonfree when you saw those issues? [01:25] Yes, I said that on your configuration it works :) [01:25] huh? [01:25] Downgrading to hardys version of the wrapper and reinstalling the non free flash works [01:26] nullack: yes. thats not the question [01:26] The Intrepid revision of the wrapper does not work [01:26] please upgrade to latest intrepid nspluginwrapper [01:26] then --reinstall flashplugin-nonfree [01:26] right doing that now [01:26] ok [01:26] i fixed most painful cases in ubuntu2 ... so when you didnt --reinstall on that version you might still see the issues from ubuntu1 ... which actually had exactly your symptoms [01:27] at least when it always crashes for you [01:29] Ok, youve hit the root cause there asac :) People who have been testing Intrepid for sometime did not get the reinstall of flash and ended up with an invalid config which is probably why Brian couldnt replicate [01:29] true [01:29] I no longer have the always crash [01:29] ok good [01:29] But, the new bug of flickeing flash is pretty bad :) [01:30] well. actually i didnt trigger the --reinstall yet because i wanted to sort out issues like ubuntu1 [01:30] unfortunately this kind of slow feedback has some colleteral damage ;) [01:30] nullack: but that also appears to happen without nspluginwrapper on x86 [01:30] so its a flash or driver issue [01:31] Im happy to work closely with you for quicker feedback - its sometimes hard to go from a bug report to knowing who to talk too, and not to feel like Im pestering the devs with my bugs [01:31] nullack: please comment on the bug so others can recover their stale ;) [01:31] I can do another test usng the NV driver on te flickering thing [01:32] commenting now [01:32] thx [01:33] nullack: how bad is performance for you now? [01:35] Its higher than what it was. Im on an AMD Sempron single core, scrolling is the worse on sites like tv.com, conky shows 85-95% utilisation while scrolling [01:36] Sitting at the top of tv.com Im floating around 50% with no scrolling [01:37] nullack: which process is taking the cycles? [01:37] the firefox or the npviewer process? [01:37] Xorg takes the most, then npviewer [01:38] nullack: whats the ratio? [01:38] like 2:1 or more like 10:1 ? [01:43] One sec I will u/l a file showing my conky which is pretty descriptive [01:45] npviewbin just crashed, sending apport [01:47] asac : http://www.fileqube.com/shared/dGrpZBSMn93509 [01:47] asac : its a png showing my conky so you can see details [01:48] You can view it fullsize or download [01:51] asac: In the cpu utilisation history you can see the graph at around 50% which was the time I was sitting at the top of the tv.com page not scrolling [01:51] nullack: as long as firefox doesnt crash its most likely just flash crashing [01:51] asac: Then you can see in the cpu graph where its maxing out/close to it where Im scrolling [01:52] The conky shows cpu top and mem top for you [01:54] ok i think i made nspluginwrapper fast enough here to see the flickering ;) [01:57] but i think that hack added more instability to it [01:57] then it helped ;) [01:57] :) [02:04] asac : I can confirm the flickering on tv.com is there for both the NV and NVIDIA drivers [02:04] Im going back to nvidia, gdm restarting [02:12] asac : Im raising a bug on the flickering - is the correct package for it the flash plugin or the wrapper? [02:14] nullack: it happens with nspluginwrapper too (according to fta) [02:14] nullack: but we cannot really fix flash plugin bugs [02:14] so no real need to file bug on that package [02:14] I could setup an upstream bugwatch [02:15] Ok Ill take it uo stream [02:15] yeah [02:15] *up [02:15] if you can really link a launchpad bug to an adobe ticket then go for it [02:15] Ok Ill create one upstream and bugwtach it here in LP [07:13] im here if needed im just working on some scripts (3) for addressbook [07:37] * gnomefreak loves the new log out dialog but it would be great if it worked [11:07] asac: you mean sound in flash? [11:07] that problem? [11:09] blacklist snd_pcsp causes no sound if you comment it out you get system sound but no speaker sound [11:10] * gnomefreak waiting for g/f to get out of shower [11:10] gnomefreak: no ... i mean general flash issues if you are running nspluginwrapper [11:10] oh im not running nspluginwrapper since its not "needed" on 32bit [11:10] i can test it when i get home [11:11] if i get out of here sometime this year === asac_ is now known as asac [14:39] Jazzva: fta: ok topic one of today will be "packaging a new ubufox release" ;) [14:40] then i try to be innovative ... maybe writing a webbrowser with xulrunner ;) [14:40] asac, ouch... I forgot about that :( [14:40] When is it? [14:40] 19 UTC [14:40] I won't make it... I'm presenting my project at 18:15 UTC [14:40] sure ... maybe lurk ;) [14:41] good luck ;) [14:41] Well, if I can stay at comp. center at school... I should finish by 19UTC [14:41] asac, thanks, I'm gonna need it. I'm scared of bugs that I didn't notice, and there is a probability there are some... [14:41] :) [14:42] Jazzva: in class presentation nobody has problems with things not working ;) [14:42] at least thats my experience [14:42] what matters is that the talk absorbs them ;) [14:42] crashing prototypes can even be good to get some fun into it ;) [14:42] Not a class presentation... I'm presenting it to teacher's assistant, and then I get points on it... usually it's either 0, or 20... nothing in between [14:42] :) [14:43] ok. i wouldnt be too scared;) [14:43] though a little tension is good to get high-performance ;) [14:43] Anyway... sorry for not mentioning this earlier... I was busy working on it, so I forgot about everything else [14:43] little? heh :) [14:44] good luck with the presentation... I'll try to come if I can :) [14:44] at least to lurk :) [14:44] Jazzva: then join #ubuntu-classroom now ;) [14:44] now? [14:44] Jazzva: well. otherwise you will forget and cannot look ;) [14:44] Isn't it at 19:00 utc? [14:44] at least that would happen to me ;) [14:45] I won't be at home ... I'll be at school :) [14:45] ah. thought you have your IRC running at home [14:45] most of the time, yes... :) [14:45] ah... I can read backlog then :)... easier than looking at logs on the net [14:46] off to finish writing the documentation for this project :) [14:46] have fun :) [15:11] asac, sorry, i'm not much into this those days, I'm having a hard time at work (heavy load). I hope i could clear this up in a week or so [15:12] fta2: that was just for your interest anyway ;) [15:12] not like a request for help :) [15:14] asac, i don't want to give the impression that i'm no longer interested, I am, i'm just able to contribute less than usual but it is temporary [15:14] fta2: thats all fine. you are doing a lot of work [15:14] even when you dont have time ;) [15:15] lets hope its temporary ;) [15:17] i take on my night hours, maybe too much, my body reminds me it needs more sleep :P [15:17] meeting in 15 min; i need to move. cu++ [16:10] asac: somebody talked to you aboug bug 262693 yesterday right? [16:10] Launchpad bug 262693 in flashplugin-nonfree "Flash not working: Intrepid, 2.6.27, FF3" [Undecided,Confirmed] https://launchpad.net/bugs/262693 [16:11] I was wondering if it was a duplicate of something or if we can close some other tasks on it [16:42] bdmurray: looking [16:44] bdmurray: adjusted stati properly [16:45] stati - heh [16:45] asac: thanks [16:45] bdmurray: i changed my mail filters [16:45] i wont see anything i am not subscribed to atm [16:45] i will also include triaged and in progress in that mailbox [16:46] but i want to ramp up step by step instead of busting my mailbox making even that approach void [16:46] bdmurray: so to summon me just subscribe me to bugs [16:46] this will make me appear quite instantly [16:46] asac: okay, thanks for letting me know [16:47] oh ... assigned bugs i will see too ;) [16:47] By the way do you know of any tools to validate /etc/network/interface files? [16:47] bdmurray: syntactically? [16:47] Yes [16:47] bdmurray: what use case? [16:47] otherwise ifupdown should complain i guess ;) [16:48] we have a parser in network-manager which could be used to make a verifier out of it [16:48] To check the files attached to bug reports [16:48] Alberto wrote one for xorg.conf and I've hooked it up to python-launchpad-bugs [16:48] bdmurray: i think the most frequent problems are not syntax problems, but semantical problems [16:48] which are really hard to detect [16:50] Okay, I was trying to think of other files attached to bug reports that might benefit from a similar process [16:51] asac: I could query the database for all the 'interfaces' files so we could get an idea of how many have syntactic errors though [17:16] bdmurray: ill think about it [21:33] asac: ok here I am, would you like to give me some pointers now ? That way I can work a bit on something this we [21:34] asac: The DistributedDevelopment project is so ambitious, its pretty mind blowing [21:36] kaaloo1: agreed [21:36] ill be off in a few. but to give you the pointers lets look at ... hmm ... gnash :) [21:37] kaaloo1: the gnash full-source ubuntu tree is: lp:~gnash/gnash/ubuntu/ [21:37] most likely the other plugins wont have a bzr tree. [21:37] asac: it doesn't work for me, I installed the gnash plugin but I don't see it [21:37] but the changes are similar [21:37] kaaloo1: true [21:37] thats what we need to change ;) [21:37] asac: :) [21:37] kaaloo1: you probably have flashplugin-nonfree installed? [21:38] asac: ok sounds cool then, yes I do [21:38] kaaloo1: so the old way how plugins were managed is by using alternatives [21:38] have you heard of update-alternatives? [21:38] asac: yes, I've used it to switch java implementations [21:39] ok basically alternatives are links created by the admin user that are system wide [21:39] for instance: [21:39] all flah plugins use: [21:39] /usr/lib/xulrunner-addons/plugins/flashplugin-alternative.so [21:39] look add ls -l /usr/lib/xulrunner-addons/plugins/flashplugin-alternative.so [21:40] it points to /etc/alternatives/xulrunner-addons-flashplugin [21:40] so the alternative name is "xulrunner-addons-flashplugin" [21:40] so in the old model when you wanted to use gnash instead of adobe you would do: [21:40] sudo update-alternatives --config xulrunner-addons-flashplugin [21:40] asac: right [21:41] the new model wants to eliminate that alternative [21:41] well. actually for some reasons we want to keep it, but we want to install all plugins also in firefox-addons/plugins [21:41] asac: wow there are a bunch of bad links in my /etc/alternatives [21:41] why not xulrunner-addons? because there are applications that use xulrunner that dont have ubufox [21:42] and thus we want to keep the option to switch the plugin through alternatives [21:42] let me check something ;) [21:42] kaaloo1: did you branch the gnash branch? [21:43] asac: not yet, give me a sec [21:44] asac:ok [21:44] asac:oh wow not much in there ok, sounds like what we did for ubufox [21:45] asac: I see a post install script that calls update-alternatives for the different browsers [21:46] kaaloo1: right. [21:46] kaaloo1: so what we want (afaict) is to keep that [21:46] as it is [21:46] on top we want to add a link [21:46] asac: ok [21:47] the link should point to /usr/lib/firefox-addons/extensions/ubufox\@ubuntu.com/plugins/ [21:47] kaaloo1: so for gnash ... create a mozilla-plugin-gnash.links file [21:47] (look at klash.links) [21:48] kaaloo1: oh [21:48] if there is no changelog entry with UNRELEASED in debian/changelog [21:48] asac:ok, didn't know it could be done that way [21:48] it means that the tree is "closed" [21:48] so as a first commit open it up [21:48] like: [21:48] dch -i -DUNRELEASED [21:48] asac:closed because of a feature freeze ? [21:48] and then commit with "* open tree for packaging" or something [21:49] kaaloo1: no ... because when the topmost changelog has "intrepid" or "hardy" [21:49] that means that the commit you are looking at is a commit that was uploaded [21:49] so good practice is to create the changelog entry like above [21:49] to open the tree [21:49] asac:ok but I dos see an UNRELEASED as the first changelog entry [21:49] (close doesnt really mean: closed until someone approves it to be opened) [21:49] kaaloo1: which version is taht? [21:50] kaaloo1: yeah right. [21:50] then the tree is open [21:50] and you can just add your changes to the changelog [21:50] kaaloo1: usually you mark your changelog entries with: [21:50] ok I'll try it [21:50] [ you name ] [21:50] e.g. [21:51] [ you name ] [21:51] * change 1 [21:51] * change 2 [21:51] kaaloo1: and remember to use debcommit ... which does magic things in case you close a bug in changelog [21:51] (which you probably dont do right here ;)) [21:51] but anyway :) [21:53] asac: right, I followed the packaging course on you tube, and I did some work on the groovy package, I learned about LP: and Closes: [21:53] asac: Thanks a lot !! I will do that then and push it on lp so you can review it [21:55] kaaloo1: rock on! [22:21] sort of back... RMI thing that worked here didn't woork at school... so I have to fix it by 15th september. [22:21] asac, how did the presentation go? :) [22:23] Jazzva: quite well ;) [22:23] given that i was completely underprepared at least ;) [22:23] i talked again too long about mozillateam in intro [22:23] so we didnt get to writing a quick xul webbrowser [22:24] oh, well... [22:24] good that it went quite well... I'm gonna look at the backlog [22:24] what didnt work about RMI? [22:25] asac, dunno... since it worked here. so I was shocked :) [22:25] any exception? [22:25] asac, sorry... I was a bit off for a moment [22:26] yeah, client-side reports it can't locate the stub class [22:26] so a ClassNotFoundException? [22:26] mhm [22:27] it happened on my computer [22:27] i'm not sure if passing java.rmi.server.codebase to jvm helped, but it seemed to work after that [22:27] but that didn't work at the school [22:28] and the TA said that java.rmi.server.codebase shouldn't solve that, and that it should work without it, too... so, i got confused :) [22:28] Jazzva: is it a classnotfoundexception or something else? [22:28] it's a ClassNotFoundException... [22:29] well, first it reports UnmarshalledException (I think), and says that it was prodused by ClassNotFoundException [22:29] Jazzva: same jre version? [22:29] no... I have 1.6.0_07, at school it's 1.6.0 [22:30] but that shouldn't really matter... it only differs in subversion [22:30] right [22:31] it also might be that I didn't know how to point to java.rmi.server.codebase at the server-side [22:31] Jazzva: quite some time since i used it, but for me it was quite simple [22:31] (if that actually matters) [22:31] and i cannot remember java.rmi.server.codebase ever being used [22:31] here it was file:///home/sasa/blabla/code [22:31] Jazzva: what you need is to have the stub classes in a jar that is in classpath [22:32] lemme check that [22:32] Jazzva: do you dynamically want to download the stubs? [22:32] thats the only purpose i can find of that property [22:32] i doubt that you want that [22:32] I'm not sure :) [22:33] i doubt that you want that [22:33] Jazzva: what you want is to use rmic to produce stubs and skels [22:34] and include the stubs in the -client.jar [22:34] and the skels in the -server.jar [22:34] but maybe thats toooo old fashioned ;) [22:35] I'll play with it a bit... [22:35] I'll go to school these days and to try to make it work there... to see what's the problem :) [22:36] thanks for the suggestions :) [22:36] Jazzva: my gues its a firewall issue or something [22:36] most likely the download of the stub doesnt work there [22:36] because your client cannot find the server [22:36] are you using rmiregistry? [22:37] to discover the server? [22:38] yep [22:38] well, I don't run rmiregistry, but call Registry.newRegistry() (I think that's the method) [22:38] but it does the same as the call of rmiregistry [22:39] Jazzva: yes. rmiregistries are either contacted explicitly or they are found through network broadcast [22:39] if network broadcast doesnt work it wont work [22:40] hmm, I would say this was explicitly. I called Registry.newRegistry() on server-side, and Registry.locateRegistry(host) on client-side [22:41] ok [22:41] and java.rmi.server.codebase on server side? [22:41] it might be the firewell... [22:41] *firewall [22:41] Jazzva: did you try on the same host? [22:42] nope... but I think that at sometime TA mentioned something about the ports we're allowed to use [22:43] anyway, I'll sort it out somehow :) [22:45] Jazzva: try on local host [22:45] on local host it works... [22:45] and tell the TA to stop wasting your time ;) [22:45] lol :) [22:46] I also tried last night with a friend... and it worked, too [22:46] ask him why this excersize is about proving him that rmi works [22:46] no need to show that it works on different hosts [22:46] (on the internet) [22:46] just painful and wastes time ;) [22:46] Wow, a chan just for Ubuntu Mozilla folk? [22:46] :) [22:46] Jazzva: well. then its certainly a port issue in the network [22:47] Lns: yes. people feel offended by good crack ;) [22:47] asac, I suppose... I'm almost sure that he said once we can use only five ports... [22:47] asac: .....ok [22:47] but not in this exam term... I'll have to see [22:47] * Lns runs away [22:47] Jazzva: tweak the rmi port then [22:47] Jazzva: should just be a -D on both sides [22:47] -D? [22:47] I think i can just do [22:47] Lns: whats up? [22:48] newRegistry(port) [22:48] Can anyone comment on this bug I just filed? https://bugzilla.mozilla.org/show_bug.cgi?id=453704 [22:48] locateRegistry(host, port) :) [22:48]