[00:00] fta: i think its our problem [00:00] cdbs orders might be difference [00:00] ? [02:22] <[reed]> asac: ping [02:36] hello [02:38] anybody there? [04:43] <[reed]> asac / fta: committed like 7 of glandium's patches... wonder if any of them obsolete some of your patches ;) [06:06] asac: you around or is it still too early? [06:07] NCommander: you have email ;) [06:09] our link is fixed as well [06:10] s/our/your [07:25] !info sunbird hardy [07:25] sunbird (source: lightning-sunbird): Sunbird stand-alone Calendar. In component universe, is optional. Version 0.7+nobinonly-0ubuntu2 (hardy), package size 7792 kB, installed size 23212 kB [07:26] !info sunbird intrepid [07:26] sunbird (source: lightning-sunbird): Sunbird stand-alone Calendar. In component universe, is optional. Version 0.8+nobinonly-0ubuntu1 (intrepid), package size 7917 kB, installed size 23908 kB [07:51] its almost 3am :( [08:01] @time new_york [08:01] Current time in America/New_York: October 21 2008, 03:01:25 - Next meeting: Community Council in 3 hours 58 minutes [08:01] Current time in America/New_York: October 21 2008, 03:01:25 - Next meeting: Community Council in 3 hours 58 minutes [08:02] ok im gone maybe ill be back for meeting but i really need sleep === asac__ is now known as asac [11:36] Hello? [11:37] anybody there? [11:43] Does anyone can help handle bug #286906? [11:43] Launchpad bug 286906 in xulrunner-1.9 "Unable to use libmozjs.so in an application, because of library path problem." [Undecided,New] https://launchpad.net/bugs/286906 [13:28] asac: hi :) [13:30] asac: I'm here to know which features are new and documentation ready on https://help.ubuntu.com/community/NetworkManager [13:30] asac: which ones still hold good in ubuntu 8.04 and ubuntu 8.10 and which things are new. [13:42] firefox is routinely getting stuck for several minutes when I come back to it. It's not paging in (it's already in memory according to top), and strace shows it calling getTimeOfDay 4 times, then write, and so forth over and over again === shirish1 is now known as shirish [13:53] shirish: wait for a few minutes pleaes ;) [13:54] asac: sure, will do [13:56] asac: please lemme know when you are free so we can start atleast knowing what spaces needed to be started to do stuff. [13:57] :/ [13:57] 1gig of resident memory used on a session that usually only uses 600mb [13:58] it's bouncing around though, up to 1g, down to 948m slowly back up to a gig, down to 946m [13:59] (2.5gb in this machine, there's 100mb not even used yet, so it's not thrashing that I can see) [13:59] I can get 1 click every 3 minutes through [14:00] if I middle click a tab, the windows will go grey, and 3 minutes later (or so) it'll close it [14:00] cwillu: try to disble all your extensions is the first advice i ususally give [14:01] cwillu: if that doesnt help try to disble your adobe flashplugin [14:03] asac: is the practise of adding $USER to netdev still desired or is that deprecated? [14:04] asac: that should have been is the practise of adding $USER to the netdev group still desired or is that deprecated practise? [14:13] shirish: huh? [14:13] shirish: oh because of dhcp? i dont think so. no [14:13] dhcp is run as root now [14:14] shirish: let me check that page now [14:14] asac: I have been having issues as to which is the correct way to configure stuff for having things managed by network-manager [14:14] shirish: Installation -> maybe we should make it obvious that its installed by default in hardy, intrepid [14:14] right [14:15] asac: hi [14:15] shirish: untick the "Enable this connection" .... thats all outdated, but even before it was called "enable roaming mode" [14:15] asac: did you get change to check my package? [14:15] nxvl: no i didnt get any coins ;) [14:15] :-P [14:16] nxvl: i have two things on my list. then i will do that [14:16] asac: hang there a minute, brother, writing the first thing about hardy and Intrepid. [14:16] shirish: well. so gutsy + hardy the checkbox is called "enable roaming" or something [14:16] asac: ok, did you have the link to the package? [14:16] in intrepid that administration menu doesnt exist anymore [14:16] by default [14:17] you need to install the network-admin thing from gnome to get that gui back [14:17] (usually NM should be able to deal with that now) [14:22] asac: in this version of nm-applet it says "enable networking" [14:22] shirish: sorry. there is confusion [14:22] shirish: previously that dialog in system was the gnome dialog [14:22] thats gone now and replaced with nm connection editor [14:23] shirish: the "configuring section is mostly about the /etc/network/interfaces " section [14:24] asac: can you please first check the installation section, do you think that is correct or incorrect? [14:24] shirish: i really think we cannot keep the documentation for all (dapper - intrepid) on the same page [14:24] asac: the part of ubuntu 8.04 and ubuntu 8.10 installation [14:24] shirish: ok let me lok there [14:25] but note that we should certainly do someting about the configuring section [14:25] especially what to do for old setups [14:25] shirish: i think NM is installed by default since feisty [14:25] asac: think the same but then how we do that, I'm not a good shot at doing stuff from scratch, but that could be taken up at the ubuntu-doc mailing list [14:26] asac: ok, what about the bit about the applet, you think that is wrong as well ? [14:26] shirish: the nm-applet doc isnt right. its a tray thing. the only thing the user can do is enable/disable it in the gnome session [14:26] Preferences -> Sessions [14:27] but i don tihnk we should mention that [14:27] nobody wants to disable it [14:27] asac: right [14:27] shirish: imo. we should state that its installed by default. then we should just document the features [14:27] and add a FAQ or Troubleshooting section [14:27] asac: right [14:27] with the most common pitfalls [14:28] shirish: so the main elements (which would deserve a short intro) are: [14:29] 1. the applet [14:29] 2. the connection editor (which is what you find in System now) [14:29] not sure how to document the applet. usually thats self explaining. to connect to a network, just click on it [14:30] also might be important to mention that NM auto creates connections on a best effort base, but if those connections dont match your need (e.g. no dhcp), then you need to use the connection editor to configure things [14:30] for the connection editor we should do multiple paragraphs: [14:30] asac: right hang on, putting up things as you speak. [14:31] 1. introduce the available connection types (wired/wireless/mobile broadband/VPN) [14:31] 2. document how to create/modify a user connection [14:32] (here we might need a hint that auto created connections are _not_ editable. and modifying them withouth changing names wont be saved) [14:32] 2a. introduce wired/wireless/mobile broadband/ppp tabs [14:33] 2b. introduce ip4settings tab in connection editor (e.g. which type - aka static dhcp, dhcp addresses only, etc.; how to edit routes) [14:34] 3. document how to install vpn support [14:34] 3a. give a quiick overview of the vpn wizards you get in connection editor when the vpn packages are installed [14:34] (not sure if this really needs to be detailed) [14:41] nxvl: do you have test packages in a PPA ? [14:42] nxvl: security updates are kind of critical and since there is no official way to stage them for broad-testing we usually put them in a PPA so w can at least verify the bits that will be released before pushing them [14:43] asac: yes, in my ppa [14:43] asac: nvalcarcel [14:44] nxvl: so i can take that orig and do a test build and get the exact same result ;) [14:44] ok getting the branch here too [14:45] yes [14:45] asac: going slow, can you re-check the Installation part once again and then the configuring part which I have put up atm. [14:45] shirish: yeah. please save what i wrote though ;) [14:46] asac: write, its there in my logs :) [14:46] shirish: the endable sentence can be dropped. the applet is there to manage your networking from the desktop ;) [14:48] asac: did that [14:48] shirish: configuring is completely outdated. needs to be dropped [14:49] shirish: the bits about the interfaces do not apply to most users. so we should add how to deal with interfaces in the troubleshooting section [14:49] oh sorry [14:49] didnt see that you still split dapper et al [14:49] asac: right, for the moment shall I put it as configuring for old versions? [14:50] asac: I dunno, there may be users who may be using and may need this documentation. [14:50] shirish: hardy + intrepid + _gutsy_ [14:50] shirish: its unfortunate that that documentation still exists [14:50] it made lots of users most likely revert our transition where we removed everything from interfaces automatically [14:50] asac: shouldn't that be gutsy+hardy+intrepid in that order or the reverse ;) [14:51] sure [14:52] shirish: add to hardy et al: "if your interfaces are shown as "unmanaged" in network manager applet, please read the troubleshooting section about "TBD" [14:54] asac: by et al you mean Hardy and after that right? [14:54] nxvl: you didnt add the .desktop file to the branch [14:54] please do that ;) [14:54] shirish: yes [14:57] asac: ok please refresh or recheck the page, the first two sections, Installations and the new one which I made about configuration, are you happy with that? [14:57] asac: mm i forgot tu push that change [14:57] asac: one second [14:58] shirish: you still have "ts used to enable or disable the nm-applet for that session. " [14:58] thats just not th epurpose of the applet [14:58] the purose is to "manage and activate/deactive your network connections" [14:58] asac: k, pushed [14:59] asac: right [14:59] asac: what it should be saying, its used to disable or enable networking for the session? [14:59] asac: ah, ok didn't see that last sentence, doing it [15:01] shirish: ok and "Specifically for Hardy and latter," would be "For intrepid or later, ..." [15:02] shirish: TBD means ... to be defined. we should replace that with the link to the troubleshooting section when we have that ;) [15:02] asac: ok cool. [15:03] shirish: imo we shouldnt really make a single document out of it for NM 0.7 [15:03] i mean everything below is completely outdated too [15:03] i mean everything below is completely outdated t [15:03] oops [15:03] ;) [15:03] no issues. [15:06] asac: corrected the former, now what about suspend support, that's also outdated/deprecated as far as Hardy, Intrepid is concerned? [15:07] shirish: i am not even sure how this applies anywhere [15:07] most likely its just dapper and feisty [15:07] but cant tell for sure [15:07] asac: so will put it like that, dapper and feisty (atleast the headings) [15:07] shirish: really. lets split this up in a) documetnation of networkmanager 0.6 (e.g. dapper, feisty, gutsy, hardy) and networkmanager 0.7 (intrepid and later) [15:08] shirish: then we can keep the document in the state as it was before [15:08] (even though its inaccurate for a bunch of things for a bunch of things) [15:08] ;) [15:08] we just run into in accuracies all over the place and fixing everything just because we want to document intrepid is too hard [15:08] asac: how do I do that? Do you want me to create a new page or how? [15:09] shirish: yes. i think we should make one main document that just states: [15:09] documetnation for network manager before 0.7 (dapper,feisty, gutsy, hardy) can be found here: ... [15:09] and for network manager 0.7 (intrepid, and later) can be found here: [15:09] asac: ah ok, that is cool. [15:10] shirish: at least i think that makes sense [15:10] asac: but then this page also needs to be renamed, ok making a new page. [15:15] nxvl: looks good. a final release commit would have helped to identify what was released later though [15:16] nxvl: you can still do that by just running dch -r -Dhardy-security [15:16] or something [15:16] and commit with "RELEASE x.y.z-0ubuntuY to WHATEVERYOURDISTROISCALLED/hardy-security" [15:16] nxvl: ^^ [15:16] nxvl: when that is done i am happy ;) [15:17] our distro is always called hardy [15:17] no -security or -updates [15:17] is always hardy [15:17] asac: have made a page called it networkmanager options https://help.ubuntu.com/community/NetworkManagerOptions is that good enough ? [15:18] asac: but got your point on the release commit [15:18] shirish: err. most likely nobody will look for that [15:18] shirish: imo the top most page should refer to NetworkManager0.6 and NetworkManager0.7 [15:18] nxvl: yeah ... then use "hardy" ;) [15:19] asac: i'm unsure about the versioning schema [15:19] nxvl: i usually say: to ubuntu/intrepid .... you could say to "oem/hardy" [15:19] ;) [15:19] asac: so you want me to integrate this at the NetworkManager doc itself. [15:19] nxvl: just the complete version from the changelog imo [15:19] asac: since i'm having a different base version, but the ubuntu part should stay the same, or change? [15:19] nxvl: you can also move the +nb1 before the +nobinonly [15:19] nxvl: the netbook thin at the end is not needed [15:19] just 8.04.1 should be enough [15:19] asac: is nxvl your new padawan? ^^ [15:20] sebner: intermediate ;) [15:20] kk ^^ [15:20] sebner: kind of, i still don't understand asac'c packaging methods [15:20] :P [15:20] nxvl: who does that .... :P [15:20] right [15:20] asac: oh! ok then [15:21] asac: can you re-check the NetworkManager top-post section, is that good ? [15:23] yeah let me install abrowser agin ;) [15:23] dont have a browser right now [15:24] asac: "abrowser" or "a browser" ^^ [15:25] its abrowser [15:25] hihi [15:31] sebner: hi :) [15:32] asac: have put up a new page where I have ported stuff from what we discussed so far, https://help.ubuntu.com/community/NetworkManager0.7 [15:32] asac: using the documentation template, so would have to excuse the stuff not written yet :) [15:35] asac: then how do you want to do it? I thought you wanted to have a new page, altogether, what should be ideal IMHO should be https://help.ubuntu.com/community/NetworkManager has stuff only for NM 0.7 while the rest should be another page altogether, all stuff on the same page doesn't work (becomes too confusing) [15:36] asac: scratch that last line. [15:36] shirish: yes. it would be ok to make NetworkManager 0.7 and give a fat reference to the 0.6 page right on to [15:36] p [15:37] asac: right, that I should have thought of before. [15:37] asac: I know you care for the baby but when does NM 0.7 gets officially matura? ca? [15:38] shirish: well. its ok as its now i think. its just that we want to make 0.7 the default at some point [15:39] maybe its ok for now to link to it from the "main" networkmanager page and keep that at 0.6 [15:39] sebner: never. dcbw said he would hang himself if not released by end of this month [15:39] but given that they now started to do refactoring i doubt that it will happen [15:39] hehe [15:39] kk thx [15:43] asac: https://help.ubuntu.com/community/NetworkManager0.7 that good, the top-most thing? [15:45] shirish: you should say "ATTENTION: if you are use dapper, feisty, gutsy or hardy, read the documentatoin for networkmanager 0.6 [15:45] btw, either NetworkManager (upstream name) or network-manager (package name) please [15:45] shirish: ^^ [15:45] asac: that's what you meant by FAT ;) [15:47] yeah ;) [15:47] shirish: also put the same on the other page for intrepid [15:48] shirish: people will most likely find one of these through google [15:48] so we cannot rely on them reading the url ;) [15:48] asac: right, updated the same. [15:51] asac: did that as well. [15:52] asac: ok now for some info. or introduction of the nm-applet and then nm-connection editor [15:52] asac: know any good content for ^^^ [15:54] what about lightning/iceowl-extension 0.9? [15:56] shirish: i wrote my ideas above ;) [15:56] is there anything that isnt clear :-P [15:56] ? [15:56] MechtiIde: didnt make the train for intrepid. jaunty will get it [15:58] jdstrand: it completely missed that we had a reason to celebrate two days ago ;) [15:58] asac: I know! I'm updating CVEs for feisty now :) [15:58] jdstrand: so here my cheers and hugs for succeeding in providing full security support coverage for feisty ;) [15:59] mission accomplished [15:59] now please let go dapper away :/ [15:59] * jdstrand hugs asac [15:59] re dapper> 8 months for you, 32 for me [16:00] jdstrand: yeah. most likely it doesnt make such a big difference for you ;) [16:00] but it certainly is much appreciated here :) [16:00] heh [16:00] jdstrand: you still will receive a bunch of tarballs on every release [16:00] oh yes, I'm well aware ;) [16:00] the same amount actually ;) [16:03] jdstrand: will 8.04.1 be supported till june or end of april? [16:05] asac: I thought april [16:06] jdstrand: wasnt sure, i rememberd that some said that 8.04 isnt LTS, only 8.04.1 [16:07] asac: I'd like to see that quoted-- I certainly never heard that === fta_ is now known as fta [16:14] mozilla bug 435088 [16:14] Mozilla bug 435088 in XPCOM "Check .autoreg in the GRE directory, too" [Normal,Resolved: fixed] http://bugzilla.mozilla.org/show_bug.cgi?id=435088 [16:18] good [16:21] asac, i guess we can drop our bzXXX_gre_autoreg.patch [16:22] yeah [16:22] since its similar ;) [16:22] err identical i mean [16:22] not identical at all. but the goal is the same [16:24] fta: ok then i mixed that up. we should verify that then after dropping [16:25] mozilla bug 412610 [16:26] Mozilla bug 412610 in Startup and Profile System "MAXPATHLEN too small for glibc's realpath()" [Normal,Resolved: fixed] http://bugzilla.mozilla.org/show_bug.cgi?id=412610 [16:27] Hi, is there anyone around to review my intrepid-nomination of bug 286688? [16:27] Launchpad bug 286688 in ubufox "restart notification erroneously keeps appearing after restart" [Undecided,New] https://launchpad.net/bugs/286688 [16:29] maxb: any idea how i can shutdown withput that being rewritten? [16:30] maxb: are you having extensoins installed (besides ubufox obviously) [16:31] No other extensions. I'm away from the computer on which I initially observed it right now, but let me just check if I can reproduce here. [16:32] asac: can you take a look at https://help.ubuntu.com/community/NetworkManager0.7 and see if things are write, specificially the part about nm-applet, nm-connections editor as well as the introduction to the various technologies [16:34] shirish: NM-Applet (a) .... imo the main focus should be on teaching users how to connect to networks and what those entries mean [16:34] and only then discuss what happens on right click [16:34] (e.g. left-click is the main use case) [16:34] How annoying, what was trivially reproducable there does not reproduce here. I shall investigate further tonight when I'm back at the problem machine and update the bug. [16:35] shirish: i think the title "Configuring Devices for Ubuntu 7.10 (Gutsy), Ubuntu 8.04 (Hardy), Ubuntu 8.10 (Intrepid)" is now obsolete and can be removed from the 0.7 document [16:35] shirish: we should later look how we can fix the 0.6 document for gutsy/hardy [16:43] asac: I dunno much about doing stuff for 0.6 simply because I don't have that any of those configurations, also going for dinner so bbiaw :) [16:44] k [16:44] shirish: for now lets focus solely on 0.7 [16:44] let me know ;) [16:51] asac, could you push your ff3 branches ? 3.0.3+nobinonly-0ubuntu2 is not on lp [17:07] suzhe: hi [17:07] hello [17:07] fta: oh yeah. we have a .intrepid branch now ... will push from there to head too [17:07] asac: what's your opinion about bug #286906? [17:07] Launchpad bug 286906 in xulrunner-1.9 "Unable to use libmozjs.so in an application, because of library path problem." [Undecided,Incomplete] https://launchpad.net/bugs/286906 [17:08] asac, at least, update .head, that's what i merge from for 3.1 [17:08] fta: done [17:08] sorry that i forgot [17:08] hmm still pushing up ;) [17:09] seems ftbfs-with-branding-dir is no longer needed, i can't find the root of this patch... [17:09] mozilla Bug 460915 [17:09] Mozilla bug 460915 in General "Build problem when using --with-branding-dir" [Normal,Assigned] http://bugzilla.mozilla.org/show_bug.cgi?id=460915 [17:11] asac, in rev6, i see + renamed debian/patches/brandir-dir-depth-fix to debian/patches/ftbfs-with-branding-dir, do you know anything about brandir-dir-depth-fix ? [17:11] suzhe: well mozjs.so doesnt really have any abi guarantees i think [17:11] which is why we dont provide it in libdir, but in pkglibdir [17:12] fta: yeah [17:12] fta: it was a bug [17:12] (not sure when) [17:12] (not sure if still a problem) [17:12] asac: back, put up some more info. in the nm-applet please re-check in the nm 0.7 page https://help.ubuntu.com/community/NetworkManager0.7 and lemme know what you think. [17:13] fta: did i do a lousy comment in commit? [17:13] i don't have enough history on that particular patch [17:14] asac: AFAIK, libmozjs.so keeps backward compatibility since version 160 quite well. [17:15] asac, ok, it appeared in rev2 [17:15] shirish: those nm applet lines should be accompanied by screenshots i guess [17:16] suzhe: yes. but is there any policy on that? if there is we can make a proper first tier lib out of it [17:16] asac: you mean the one where I give the link to the D-Link stuff? [17:16] shirish: no. every line ;) [17:16] asac: damn, ok cool. will try [17:16] asac: And I don't think the abi will be changed in xulrunner 1.9 series [17:17] shirish: e.g. "he network-manager or the nm-applet is the one which is found in the systray. The icon of two computers, one below to the other on the left-side." -> screen snippet shich shows the tray with NM [17:18] asac: now that your talking about nm, has there been any changes to it so that I have to do some manual configuration lately ? [17:18] asac: I'm not sure if there is any policy on that. [17:18] asac: let me have a check. [17:20] asac: because, nm doesn't show up in systray when I login any more, have to do a /etc/init.d/NetworkManager restart to get it to show up [17:21] niki: if nm-applet is running and NetworkManager you need to restart NetworkManager ... yes. [17:22] s/and NetworkManager/and NetworkManager not/ [17:22] asac: this in on a fresh boot up, so, shouldn't it be started automatically ? [17:23] niki: it should [17:23] niki: so please check if its not running [17:23] if so, post your complete syslog somewhere [17:24] I will [17:26] asac: another issues with nm is that it seems to have "lost" all the interfaces, no wired interfaces (only an usb0 which is defined in /etc/network/interfaces which is greyed out), and no wireless interface, is that a known issue ? [17:26] asac: you just want the systray or the whole desktop there? [17:26] niki: interestee in interfaces are always not-managed by default (same as in gutsy) [17:27] niki: should be "unmanaged" in applet [17:27] niki: if wireless isnt seen, then that device isnt known to hal ... not a NM issue. you might have disabled wireless in nm right-click menu though [17:27] asac: though there is no explicit policy about abi compatibility issue of libmozjs.so, I believe that the abi won't be changed during xulrunner 1.9 lifetime. [17:28] shirish: well. given that each line gets a snapshot the shown section should be as small as possible - while still giving user enough context to spot that [17:28] suzhe: well. believing is good [17:28] we need a policy or upstream needs to use SONAMEs to express how they change abi/api [17:28] asac: right, ok will work on that. [17:30] asac: wireless is enabled, just wonder why hal had dropped "support" for my "Intel Corporation PRO/Wireless 5100 AGN [Shiloh]" as it has been working, but i'll investigate that [17:31] niki: try hal-find-by-capability --capability net.80211 [17:31] asac: can you help ask upstream developer about this issue? [17:32] root@niki-laptop:/etc/default# hal-find-by-capability --capability net.80211 [17:32] /org/freedesktop/Hal/devices/net_00_16_ea_7a_d2_8a [17:32] ^ asac [17:32] asac, where did you handle firefox-trunk before you created the bzr branch? [17:32] branch nick: firefox-trunk [17:32] timestamp: Wed 2007-03-21 18:21:52 +0100 [17:32] message: [17:32] initial firefox-trunk debian/ package [17:32] seems to find it fine [17:32] niki: do you see "wirless networks" in the applet at all? have you checked with right clicking that wireless isnt disabled? [17:33] "wirless networks" -> thats a grey text in the applet drop down [17:33] yep, but in the list it just says "device is unmanaged" [17:34] niki: then you have a configuration for that in /etc/network/interfaces [17:34] ok, so I my not have that then anymore ? [17:34] niki: NM _never_ managed devices in /etc/network/interfaces for you [17:35] niki: or maybe ... can you post your interfaces please? [17:35] well sometime ago it kinda ignored the entry in interfaces, so that I could managed it both places [17:35] asac: I think the solution used by opensuse is good enough. [17:35] fta: isnt there a branch for that in firefox project? [17:36] suzhe: no. people will create packages that link against the libs directly which we do not want unless we know that we can track abi [17:36] had to have the entry because nm didn't manage to associate with my ad-hoc network [17:36] asac, that's rev2 of all ff3 branches, rev1 being the tarball [17:36] sec, i'll paste it in a sec [17:37] suzhe: i men what does it help if it works, but you cannot really maintain this stuff in a distro ... we need a real solution [17:37] asac: then do you have a good solution for this problem? I searched on google and found that there are some other applications that require libmozjs.so [17:37] asac: many ubuntu users reported similar issues. [17:38] suzhe: many? ididnt get to know about that yet. [17:38] do you have references? [17:38] fta: maybye https://code.edge.launchpad.net/~mozillateam/firefox/trunk.dev [17:39] fta: i think that "restart" was done because you wanted to get rid of the tarball cruft in the branch [17:39] might be wrong of course [17:39] asac: http://pastebin.com/d5fcc918c [17:39] fta: hmm. thats that branch [17:39] asac, no, we never restarted, this branch starts like all the others [17:40] yeah [17:40] asac: you can search ubuntu libmozjs.so on google. [17:40] asac, i wasn't there yet at that time so i don't know [17:40] suzhe: only see old noise on the first page [17:40] about firefox broken etc. [17:41] fta: most likely i did all at once and imported that [17:42] fta: the firefox-trunk branch nick was just the name of that branch for me locally at that time [17:42] asac: BTW, any progress with your google chrome/chromium import to LP? [17:42] sebner: jcastro would know more ;) [17:43] asac: for example: http://itsiriuscn.yo2.cn/articles/ubuntu%E4%B8%AD%E5%AE%89%E8%A3%85chmsee.html [17:43] jcastro: pingeling :) [17:43] last i know is that we had a pending request to create a "super" project on launchpad [17:43] lol [17:43] damm, your quite busy asac ;-) [17:44] niki: he is always busy so just ignore it ^^ [17:44] niki: so ... you sure you could use wireless in the NM menu? [17:44] in hardy? [17:44] intrepid [17:45] niki: err. i mean if you could use the exact same config and still select your access points in the NM menu [17:45] on hardy [17:45] asac: well I haven't been able to get it to associate with an ad-hoc network yet, that is why I have the entry in interfaces. [17:46] I bougth this lappy some months ago, and went strait to intrepid, so actually don't know [17:47] niki: ok. then thats a problem of you ;) [17:47] niki: currently every device configured in interfaces is unmanaged in nm [17:47] niki: you can do adhoc stuff in NM directly [17:47] sebner: it's been imported [17:47] not sure why you use interfaces [17:47] jcastro: everything? [17:47] sebner: it needs to be shuffled around namespace wise [17:47] jcastro: is the "master" project set up? [17:47] oh [17:47] no, I never got a response about that [17:48] "chromium" is importing though [17:48] hehe, yeah, I know, had to "skip" hardy because the wireless card wasn't supported, support was first added in 2.6.27 [17:48] jcastro: jcastro you mean "chrome"? [17:48] ^ asac [17:48] jcastro: we really need all parts. otherwise that single import isnt worth much ;) [17:48] yeah it needs to be renamed [17:48] jcastro: could you ping lp folks again? [17:48] asac: yeah [17:49] sebner, i follow chromium quite closely [17:49] jcastro: please dont rename it to chromium ;) ... chromium-project is the "main" project. [17:49] and chrome/ was a subtree [17:50] asac: I'll comment out wlan0, and reboot to see what happens, i'll be back in about 10 15 min. [17:50] jcastro: fta : already building on linux? [17:50] sebner, but there's not much activity for linux, just 1 or 2 commits a week to build various small things already in the tree for linux [17:50] sebner: hahahaha not even close. [17:50] rofl [17:50] jcastro: it'S opensource. fix it :P [17:50] not, most of the code remains to be written/ported [17:51] -not+no [17:51] and now .. who kicks google's ass to do that? [17:51] brb... [17:52] i think they are working on it internally, it's just not public yet [17:52] yeah ... most likely throwing stuff over the wall ;) [17:53] asac: are you guys talking about google chrome or something else altogether? [17:53] chromium [17:54] asac, ping [17:54] fta: what is chromium? [17:54] shirish: chromium ... yes [17:54] Jazzva: ? [17:54] chromium = google chrome minus branding and some features [17:55] Jazzva: i think MOTU freeze will start on 23rd ;) ... whatever we want to get in extension wise should be done soonish ;) [17:55] fta: ah, ok cool. [17:55] Jazzva: also a review of the app-install-data ;) do you know how many changes we have in there? [17:55] http://blog.chromium.org/2008/10/google-chrome-chromium-and-google.html [17:56] asac, is this procedure still right for getting Profile dir? http://developer.mozilla.org/En/Code_snippets/File_I%2F%2FO , section "Getting special files"? I tried it two nights ago in order to fix foxyproxy, but it didn't work, I tested with file.path, but it didn't write anything [17:56] asac, 23rd? Oh... that soon. I'll be quick :) [17:56] Jazzva: yes thats correct [17:56] Jazzva: you have the diff you tried? [17:57] asac, app-install-data will have removal of mime-type for firefox-2 extensions. And there should be few files added for new extensions [17:57] asac, no. It's a one line change, I'll paste it now [17:58] hello jorendorff [17:58] hi [17:58] So, we were discussing https://bugs.launchpad.net/ubuntu/+source/xulrunner-1.9/+bug/286906 [17:58] Launchpad bug 286906 in xulrunner-1.9 "Unable to use libmozjs.so in an application, because of library path problem." [Undecided,Incomplete] [17:58] And I think a big part of the question is whether the ubuntu moz team *wants* to provide a binary ABI that's stable over time [17:59] asac: jorendorff is a developer from spidermonkey team. [17:59] a policy question I know nothing about, hoping the folks here can help [17:59] (just a developer, but hoping I can help) [18:01] jorendorff: well. the ubuntu moz team doesnt have a will on its own. its just that when developers want to develop against mozjs.so they want that lib to be available in the distro [18:01] but unless there are no ABI/API guarantees/policies/procedures, we cannot ship libraries in decent fashion [18:01] asac: Sorry if I don't understand the issues properly -- I really know next to nothing about packaging [18:02] asac, http://paste.ubuntu.com/60599/ [18:02] jorendorff: for instance. we currently build libmozjs.so as part of the xulrunner [18:02] jorendorff: there libmozjs.so is shipped in pkglibdir ... [18:02] I think that's the place where it gets the settings dir, and then it checks it for foxyproxy.xml. If it doesn't exist, it creates new file, if it exists it reads from it. [18:02] this means de-facto that those libs are not available for public consumption ... just for internal use (e.g. xulrunner itself) [18:03] jorendorff: when it comes to decide whether we can make a lib available in libdir /(e.g. /usr/lib) we usually need the lib to use -version-info (or something similar if your lib doesnt use libtool) [18:04] which properly indicates abi/api compatibility [18:04] OR we need a guarantee that there will be no abi/api breakage as long as libmozjs.so is shipped under that name [18:04] the latter is what nspr and nss are doing [18:04] ok, I'll paste this into a bug on our end. [18:05] It looks like we are at least weeks away from providing either of those. [18:05] asac: In this case, can pkglibdir act as the version info? [18:06] suzhe: Right now, my guess is, nobody currently wants to guarantee that there will even be a libmozjs.so file there [18:06] much less anything about its remaining stable in any way over time [18:07] nobody meaning neither the Ubuntu moz team nor the moz xulrunner team nor the moz spidermonkey team [18:07] jorendorff: it's guaranteed by pkg-config file mozilla-js.pc [18:07] Well, that is interesting [18:07] where does that file come from? [18:07] xulrunner-1.9-dev package. [18:08] and I believe it comes from upstream. [18:08] ah, looks like that's coming from xulrunner [18:08] content of mozilla-js.pc is pasted in that bug report. [18:09] I saw it; it's coming from here http://mxr.mozilla.org/mozilla-central/source/xulrunner/installer/mozilla-js.pc.in [18:10] But it looks suspicious to me. I'm pretty sure xulrunner just takes whatever version of SpiderMonkey is in the tree, so it seems wrong to slap %MOZILLA_VERSION% on that (since the two are ostensibly independently versioned) [18:11] jorendorff: will xulrunner 1.9 sticks to a specific version of spidermonkey? [18:13] No, they take whatever version of SpiderMonkey is in the tree -- i.e. the latest [18:14] For all I know the ABI may remain compatible from 1.7 to trunk, for features exposed in jsapi.h [18:14] but if so, as much by coincidence as anything else [18:15] I know that other functions, the ones not exposed in jsapi.h, tend to come and go. [18:15] jorendorff: do you have any policy to maintain ABI compatibility for features exposed in jsapi.h? [18:16] no [18:16] suzhe: at this point the pkglibdir acts as version info [18:16] suzhe: just that it doesnt help for you as it changes on each and every upstream release [18:16] recently we considered changing jsval from pointer-sized to a 64-bit type, for example. [18:16] (which is intended just for things like this) [18:16] asac: now is the time to sign up! https://wiki.ubuntu.com/UbuntuOpenWeek/Prep [18:16] asac: no luck, but I did get a segmentation fault from nm [18:16] niki: no luck with what? [18:16] adhoc? [18:17] adhoc doesnt work for most modern chipsets i think [18:17] (funnily) [18:17] asac: the interface didn't show up [18:17] niki: remove everything from interfaces [18:17] asac: well i'm running adhoc atm [18:17] niki: it will show up for sure after reboot [18:17] niki: good to know ;) .... wlang? [18:17] iwlang ;) [18:18] yep [18:18] thats good news [18:18] didnt work for me on my plain 4965 here [18:18] asac: then, my question is: in which xulunner version range the ABI will keep stable? 1.9.x.x or 1.9.0.x? [18:18] suzhe: we dont know anything. and there are no upstream guarantees. which means we only guarnatee that its stable for each upstream release [18:18] see my paste bin from before, that sets my wlan up for adhoc [18:18] suzhe: which is what we reflect by shipping stuff that way [18:19] (which actually is the official upstream way to ship it ... e.g. run make install on unpatched sources) [18:19] everyone who patches that out of the build system hides the problem and will make it less likely to be fixed [18:19] asac: especially, see the pre-up command i have, I need that else adhoc won't work [18:19] becaues everybody just accepts that people will "just" linke -rpath and then prey ;) [18:20] niki: you can configure in NM that adhoc should share the NM [18:20] asac: Hmm, then my package must be rebuilt every time when xulrunner upgrades. It's really bad. [18:20] niki: NM should support it out of box [18:20] niki: its just that it doesnt even work manually here [18:20] e.g. with iwconfig only for instance [18:21] I did the setup in iwconfig before doing it in ifupdown [18:21] suzhe: right. but thats what upstream suggests. we need to convince spidermonkey developers to take over responsibility and make a properly versioned library out of it [18:21] jorendorff: Can you help fix this issue from upstream, by adding version info to libmozjs.so? [18:21] or release a policy ... or what ever [18:21] but ifupdown needs pre-up iwconfig wlan0 mode ad-hoc [18:22] suzhe: most likely thats not feasible for them they rather want to go the "strict" policy road and dont break abi/api until they go for libmozjs2.so [18:22] asac: OK, I am convinced you guys are doing the right thing. I'm going to ask the xulrunner guys; surely someone has given this some thought [18:22] etc. [18:22] we need a policy [18:22] jorendorff: well. this has been long discussed i think. but you can try [18:22] jorendorff: the xulrunner position is that caring about abi/api isnt necessary as xpcom will properly protect stuff [18:23] jorendorff: but you never know ;) [18:23] yeah, I think the tendency is away from binary compatibility altogether and toward JS-only components and stuff, but I don't know how that fits in with SpiderMonkey [18:23] libmozjs that is [18:23] only thing i can say is that without any of these mechanisms its really hard to ship libmozjs.so ... which would be a shame imo [18:24] Well -- we have some nuts and bolts we have to take care of, and believe it or not we are working on it... [18:24] but boy has it been a long haul https://bugzilla.mozilla.org/show_bug.cgi?id=97954 [18:24] jorendorff: at best try to listen to what bsmedberg thinks. and also how he thinks libmozjs.so can be properly shipped in such a way that it can become a proper top-level lib on linux systems [18:24] asac: the other end of the adhoc network is my firewall, which too is configured with ifupdown, but watch out, i had been trying to configure the interface on the firwall for weeks, with no luck, then one day I had to reboot the machine, and by magick adhoc worked on it, so my guess is that when you configure/deconfigure it leaves some garbage in the interface config or something like that. [18:24] in my opinion libmozjs.so is really important for 3rd applications who wants to interpret and run javascript. [18:24] jorendorff: i would be happy to get more ideas from him ;) [18:25] excellent -- I will do that [18:25] niki: yeah. thats what i ment. for ages adhoc didnt work on modern intel [18:25] keep in mind that libmozjs as a library for applications that want a scripting engine [18:25] asac: I'll try to empty the interfaces file, after filing the nm crash [18:25] if it works now fine. [18:25] niki: but then NM should work ttoo [18:25] niki: give it a try [18:25] is an extremely different thing from libmozjs for things that want to hook into mozilla [18:25] niki: thanks [18:25] asac: both ends is intel fyi. [18:25] Mozilla bug 97954 in JavaScript Engine "autoconf build environment for spidermonkey" [Enhancement,Assigned] [18:26] niki: after empying you just need to restart ... or sudo killall nm-system-settings [18:26] ok, ahould I also remove the lo interface ? [18:26] jorendorff: right. i am all with you. imo mozjs should become a real project with its own build system and a decent library versioning policy [18:27] jorendorff: and then xulrunner/firefox should be able to use mozjs from system-library [18:27] just as it uses gtk now (e.g. with proper minimal versoin requirements and such) [18:27] the latter will take some selling :) but I'm all with you on the former. [18:27] of course upstream firefox builds could still use their own in-binary-blob libmozjs if they dont want to rely on that [18:28] jorendorff: well. do the former and i am sure we can get the latter supported in the build-system [18:28] thats the easier task ;) [18:28] (e.g. making xulrunner use system-mozjs) [18:33] asac: regarding this bug https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/269474 , I now have one more crash (exactly the same) [18:33] Launchpad bug 269474 in network-manager "NetworkManager crashed with SIGSEGV in strlen()" [Medium,Incomplete] [18:35] asac: do you have a workable solution for this issue at current stage? [18:36] niki: do you have the the syslog for this crash too? [18:36] suzhe: i will think about it ... have to run now ... will be back in a few hours [18:36] sec [18:36] ok have to run ... culater [18:36] asac: thanks a lot. [18:37] suzhe: dont hope too much for now [18:38] asac: http://pastebin.com/d3afb58e9 [18:39] in the first section marked ### NM Restart ### [18:39] that is with everything exceplt lo commented out [18:39] in interfaces file that is [18:46] asac: gonna reboot, with an empty interfaces file, except lo is still defined. [18:46] be back shortly [18:56] hi, are there known probs with todays ff update? [18:56] i.e. ff crashing every time you try to open it? [19:00] asac: success, nm connected to adhoc here now, with empty interfaces file both wlan0 and eth0 is seen in nm applet, although eth0 wasn't specified in the interfaces file. [19:01] but to get nm to connect to my adhoc network I had to run "iwconfig wlan0 mode ad-hoc" in a term, after that it connected on first try [19:01] ^ asc [19:01] ^ asac [19:05] my prob seems to be related to firebug. :-( Will investigate further. [19:26] asac, why are we dropping other-licenses/7zstub in nobinonly? [19:31] fta: because its binonly? [19:31] removed `other-licenses/7zstub/src/Windows/Control/Dialog.cpp' [19:31] removed `other-licenses/7zstub/src/Windows/Control/Dialog.h' [19:31] removed `other-licenses/7zstub/src/Windows/Control/ProgressBar.h' [19:31] niki: hmm [19:31] removed `other-licenses/7zstub/src/Common/ComTry.h' [19:31] removed `other-licenses/7zstub/src/Common/CRC.cpp' [19:31] removed `other-licenses/7zstub/src/Common/CRC.h' [19:31] removed `other-licenses/7zstub/src/Common/Defs.h' [19:32] niki: sounds like a bug in wpasupplicant then [19:32] niki: would be interesting to get a syslog after reproducing [19:32] we drop the whole thing, not just binaries [19:32] asac: reproducing what ? [19:33] asac, removed `./js/src/liveconnect/jsj_nodl.c' [19:35] [reed], what is other-licenses/7zstub/ used for? [19:35] niki: conneting to adhoc without running iwconfig first [19:35] fta: i can tell you that [19:36] fta: its the stub used to pack the iwndows installer [19:36] <[reed]> fta: NSIS Installer [19:36] <[reed]> yeah [19:36] <[reed]> what asac said [19:36] fta: previously there was only a binary ... now there appears to be sources (which is good) [19:36] I didn't run iwconfig first, ran it after trying to use default nm [19:36] fta: if the binary stuff isnt in there then i dont mind if we ship it in the nobinonly sources. [19:36] fta: but we dont need it as we have 7z in the archive anyway [19:37] and linux builds dont use it [19:37] i would love to see that +nobinonly sig disappear for good [19:39] is that a fake ? http://www.engadget.com/2008/10/20/apples-new-macbook-is-mm-mm-goo/ [19:39] fta: i dont see any benefit from making the version better readable [19:39] ;) [19:39] asac: do you want a copy of my log ? [19:39] niki: i want that as a bug ... attached [19:39] sure [19:40] niki: 1. reproduce 2. attach complete log of failed attempt [19:40] niki: 3. workaround 2. attach complete log after the success [19:41] I have that in my current log, tried to associate 2 times before using iwconfig, so I should have what you want ;-) [19:42] I'll finish what i'm doing right now, and then clean the log a little, and the upload it [20:25] asac: you still up buddy? [20:51] asac: Ok, this is weired, I just rebooted to catch some logs and such for you, but this time it associated the wlan connection automatically without me doing anything [22:44] niki: thats expected. once you connect for first time NM will auto connect there when it sees the same AP next time [23:24] asac, fta: willing to do a quick and easy upload :)? bug 287253 [23:24] Launchpad bug 287253 in diggler "diggler 0.9-16ubuntu1 depends on firefox-3 instead on firefox" [Undecided,New] https://launchpad.net/bugs/287253 [23:24] thanks [23:27] noticed that I haven't set deb mail... stupid [23:28] everything good now [23:28] is it stil open? [23:28] still [23:41] fta: yes [23:42] fta: universe still has normal FFe rules [23:46] asac, bug fix doesn't need permission, right? [23:48] fta: clearly mark it as such in the changelog and keep changes minimal [23:49] i think you have to ping RMs to get it approved, but they wont do a thorough review [23:49] there was an announcement mail about this on -devel-announce recently i think [23:50] reading https://wiki.ubuntu.com/FreezeExceptionProcess is not clear for bug fix only [23:51] fta: https://lists.ubuntu.com/archives/ubuntu-devel-announce/2008-October/000502.html [23:51] thtas the announcement of how universe/multiverse are handled during main hard freeze [23:52] fta: if in doubt bug sispoty [23:52] ok, so 23rd will be the date, i'm still fine without approval [23:52] Jazzva, i'll push it [23:53] fta: i think just uploading and seeing if archive admins auto approve it within 15 hours or so ...otherwise pinging them [23:53] archive adminst -> Release Managers [23:54] fta, during hard freeze they do [23:55] BUt you upload normally, and it enters the unapproved queue, someone from motu-release has to kick it into the archive [23:55] fta, thanks :) [23:55] asac, I'm done with extensions list... [23:55] There wasn't much to update... [23:55] you can see in the changelog [23:55] https://code.edge.launchpad.net/~jazzva/app-install-data-ubuntu/ubuntu.mozilla-extensions [23:55] Jazzva, any reason to not add abrowser ? [23:55] abrowser? [23:55] in MimeType? [23:55] in depends i guess [23:56] for diggler? [23:56] diggler [23:56] or recomends (but then both) [23:56] -Depends: firefox-3 | seamonkey-browser [23:56] +Depends: firefox | seamonkey-browser [23:56] I forgot to do that... I'll upload a new diff [23:56] no need, i can add it [23:58] i think its depated that extensions must depend on the browser [23:58] most likely a recommends is better now tha tthose get auto installed by default [23:59] fta, ok. thanks [23:59] but i guess we have those depends everwhere and shouldnt start to do something righ tnow