[00:00] you run a tracking on chromium? [00:00] what's that realtime search thingy? [00:00] * BUGabundo google real time search engine [00:00] http://venturebeat.com/2009/06/20/who-rules-real-time-search-a-look-at-9-contenders/ [00:00] grreee [00:00] bad google [00:01] 2nd hit http://collecta.com/ [00:06] * BUGabundo echo sleep > /etc/mode && /home/BUGabundo reload [00:07] BUGabundo, you should really learn some shell === asac_ is now known as asac [00:07] its a joke [00:07] come on!!! [00:08] its not even meant to work [00:08] just to say gudd night [00:08] grrr [00:08] some day I'll have to install a spellchecker on this brain of mine [00:36] asac: are you still here? [00:46] micahg: for a minute or so ... yes [00:46] ok, first, I rejected a bug to pacakge FF3.6a1 [00:46] micahg: wishlist bug? [00:46] second, I just saw that mozilla wants to release ff3.6 in quick order [00:47] Well, is there any chance we'd consider packaging individual pre-releases [00:47] I would think that time would be better spent elsewhere [00:47] not sure what you mean by that [00:47] as we have the dailies [00:48] packaging 3.6a1 in an official repo [00:48] ? [00:48] not a1 [00:48] micahg: so for official archive we usually start on late alphas [00:48] like 5 or 6 in the past [00:48] or even early betas [00:48] we want to create a milestone ppa though [00:48] that would fit in there imo [00:48] milestone/backports ppa [00:48] should I reopen the request? [00:49] micahg: a general wishlist bug "3.6aX in the archive can be kept open [00:49] at best give us the bug id so we can alread add it to changelog [00:49] ok, I'll just open as packaging ff3.6 [00:49] yeah [00:50] second [00:50] micahg: and xul 1.9.2 ;) [00:50] separate bug? [00:50] or same? [00:52] bug 410959 [00:52] Launchpad bug 410959 in ubuntu "[needs-packaging] Firefox 3.6" [Wishlist,Triaged] https://launchpad.net/bugs/410959 [00:53] should xul192 be a separate bug? [00:54] not sure. i think most will look for ffox bug so it should be fine [00:54] ok [00:54] so just add xul192 to it? [00:54] or jsut leave it [00:54] micahg: at best link our .head branch to it so noone wastes time and does a package ;) [00:54] micahg: add both branches [00:56] This one? https://code.edge.launchpad.net/~mozillateam/firefox/firefox-3.2.head [00:56] yes [00:56] we should rename it at some point [00:58] Is this the head for xulrunner? lp:~mozillateam/xulrunner/xulrunner-1.9.2.head [00:58] http://benjamin.smedbergs.us/blog/2009-08-07/pyxpcom-welcome-todd-whitema/ [00:58] micahg: yes [01:00] ok [01:00] last issue for the night [01:00] mozilla was talking about releasing 3.6 quickly [01:00] from what i know they try to aim for 9 month cycles since 3.0 [01:00] 3.5 took 12 month though [01:00] would it follow that 3.5 support will only be 6-9 months after 3.6 release? [01:01] i guess they will not extend their support timeframe unfortunately [01:01] so yes. [01:01] so, maybe we should include ff 3.6 in universe in karmic [01:01] that way it'll have a browser for its entire life [01:02] problem is that we cant really make it the default so this wouldnt help much [01:02] asac, we should do something to make the bot abort if the tarball is not for the right branch.. like foo-1.2 version 1.3a1pre [01:02] fta: yeah [01:03] otherwise we are doomed when they bump version i guess [01:03] well, for hardy was ff3 default or ff2? [01:03] we have now moz-version --compare in mozilla-devscripts. we could use that to test/declare upper/lower version bounds [01:03] either manualyl maintained or auto generated from changelog upstream version [01:03] ... [01:06] micahg: for hardy ff3 [01:07] ah [01:07] maybe abort in MOZCLIENT_POSTCOCMD and propagate the failure.. not sure the rest will follow though.. [01:07] well, what happens to jaunty after january? [01:07] default browser loses support [01:07] micahg: backports [01:08] micahg: what happens to hardy-jaunty after january [01:08] ok, but at least ff3.5 is in the repos and will have support at least though 2010 [01:08] that's why I'm wondering about ff3.6 [01:08] not in hardy [01:08] yeah [01:08] the universe package are not a valid support option [01:09] the problem is xulrunner ... which has rdepends [01:09] it would be quite an effort to backport all rdepends to new xulrunner [01:09] but might be the best way indeed ;) [01:09] maybe we'll have chromium by then ;) [01:09] not so sure yet if i want to do that [01:10] !info firefox dapper [01:10] fta: it doesnt matter. rdepends is the problem [01:10] firefox (source: firefox): lightweight web browser based on Mozilla. In component main, is optional. Version 1.5.dfsg+1.5.0.15~prepatch080614l-0ubuntu1 (dapper), package size 7810 kB, installed size 23108 kB [01:10] we still have that, so why not 3.0 [01:10] did i say something different? [01:10] fta: dapper isn't receving desktop support [01:10] i will backport patches until i die ;) [01:11] i really wonder how many users are still using that... [01:11] fta: its EOL now. [01:11] (dapper) [01:11] asac: can we add a note about FF EOL like debian does? [01:11] only server is still supported [01:11] the desktop [01:11] micahg: no [01:11] micahg: debian surrenders ... which is not something we can do [01:11] lol [01:12] actually we supported in debian for full stable [01:12] just not for oldstable [01:12] because we have no folks that are willing to maintain oldstable long enough [01:12] so you intend to backport fixes to 3.0.x until karmic is EOL? [01:12] and i tried as best as i could, but my time is also finite [01:12] I know [01:13] micahg: until hardy is EOL [01:13] not sure what comes first [01:13] same time :) [01:13] 4/11 [01:13] 3.0 will ride along with that [01:13] so yeah [01:13] good luck for karmic then [01:13] i think its 12 month backports [01:13] should be doable [01:13] 3.5 branch is not that much different [01:13] i expect most issues to popup in tracemonkey ... which dont need to be backported in most cases [01:14] as 3.0 has no tracemonkey [01:14] the layout stuff etc. is hopefully similar enough to make backports possible [01:15] ok, so only CVE's would be backported? [01:16] in the past i backported all sg:* bugs fixed upstream [01:16] I'm just wondering if I should be marking stuff Won't Fix for ff3.0 [01:16] not all get a MFSA/CVEs, but most [01:16] of course MFSA/CVE bugs have priority [01:16] what is sg:*? [01:16] micahg: everything except security group bugs or top crashers are wont fix in 3.0 [01:16] (and packaging bugs) [01:17] ok, I'll keep marking them then ;) [01:17] micahg: please ask reports to add firefox 3.5 task on their own if they dont see it [01:17] in that way we might loose some bugs that just sit there for nothing ;) [01:17] i mean for low/undecided and maybe medium bugs [01:18] * micahg usually adds it...is that not right? [01:18] if you touch bugs you can add it [01:18] but dont go through the list and add it everywhere... its a good time to let users confirm it [01:18] ok, I usually add it if it's already upstream [01:18] yes. upstreawmed bugs should be moved too [01:19] I'll do that...ping user for update [01:19] I hope next weekend to do more triage [01:19] I hope this is the last Sunday I have to work for a while [01:29] scons: *** [/build/buildd/chromium-browser-3.0.198.0~svn20090809r22888/build-tree/src/sconsbuild/Release/lib/libskia.a] /build/buildd/chromium-browser-3.0.198.0~svn20090809r22888/build-tree/src/sconsbuild/Release/.sconsign.dblite: No such file or directory [01:29] hmm [01:30] so system scons is not good enough in intrepid [01:31] bugzilla is broken [01:31] "Repairs in Progress" [01:32] duh-RAM-a! [01:44] yeah [01:46] lots of other sites too [01:46] MDC for instance is down [01:46] https://developer.mozilla.org/en/GRE [03:48] firefox 3.5 cannot use the URW Chancery L font [03:49] http://www.codestyle.org/servlets/FontSampler?class=URWChanceryL&font=URW%20Chancery%20L&compare=normal [03:49] it can use that font on mac [03:50] why not? [03:50] looks fine to me [03:50] really? it's a script font. you're seeing that on that page? [03:51] oh maybe not then [03:51] it would need to be installed on teh system [03:51] it's showing you a default font [03:51] it's instaleld in ubuntu by default. you've got it [03:51] it looks different [03:51] try that page in ff 3 [03:51] or any other browser [03:52] It shows me that font family as being nrendered [03:53] as being what? [03:53] ah, I see [03:53] hmmm [03:54] in Firebug [03:54] ok, I'd say open a bug [03:54] no, i can't do that yet [03:55] guhhh [03:55] why not? [03:55] i have to test all of the other fonts too [03:55] how can i trust it? [03:55] if it broke this one why didn't it break any others? [03:55] ok, well, when you are done, please open a bug report in firefox-3.5 and I will check upstream [03:57] the common fonts page shwos discrepencies [03:57] huh? [03:57] http://www.codestyle.org/css/font-family/sampler-UnixResults.shtml [03:58] between ff3 and ff3.5 [03:59] oh crap [03:59] how many, do you suppose? [04:00] looks like quite a few [04:00] guhhhhh [04:00] most are unnoticable unless you are specifically looking for it [04:01] i'm having trouble coming up with this bug report title: "font rendering is all effed up in FF 3.5" [04:01] non standard fonts aren't rendering correctly? [04:02] i dunno [05:50] I attempted to update my firefox-3.5 this morning and it seems like firefox.desktop would be updated after the update is completed, is it intended? [05:51] apt refused to overwrite firefox.desktop saying something like it belongs to firefox-3.0 (???) [05:52] i then uninstalled firefox-3.5 and revert back to firefox-3.0 for now (using hardy 64bit) [05:53] Jeffrey04: is someone helping you? [05:53] no, I don't know where I can report this [05:53] ok [05:53] can you start over [05:53] sorry [05:54] what is the problem? [05:54] I ran apt-get upgrade this morning [05:54] and then updated firefox-3.5 [05:54] oh hardy? [05:54] *on [05:54] but the upgrade failed saying that firefox-3.5-branding cannot overwrite firefox.desktop [05:54] yea, hardy 64bit [05:55] from where did you get a firefox-3.5 package? [05:55] ubuntu-mozilla-daily ppa [05:55] ah [05:55] not sure what causes the problem and i uninstalled firefox-3.5 for now [05:56] just being curious whether the overwritting of firefox.desktop is intended??! [05:56] well, they're getting ready to switch branding for Karmic [05:57] :D [05:57] if you look at the changelog, it seems like they are indeed changing/breaking stuff [05:57] fta was talking about this earlier I think [05:58] iC [05:59] thanks for the information [05:59] * micahg is checking the logs now [06:00] ok, so it seems that it will be fixed in the next few days [06:01] Jeffrey04: still there? [06:01] yes [06:01] ok [06:01] yeah, so they caught it today [06:01] was reading through the changelog [06:02] http://irclogs.ubuntu.com/2009/08/09/%23ubuntu-mozillateam.txt about 4PM [06:02] 4:30PM [06:02] oops [06:02] 5:30PM [06:05] OK :D guess I just wait and reinstall after a few days :) [06:05] i spend most of the time under google-chrome (unless I'm doing javascript) though [06:06] thanks for the help @micahg [06:06] no problem Jeffrey04 [06:07] the team is still working on a more permanent solution for FF3.5 on hardy [06:07] yea, [10:13] fta: so i think we need to add firefox 3 top dailies (and milestone) ppa to properly ship a firefox-3.0.desktop thing [10:22] asac: bahh, mozilla didn't fix the flash-fullscreen bug yes (not even with 3.5.2) :( [10:28] Err http://ubuntu.intergenia.de karmic Release.gpg Something wicked happened resolving 'ubuntu.intergenia.de:http' (-11) [10:28] hmm. /me unable to setup chroot as it seems [10:37] 11:36 < lool> Ah now I find it in https://launchpad.net/ubuntu/+source/thunderbird-3.0 I had checked thundebird3 [10:37] 11:36 < asac> oh. so launchpad creates packages pages for ppa things now?` [10:37] 11:36 < asac> interesting [10:37] 11:36 < asac> one even can file bugs [10:37] 11:37 < asac> i guess that answers the question: "where to file bugs against those packages not yet in ubuntu for real" [10:37] fta: ^^ [10:37] sebner: sorry for that ;) ... but isnt that a flash problem [10:44] fta: https://edge.launchpad.net/ubuntu/+source/chromium-browser [10:44] ;) [10:44] bugs [10:45] btw, there's some (maybe related) issue with bug #406352 [10:45] Launchpad bug 406352 in firefox-3.5 "Needs correctly set Preferred applications>Browser gconf key (fails on manual Firefox 3.5 installs)" [Undecided,New] https://launchpad.net/bugs/406352 [10:49] sladen: not sure how we can address that in the packaging (except when doing the full migration) [10:49] migration==transition [10:51] not sure why you have firefox.desktop if firefox-3.0 is removed (unless you run dailies, where we migrated firefox 3.5 to default now) [10:52] hmm ... so probably -branding package is left over [10:52] after removal [10:55] not sure if i like circular dependencies [10:55] but i guess that would be needed here [10:56] hmm. firefox-3.0-branding depends on firefox-3.0 ... so that shouldnt be a problem [10:56] asac: the fix is for xulrunner so I blame mozilla .. ok, flash. We always can blame flash too \o/ [10:57] ack ;) [11:01] looking through gconftool-2 -R / some things hardcode "firefox" some "firefox %s" and some "firefox.desktop" [11:02] but I guess the only alternative would be to have those hardcoded to /usr/bin/default-distro-browser and have the packages dpkg-divert that [11:05] i think those are bugs from gnome-panel etc. [11:05] some get a full copy of .desktop and some get a reference to .desktop [11:06] in any case. the default transition will make this work [11:06] before that we dont really support uninstalling firefox-3.0 [11:08] asac: right, elephant in the room time. are you just counting the total hits to :hsw1kx-3zxg :jhy8zj-qv39 :getzo5-xcfi [11:08] asac: if so, what's the fourth one? [11:08] sladen: the start.ubuntu.com/9.10 code [11:09] asac: ah, of course [11:09] sladen: btw, i didnt want to merge all multisearch bug into a single huge one ... ;) [11:09] thats why we used the "multisearch" tag [11:09] but doesnt matter ;) ... now all subscribers are on that bug :) [11:10] asac: yeah... I merged them because, AFAICT, the only way to pacify the various technical breakages, is not to use CSE [11:10] sladen: well. but still they are individual issues [11:10] asac: what are Mozilla doing; is their returned page also a CSE, or is it a "special" with Google? [11:10] of course you can always create a "fix all" and merge in there, but that doesnt make it easier [11:11] sladen: they have a special page [11:12] asac: so the situation is either that Canonical either need to also get a special arranged, or need to drop CSE [11:12] thats one opinion. yes. [11:12] there are various compromises though [11:12] what do you have in mind? [11:12] of course having special page would be the perfect fix [11:13] i think the most annoying things are: [11:13] 1. awesomebar does not have feeling lucky [11:13] 2. chrome changes google searchplugin [11:13] 3. newtab not configurable [11:13] is that special page (or rather /lack/ of obviously non-standard page) done on the by the User-Agent, or just via the partner = {firefox,firefox-a} partner ID [11:14] one way would be to drop 1. and 3. and make 2. an "ubuntu search" plugin living next to the real google one [11:15] sladen: the cse page happens because of the action url we are using [11:15] custom search has a different path than the real google search [11:15] the /cse [11:15] yes. [11:15] but its odd. you can add a few of our arguments to the real google search and get something in the middle ;) [11:15] can the real Google search not just be fed with the changed partner-id [11:16] http://www.google.com/coop/cse/ [11:16] sladen: that doesnt work. otherwise noone would probably use custom search ;) [11:17] it works for Mozilla [11:17] yeah. they are lucky ;) [11:17] see, that page is higlighting three bullet points, of which none are actually a desired outcome [11:18] the only one that is really desired is revenue shared (and maybe some numbers) [11:18] sladen: not sure what you mean ... which page are you referring to? [11:21] http://www.google.com/coop/cse/ "Include one or more websites, or specific webpages" (not required) "Host the search box and results on your own website" (not, except for start.ubuntu) "Customize the look and feel of the results to match your site" (negatively desired) [11:21] sladen: we dont host the results in a search box, but direct to google [11:22] also we dont really customize it much ... what you see is basically the default. [11:22] everyone can create a cse account and suggest how to do that better [11:22] okay, in that case, all *three* are *not* desired :) [11:22] so (an unrelated) in-box approach can be found here: http://people.canonical.com/~kirkland/search.html [11:22] Google's front page is the peak of usability---virtually anything done to it has negative impact [11:23] and, Google's results page is the peak of usability---virtually anything done to it has negative impact [11:23] be it (consults list) being in my local language, handling stock quotes, providng a list to gmail, providing a link to images [11:25] asac: kirkland's use of CSE is a valid one; restricting to a subset of webpages [11:26] asac: at the moment, the geeks are peeved, and the grannies are complaining "Google doesn't work on Ubuntu" [11:26] we repeated that we are aware that the current solution is sub-optimal. And one reason of this experiment was to get a detailed list of issues, so we can mitigate them if possible and balance them when with the win that would come out of the various options for the ubuntu project [11:27] *blink*. But can you expand various options? [11:27] so there are a few potential outcome of this: [11:27] 1. we get special page -> best [11:28] 2. we address as many issues as possible in a fashion that we think is the most diligent one (like what i suggested above) [11:29] I'm still slightly lost as to why a special page is (ever) going to be better than the Google default [11:29] 3. we drop all this [11:30] sladen: there are also benefits for the project that are not insignificant. we need to balance them with the loss of usability carefully. [11:30] asac: expand benefits [11:31] revenue and numbers? [11:31] like what rick and me already mentioned in the bug [11:31] from my blog: "will help to fund more talented developers" [11:32] we are not really interested in the numbers [11:32] and those would be hard to defer from it anyway afaik [11:33] yeah, your method of statistics gathering is sub-optimal anyway (no control group, and impacted behaviour) [11:34] has a fortnight of deployment delivered any meaningful results yet? [11:34] sladen: we dont need accurate numbers to get an idea [11:34] is extending the experimental likely to massively altered the results? [11:35] we really just look at the generic numbers to understand the importance etc. [11:35] asac: no, but is something is changed (I'm feeling lucky) then you'll get a notch on the first use; maybe the second, but not by the time somebody has changed their preferences, or switched to another method (eg. doing searches via Pidgin, to quote above) [11:36] and as of a fortnight of deployment what do the "generic numbers" show? [11:37] i dont know about final numbers. we wanted to wrap up next week [11:38] presumebly somebody has frequenctly checked the numbers to see that the experiement is actually working [11:39] no we verified it once after starting it [11:39] that was a day or two afterwards iirc [11:39] not sure if someone monitored the data eagerly ;) ... could be. [11:39] sladen: what would you want to know about those numbers? [11:40] absolute numbers are really meaningless [11:41] asac: not really, I'd like to know if the observed ratios have been useful [11:46] yes ... otherwise we wouldn't have kept it in for the full alpha [11:48] is the intention to keep this going for one alpha, or for two? [11:49] s/is/was/ [11:49] ~1 alpha [11:51] seems we will miss the alpha release though. i am currently working on ffox 3.5 by default and also have a to fix knetworkmanager to work somehow [11:51] so next week is idea of starting to work on this [11:51] and wrapping up etc. [11:52] also a few are on holiday this week, so not really best time to drive things in any direction [11:52] asac: if it's a time thing; I think there are volunteers who will upload a clean version [11:52] as i said. its not decided what exactly to do [11:52] we just know that something will be done ;) [11:53] who needs to make that decision (if it's not you) [11:54] more than one person [11:54] not all are available atm. that blocks it for now [11:54] anyway. have to work now ;) [11:55] from what I see now on CD there won't be multisearch. its shipped as part of 3.0 package and 3.5 will be on CD unless i have no time to work [11:56] okay, so it's only in the 3.0 upload [11:56] I can mark the 3.5 stuff as invalid? [11:57] sladen: i would say yes. folks see it in firefox-3.5 if they have 3.0 installed though [11:57] sladen: at best keep it open [11:57] otherwise folks will rant and bitch about it [11:57] i we will close the bugs when things get done [12:02] how can i see if a package is in "auto-installed" mode? [12:30] asac: apt-mark [12:36] thx [12:37] asac: grep -A1 firefox /var/lib/apt/extended_states [12:37] ah extended states [12:37] great [12:37] thx [13:08] ok that was odd, firefox-3.5 asked if i wanted to import my 3.6 settings [13:09] gnomefreak: 3.6? sure it wasnt 3.0? [13:09] gnomefreak: what did you select? [13:09] decide later? [13:11] asac: decide later. i think it was 3.6 but im still sleeping [13:13] gnomefreak: can you stop and start it again? [13:13] gnomefreak: and read if its 3.0 ;) [13:13] then select decide later :) [13:15] asac: ok 1 sec, can you please explain comments in bug 18995 [13:15] Launchpad bug 18995 in mozilla-thunderbird "[MASTER] "Open With" dialog not user-friendly" [Undecided,Confirmed] https://launchpad.net/bugs/18995 [13:15] checking now [13:16] gnomefreak: open with dialog is not user friendly. thats known [13:16] asac: comments on why we dont patch it i mean [13:16] because we dont have the proper code [13:16] the one suggested istn good enough afaik [13:16] if there is any suggested at all there [13:17] asac: the ones we gave is not good enough to certain people [13:17] comparing the start page to this issue [13:17] not sure what you mean [13:17] i did not get the dialog agian. do i need to restart system? [13:18] gnomefreak: ensure that no process is running [13:18] maybe its a bug [13:19] gnomefreak: try to re-login [13:19] if it doesnt re pop up its a bug [13:19] file one and ping fta about it ;) [13:20] asac: here are 3 of the many comments. this is after telling them we are waiting for upstream to do it due to major code changes http://paste.ubuntu.com/250803/ [13:20] asac: ill be right back going to try to get it to pop-up [13:23] gnomefreak@Development:~$ ps aux | grep firefox-3.5 [13:23] 1000 4370 0.0 0.2 1856 576 pts/1 S+ 08:23 0:00 grep firefox-3.5 [13:23]