| micahg | asac: unless we backport kmozillahelper, there's no reason to include the patches in < lucid | 00:00 |
|---|---|---|
| asac | chrisccoulson: right. but lets keep it apply, just make it so it doesnt change anything if qt version is lower | 00:00 |
| fta | btw, i suck in css, if someone is willing to improve the look, please ping me | 00:00 |
| micahg | chrisccoulson: I hope you mean jaunty ;) | 00:00 |
| asac | micahg: true, but having just one branch would be better | 00:00 |
| asac | my 2c | 00:00 |
| chrisccoulson | micahg - yeah ;) | 00:00 |
| micahg | asac: right, same branch, but it would only apply the patches for >= lucid | 00:00 |
| asac | if quilt has that feature tats fine | 00:00 |
| micahg | asac: like we do now with apparmor | 00:00 |
| asac | but if it would mean a manual patch in the rules | 00:01 |
| asac | that would be bad | 00:01 |
| asac | because then you cannot just update patches etc. | 00:01 |
| BUGabundo | fta: http://android.modaco.com/content/google-nexus-one-nexusone-modaco-com/309717/its-dead-new-screen-needed/ | 00:05 |
| BUGabundo | wow | 00:05 |
| BUGabundo | that page crashes chromium | 00:05 |
| asac | chrisccoulson: i think in nsKDEUtils.h we could make a #if QT_VERSION or something that always returns false for qt < lucid | 00:05 |
| asac | static bool getKdeSession() | 00:05 |
| asac | that is | 00:05 |
| BUGabundo | time to debug test it | 00:05 |
| BUGabundo | $ chromiumdatadir="$(mktemp -d)";chromiumdiskcache="$(mktemp -d)";chromium-browser -g --user-data-dir=$chromiumdatadir --disk-cache-dir=$chromiumdiskcache' http://android.modaco.com/content/google-nexus-one-nexusone-modaco-com/309717/its-dead-new-screen-needed/ | 00:06 |
| chrisccoulson | asac - yeah, could do. i will try and look at that tomorrow | 00:06 |
| asac | cool | 00:06 |
| asac | fta: antoher idea would be to add a bot to this channel where i can just say !buildstatus | 00:07 |
| asac | and it dumps the current status somehow ;) | 00:07 |
| asac | but dashboard is good enough i guess ;) | 00:07 |
| asac | we could teach the bot to return the url on some command | 00:07 |
| fta | asac, i already wrote it, but i wasn't sure it would be welcome, so i never plugged it here | 00:08 |
| asac | fta: in this channel i am sure its welcome ;) | 00:08 |
| fta | i will have to dig into my code repository see in what state i left it | 00:09 |
| fta | BUGabundo, that page loads fine here | 00:09 |
| asac | http://identi.ca/notice/33824116 | 00:09 |
| asac | ;) | 00:09 |
| BUGabundo | fta: yeah on a clean profile does too | 00:10 |
| BUGabundo | guess its one of my addons | 00:10 |
| asac | addons crashing chromium? how many do you have? 50? | 00:10 |
| BUGabundo | oh come on asac | 00:10 |
| asac | ;) | 00:10 |
| chrisccoulson | ah, bugger. i was hoping to do the extension updates in hardy with the minimum of updates, but we need to do the binary package name transition (to xul-ext-*) in hardy, else upgrades break :( | 00:24 |
| chrisccoulson | ^^micahg | 00:24 |
| micahg | chrisccoulson: why? | 00:24 |
| micahg | chrisccoulson: all lucid packages should have xul-ext migration paths | 00:25 |
| chrisccoulson | micahg - the packages in lucid have conflicts/replaces to ensure the upgrade works, but they are all versioned | 00:25 |
| chrisccoulson | and the version numbers are too low | 00:25 |
| micahg | chrisccoulson: 1.1+really2.0 :) | 00:25 |
| chrisccoulson | so, we'd need to either SRU all the extensions in lucid to change the versioned replaces, or do the transition in hardy | 00:25 |
| chrisccoulson | heh, we could do that as well | 00:26 |
| chrisccoulson | if it wasn't for gcc, i'd probably have taken over the whole build farm by now ;) | 00:31 |
| asac | i would go for the transition in hardy | 00:39 |
| asac | hmm. actually dont do that for ubufox ;) | 00:40 |
| crimsun | fta: what is that pastebin for? | 00:41 |
| chrisccoulson | asac - would that cause an issue wit ubufox? | 00:45 |
| chrisccoulson | s/wit/with/ | 00:45 |
| asac | chrisccoulson: plugins need to move their location too | 00:46 |
| asac | e.g. they ship in /usr/share/ubufox/plugins | 00:46 |
| asac | so the alternative switcher works | 00:46 |
| asac | not sure if any plugin dose that in hardy though (cant remember when that feature was added) | 00:46 |
| asac | anyway, i think plugins also need updates | 00:46 |
| chrisccoulson | asac - ah, i wasn't suggesting to do the whole transition we did in lucid, but just the binary package name change only | 00:46 |
| asac | ah ok | 00:47 |
| asac | that works i think ... though we need to check with mvo | 00:47 |
| chrisccoulson | as it stands at the moment, the conflicts/replaces in lucid aren't tight enough to handle the name transition during the hardy->lucid upgrade | 00:47 |
| asac | if stable updates in update-manager pull new packages | 00:47 |
| asac | chrisccoulson: yeah. i understand the problem. the version hack suggested by micahg also works. | 00:47 |
| chrisccoulson | asac - they do, i tried that in last week when i was testing the ff30.->3.6 transition | 00:48 |
| asac | ok cool | 00:48 |
| chrisccoulson | (new packages get installed by u-m) | 00:48 |
| asac | so we hav three ptions: a) do full transition, b) do name transtion only, c) do fake version update | 00:48 |
| chrisccoulson | i think i would prefer the 3rd option, but i will probably have to do the second option anyway for the extensions that are already in the PPA | 00:49 |
| asac | yeah maybe. however, option a) would probably the least work intesive one | 00:49 |
| BUGabundo | nite | 00:49 |
| BUGabundo | see you all tomorrow | 00:49 |
| asac | e.g. just throw them from lucid to hardy | 00:49 |
| asac | BUGabundo: night | 00:49 |
| chrisccoulson | yeah, we could do | 00:50 |
| === almaisan-away is now known as al-maisan | ||
| === bdrung__ is now known as bdrung | ||
| fta | question for the python experts, i have some dates like this one u'2010-05-28T10:51:47.486204+00:00', i want to compute durations/deltas | 13:04 |
| fta | nxvl, ^^ ? | 13:06 |
| BUGabundo_remote | seems no one is expert enough | 13:13 |
| fta | yeah | 13:14 |
| fta | i need a datetime or something like that | 13:14 |
| === al-maisan is now known as almaisan-away | ||
| fta | tricked it.. datetime.datetime.strptime(date_first_dispatched, "%Y-%m-%dT%H:%M:%S.%f+00:00") | 13:39 |
| chrisccoulson | bugger, we've ran out of space in the u-m-s PPA | 14:56 |
| chrisccoulson | ^^asac - do you think we should delete stuff or ask for more space? | 14:56 |
| fta | probably both | 14:58 |
| asac | chrisccoulson: ums? security ppa? | 15:20 |
| asac | chrisccoulson: whats the current max space? | 15:21 |
| asac | 4G? | 15:21 |
| asac | lets get 10 then | 15:21 |
| chrisccoulson | asac - it's 10G already | 15:21 |
| asac | hmm. | 15:21 |
| asac | chrisccoulson: go into #is and ask if there is a way to increase size at all | 15:21 |
| asac | ask bigjouls | 15:21 |
| chrisccoulson | ah, bigjouls doesn't seem to be in there | 15:22 |
| asac | chrisccoulson: wait till he is back ;) | 15:23 |
| asac | chrisccoulson: so sorry ... #soyuz ;) | 15:23 |
| chrisccoulson | on freenode or canonical? | 15:23 |
| asac | chrisccoulson: latter | 15:23 |
| chrisccoulson | heh, he's not in there either ;) | 15:23 |
| asac | 11.5 GiB (1.15%) of 1000.0 GiB | 15:34 |
| asac | nice | 15:34 |
| asac | lets share warez there ;) | 15:34 |
| BUGabundo_remote | new disk? | 15:34 |
| asac | no ... security ppa | 15:34 |
| asac | https://edge.launchpad.net/~ubuntu-mozilla-security/+archive/ppa/+packages | 15:34 |
| BUGabundo_remote | 1T | 15:35 |
| BUGabundo_remote | nice | 15:35 |
| asac | yep | 15:35 |
| chrisccoulson | asac - that should last me until tomorrow :P | 15:35 |
| asac | at least ;) | 15:35 |
| * chrisccoulson sees how quickly he can fill it up | 15:35 | |
| nxvl | fta: time object | 15:52 |
| nxvl | fta: datetime actually | 15:52 |
| nxvl | fta: import datetime, then you put that in a datetime object and you can get days for example | 15:52 |
| === almaisan-away is now known as al-maisan | ||
| chrisccoulson | come back micahg :) | 17:42 |
| nxvl | chrisccoulson: did you got the change to prepare the document of the firefox backport? | 17:48 |
| chrisccoulson | nxvl - not yet, i really need to get these packages all in to the PPA first | 17:49 |
| nxvl | sure | 17:49 |
| nxvl | let me know when you have it please | 17:49 |
| chrisccoulson | nxvl, rick created this page earlier: https://wiki.ubuntu.com/DesktopTeam/Mozilla/FirefoxHardyJaunty | 17:49 |
| chrisccoulson | and it contains the contents of an e-mail i sent out, which has a few details in | 17:50 |
| nxvl | and it refers to some lists that are not there | 17:53 |
| chrisccoulson | nxvl - ah, rick forgot the links i put at the bottom of the e-mail | 17:58 |
| chrisccoulson | what's your e-mail address? | 17:58 |
| nxvl | that's what i assumed | 17:58 |
| nxvl | nicolas.valcarcel AT canonical DOT com | 17:58 |
| chrisccoulson | nxvl - ok, i forwarded the mail to you too | 17:59 |
| nxvl | thanks | 17:59 |
| fta | chrisccoulson, http://people.ubuntu.com/~fta/ppa-dashboard/ums.html | 18:38 |
| asac | fta: can you make http://people.ubuntu.com/~fta/ppa-dashboard/ list all the dashboards rather than showing the default one? | 19:01 |
| asac | and maybe move the current index.html to daily.html | 19:01 |
| asac | or somehting? | 19:01 |
| asac | just and idea so one can browse all the dashboards | 19:02 |
| fta | asac, most probably, but i'm still experimenting | 19:02 |
| asac | without making an overview page | 19:02 |
| asac | kk | 19:02 |
| fta | the API is weird | 19:02 |
| fta | and slow | 19:02 |
| fta | and buggy | 19:02 |
| asac | oh you are using API ... and dont parse html ;) | 19:02 |
| asac | good | 19:02 |
| asac | file bugs ;) | 19:03 |
| asac | even if only 5% get fixed its a win | 19:03 |
| fta | the API is not on par with the website :( | 19:22 |
| jdstrand | chrisccoulson: hi! so, I have that usn 935-1 was given to you for ff 3.5.10. is karmic going to get 364 or 3.5.10? | 19:31 |
| micahg | jdstrand: 3.5.10 ATM | 19:31 |
| jdstrand | chrisccoulson: ^ | 19:31 |
| jdstrand | micahg: thanks | 19:31 |
| chrisccoulson | micahg - ah, you missed the discussion in #ubuntu-desktop ;) | 19:32 |
| micahg | jdstrand: yes | 19:32 |
| micahg | oops | 19:32 |
| micahg | chrisccoulson: yes | 19:32 |
| micahg | was trying to get in there | 19:32 |
| * micahg goes to read the backlog | 19:32 | |
| chrisccoulson | we need to get 3.6.4 in karmic before jaunty ;) | 19:32 |
| chrisccoulson | that's basically the summary | 19:32 |
| micahg | chrisccoulson: why? | 19:32 |
| chrisccoulson | we still need to preserve the upgrade path through karmic | 19:32 |
| micahg | chrisccoulson: I know there will be upgrade issues, but that should only affect the next update | 19:32 |
| micahg | chrisccoulson: we should have a month to do that | 19:33 |
| chrisccoulson | yeah, but we can't have a month where upgrades break ;) | 19:33 |
| chrisccoulson | users wouldn't be very happy | 19:33 |
| micahg | chrisccoulson: well, ugh | 19:33 |
| micahg | chrisccoulson: so let's use the karmic xul rdepends for jaunty and then upgrades won't break | 19:34 |
| jdstrand | chrisccoulson: if karmic gets 3.6.4, please use 930-1 for it like the others | 19:34 |
| * jdstrand holds onto 935-1 for safekeeping for now | 19:34 | |
| micahg | chrisccoulson: and version all the extensions for jaunty so the upgrades won't break | 19:34 |
| micahg | chrisccoulson: ugh, I guess if we're doing all that work we can do karmic also :( | 19:35 |
| micahg | chrisccoulson: ok, so what's the timeline then, hardy, then karmic the week after, then jaunty? | 19:35 |
| chrisccoulson | micahg - yeah, that's the plan now | 19:35 |
| micahg | chrisccoulson: I think we should push 3.5.10 to karmic on tuesday regardless | 19:36 |
| chrisccoulson | micahg - yeah, that might be what happens. it all depends on how fast we can have karmic ready ;) | 19:36 |
| micahg | chrisccoulson: well, my goal is to have whatever is left for hardy done by Tuesday | 19:37 |
| micahg | chrisccoulson: BTW, should I reversion my uploads in the transition PPA so you can copy/rebuild to security PPA? | 19:38 |
| chrisccoulson | micahg - yeah, can do | 19:38 |
| micahg | chrisccoulson: wait, as in I should ? | 19:38 |
| chrisccoulson | yeah, you can do that | 19:39 |
| micahg | chrisccoulson: k | 19:39 |
| micahg | chrisccoulson: will you be signed in over the weekend? Those should be ready Sun morning | 19:40 |
| chrisccoulson | i will be at some point | 19:40 |
| micahg | chrisccoulson: k | 19:40 |
| micahg | chrisccoulson: so which extension plan are we using? I saw the transcript between you and asac | 19:40 |
| chrisccoulson | well, i had to fix the ones i uploaded already to rename the binaries to xul-ext-* and provide the transitional package | 19:42 |
| jdstrand | chrisccoulson, micahg: in thinking about 3.5.10 vs 3.6.4 for karmic for a second, I have an opinion | 19:42 |
| micahg | chrisccoulson: so we're doing the transition in hardy? | 19:42 |
| micahg | chrisccoulson: won't that cause more problems with additional new binaries? | 19:43 |
| jdstrand | unless 3.5.9 has a critical vulnerability that doesn't exist in 3.0 or 3.6, we should just publish 3.6.4 for karmic | 19:43 |
| chrisccoulson | micahg - it's up to you. i had to do that for the ones i already uploaded, as i had no choice once i uploaded them | 19:43 |
| jdstrand | reason being, that is a lot of archive churn and disruption for karmic users | 19:43 |
| chrisccoulson | jdstand - yeah, i'm not sure what 3.5.10 fixes | 19:43 |
| micahg | jdstrand: the problem is the time between we can have all the extensions/rdepends ready for karmic | 19:43 |
| jdstrand | more than likely, 3.6.4 fixes whatever would be in 3.5.10 anyway, so focusing on 3.6.4 for all releases is best | 19:44 |
| micahg | jdstrand: I would think we can get up 3.5.10 with minimal effort to at least have a browser w/out known issues | 19:44 |
| micahg | jdstrand: my guess is it'll take another week to have karmic ready | 19:44 |
| jdstrand | micahg: yes, but only to turn around 2-3 days later with a new update | 19:44 |
| jdstrand | ok, 7 days | 19:44 |
| micahg | jdstrand: your call | 19:44 |
| jdstrand | normally, we release new packages for all supported releases at once | 19:45 |
| micahg | jdstrand: if for some reason we don't get karmic to 3.6.4 w/in a week, then what though? | 19:45 |
| jdstrand | I realize this isn't a normal update | 19:45 |
| micahg | chrisccoulson: what do you think? | 19:45 |
| jdstrand | I don't think I have enough info to make the call | 19:46 |
| chrisccoulson | i'm not sure, but 3.5.10 is ready to go anyway | 19:46 |
| micahg | chrisccoulson: WRT extensions, I'll do it whatever way you want | 19:46 |
| jdstrand | I had assumed all releases were getting 3.6.4, until a few minutes ago | 19:46 |
| chrisccoulson | jdstrand - i'm more concerned about what we do with 3.6.4 on tuesday when hardy, karmic and jaunty aren't all ready ;) | 19:46 |
| jdstrand | of course | 19:47 |
| jdstrand | but, it wouldn't be the first time that mozilla updates didn't go out on the day upstream released them | 19:47 |
| micahg | jdstrand: if we push 3.5.10 to jaunty/karmic and hardy up to 3.6.4, we'll at least have up to date browsers on all releases | 19:47 |
| jdstrand | so what I am hearing is that hardy and jaunty are fairly close to making it midweek next week, but karmic is not? | 19:48 |
| chrisccoulson | jdstrand - that was the plan, but then we realized we still need to maintain the jaunty -> karmic upgrade path | 19:48 |
| chrisccoulson | so we need to do karmic before jaunty (or at the same time) | 19:48 |
| jdstrand | well, most jaunty users are using 3.0 I'm sure, since that is in the default install | 19:49 |
| chrisccoulson | yeah, that's right | 19:49 |
| jdstrand | so upgrading it to 3.5.10 doesn't do a lot for them | 19:49 |
| jdstrand | so, hardy is ready for midweek next week, but 364 for jaunty and karmic is another week out? | 19:49 |
| chrisccoulson | jdstrand - i think that's how it's looking atm | 19:50 |
| jdstrand | while not optimal, since this is a very special case, I'm ok with hardy 364 going out before jaunty and karmic | 19:51 |
| jdstrand | (hardy+lucid) | 19:51 |
| micahg | chrisccoulson: jaunty is also more complicated as we should upgrade/import profiles from both 3.0 and 3.5 | 19:51 |
| micahg | chrisccoulson: or at least ask which one | 19:51 |
| jdstrand | so, let's just say definitively that hardy and lucid will get the 3.6.4 update next week, and they should use 930-1 | 19:51 |
| jdstrand | that leaves jaunty and karmic | 19:52 |
| jdstrand | karmic-lucid upgrade is ok | 19:52 |
| jdstrand | if we don't update jaunty, jaunty to karmic upgrade is ok | 19:52 |
| chrisccoulson | micahg - yeah, i've not tested that yet | 19:52 |
| chrisccoulson | jdstrand - yeah, seems sensible | 19:53 |
| jdstrand | so, hardy and lucid go together, and jautny and karmic go together a week later with 930-2 | 19:53 |
| micahg | jdstrand: also if we run into regressions in hardy, I would assume the priority would be those regressions, so karmic/jaunty would get pushed back | 19:53 |
| jdstrand | micahg: yes | 19:53 |
| micahg | jdstrand: that's why I think we should push 3.5.10, just in case | 19:53 |
| jdstrand | with this in mind, the only real course is that jaunty 3.5 and karmic 3.5 should get 3.5.10 | 19:53 |
| jdstrand | (after all) | 19:53 |
| micahg | jdstrand: :) | 19:53 |
| jdstrand | that leaves default jaunty 3.0 users out for a bit though, but that is about the best we can do atm | 19:54 |
| jdstrand | again, I didn't realize only hardy was ready for 3.6.4 | 19:54 |
| micahg | jdstrand: can we make an annoucement that they should switch to firefox-3.5 for a week or is that too confusing? | 19:54 |
| micahg | jdstrand: I didn't work fast enough on this project, that's why it isn't ready | 19:55 |
| jdstrand | micahg: I think that would be too confusing, and quite unprecedented | 19:55 |
| micahg | jdstrand: k | 19:55 |
| jdstrand | micahg: oh, no blame at all -- I just didn't know the status and assumed we were talking about all three | 19:55 |
| jdstrand | it will be clear enough in the usn that jaunty didn't get an update, and our cve tracker will also reflect it. then a week or so later, we will get it out | 19:56 |
| chrisccoulson | micahg - ok, so i'm just going with doing the binary name transition in hardy for the extensions, it's not really much work to do that | 19:56 |
| jdstrand | actually... | 19:56 |
| chrisccoulson | and i'm not updating extensions that currently don't work either | 19:57 |
| micahg | chrisccoulson: k, will do WRT binary name change, what do you mean by not working? | 19:57 |
| chrisccoulson | we should only do what is necessar to make sure that things which currently work carry on working | 19:57 |
| jdstrand | no, scratch that | 19:57 |
| chrisccoulson | micahg - some extensions have unresolvable dependencies, or depend only on firefox 2 etc... | 19:57 |
| chrisccoulson | we shouldn't worry about those ones | 19:58 |
| micahg | chrisccoulson: k | 20:00 |
| jdstrand | nxvl: you probably want to read backscroll ^. basically next week hardy and lucid are getting 3.6.4, jaunty and karmic 3.5 will get 3.5.10 with 3.6.4 coming a week or so later. this means jaunty 3.0 will not be patched for roughly a week and half | 20:00 |
| nxvl | jdstrand: thanks for letting me know, will check the backscroll | 20:01 |
| micahg | chrisccoulson: anything else we need to chat about before the wekeend? | 20:01 |
| chrisccoulson | micahg - i don't think so. i'll be around for most of the evening anyway (although i'm going for dinner in a few minutes) | 20:02 |
| micahg | chrisccoulson: k, I'll be around for a few more hours, but will be running around the office | 20:02 |
| === al-maisan is now known as almaisan-away | ||
| fta | http://blog.chromium.org/2010/05/desktop-notifications-now-available-to.html | 20:30 |
| micahg | fta: idk if these work with notify-osd | 20:41 |
| fta | i'd like something to interact between chromium and the app indicator | 20:42 |
| fta | (i use chromium in --app mode a lot) | 20:42 |
| fta | http://arstechnica.com/gadgets/news/2010/05/intel-to-hardware-accelerate-webm-if-it-becomes-popular.ars | 20:47 |
| fta | http://arstechnica.com/security/news/2010/05/adobe-considers-monthly-patches-to-improve-security.ars | 20:49 |
| jdstrand | micahg: are all the backported items for hardy in the ubuntu-mozilla-security ppa now? | 20:51 |
| micahg | jdstrand: not yet, we still have another 30 packages I think | 20:51 |
| jdstrand | micahg: ok. for my part, I won't be able to start testing til tuesday then | 20:52 |
| jdstrand | micahg: what is the url describing everything that is being backported? | 20:52 |
| micahg | jdstrand: https://wiki.ubuntu.com/DesktopTeam/Specs/Lucid/FirefoxNewSupportModel/xulrunner-list has teh packages and there is a gobby doc for the extensions that we will probably move to the wiki | 20:53 |
| micahg | jdstrand: the table at the bottom is what we're backporting | 20:53 |
| jdstrand | micahg: yes, that's it. thanks! | 20:54 |
| micahg | jdstrand: chrisccoulson wrote up a plan as well | 20:54 |
| micahg | jdstrand: https://wiki.ubuntu.com/DesktopTeam/Mozilla/FirefoxHardyJaunty | 20:54 |
| micahg | jdstrand: if the porting is done (I hope by Tuesday) if you want, I can help with some testing in the mornings (not sure if that's more important than the packages for jaunty/karmic) | 20:55 |
| jdstrand | micahg: when do you expect the call for testing email to go out? | 20:55 |
| jdstrand | micahg: testing is everything | 20:56 |
| micahg | jdstrand: I'm hoping tuesday, I'll be working Sun/Mon to finish any packaging | 20:56 |
| jdstrand | micahg: we can potentially identify problems that way, and incorporate fixes in the jaunty/karmic stuff | 20:56 |
| micahg | jdstrand: k, well, when do you expect to be on Tues morning? | 20:57 |
| jdstrand | micahg: hmmm.. if the call for testing goes our tuesday, then when is it hoped these will hit -security? | 20:57 |
| micahg | jdstrand: I hope I can find people to upload/copy&rebuild what I need to the security PPA over the weekend and hopefully < 10 packages left to build tues morning | 20:58 |
| jdstrand | micahg: I have monday off. Tuesday I will start at 13:00 UTC | 20:58 |
| micahg | jdstrand: I will make a point to be on then and hopefully will have had a chance to catch up w/ chrisccoulson by then | 20:59 |
| jdstrand | micahg: my plan for next week is testing of this update. I think we need all hands on deck for testing though, and not just a few of us | 20:59 |
| micahg | jdstrand: my only concern is I can't do testing and karmic/jaunty porting | 21:00 |
| jdstrand | so I'm hoping the call for testing will elicit lots of feedback (positive I hope!) | 21:00 |
| micahg | jdstrand: I can send a message to the bugsquad as well to help w/testing | 21:01 |
| jdstrand | micahg: I realize that, but the testing will ultimately help your confidence with the porting | 21:01 |
| jdstrand | this is a huge update and while I'll test it, I don't want to be the only one doing that | 21:01 |
| micahg | jdstrand: k, that's fine, I've just never done anything like this before and didn't know where the priorities are | 21:02 |
| jdstrand | I'm not suggesting I will be, I just think this requires a big effort | 21:02 |
| jdstrand | micahg: if that requires talking to your manager/etc, I understand | 21:02 |
| jdstrand | this is a big update on a stable LTS release | 21:02 |
| jdstrand | we strive for and agressively test for no regressions | 21:03 |
| jdstrand | we won't be able to test everything, but the more people we have, the better chance we have at finding things/demonstrating things are ok for most | 21:03 |
| micahg | jdstrand: it will be chrisccoulson's call then, I'll discuss with him on Tuesday | 21:03 |
| jdstrand | micahg: cool, thanks | 21:03 |
| micahg | jdstrand: you have a minute for an OT question? | 21:05 |
| jdstrand | micahg: in terms of support, we have a responsibility to the jaunty 3.0 users, of course. however, hardy LTS users are likely more numerous and definitely less forgiving with regressions | 21:05 |
| micahg | jdstrand: k, makes sense | 21:05 |
| jdstrand | micahg: so the biggest impact we can have is to test the crap out of hardy. that will build confidence in your work on jaunty, so it isn't wasted time either way | 21:05 |
| jdstrand | micahg: ok, shoot | 21:06 |
| micahg | jdstrand: k | 21:06 |
| micahg | jdstrand: do you know of a good central password repo tool? | 21:06 |
| jdstrand | micahg: can you explain your use case? | 21:06 |
| micahg | jdstrand: root password store for lots of servers | 21:07 |
| chrisccoulson | micahg / jdstrand - i've not read the whole scrollback yet, but if you;re talking about testing, then i'm already discussing with ara about coordinating that | 21:07 |
| jdstrand | chrisccoulson: that is good news | 21:07 |
| jdstrand | chrisccoulson: my only concern is that enough people are testing it | 21:08 |
| micahg | chrisccoulson: I was wondering if my time was better spent continuing porting once hardy is done or to test hardy | 21:08 |
| jdstrand | s/only/biggest/ | 21:08 |
| jdstrand | micahg: re passwords> is this just for you to maintain locally or for all those servers to access? | 21:09 |
| chrisccoulson | micahg - your time is probably better spent porting | 21:09 |
| micahg | jdstrand: well, it's for a team to manage these devices | 21:09 |
| BUGabundo | evening | 21:09 |
| micahg | chrisccoulson: k | 21:09 |
| jdstrand | chrisccoulson: mind you, I haven't been a part of your conversation with ara, but imho the most important thing is testing | 21:10 |
| jdstrand | chrisccoulson: if the resources are there without micahg, then that's fine | 21:11 |
| jdstrand | micahg: otoh I don't have a recommendation, sorry | 21:11 |
| chrisccoulson | jdstrand - yeah, we haven't discussed how exactly to test it yet, but she is aware of this and is going to help out | 21:11 |
| chrisccoulson | tbh, the thing i'm most worried about is getting the upgrade paths right | 21:11 |
| micahg | jdstrand: k, was just wondering if your team had a favorite tool for this use case | 21:11 |
| chrisccoulson | i'm quite confident that everything will be ok functionally | 21:12 |
| micahg | chrisccoulson: I'm not sure about that | 21:12 |
| jdstrand | chrisccoulson: the upgrade paths is part of testing | 21:12 |
| micahg | chrisccoulson: I'm referring to the rdepends | 21:12 |
| jdstrand | chrisccoulson: but this is I think the largest security update ever performed within Ubuntu | 21:12 |
| chrisccoulson | jdstrand - yeah, i need to get it right ;) | 21:13 |
| jdstrand | chrisccoulson: heh | 21:13 |
| jdstrand | of course | 21:13 |
| chrisccoulson | i think i'm going to have to work on monday | 21:13 |
| chrisccoulson | i wonder if i can defer my vacation for a few days ;) | 21:13 |
| jdstrand | at the risk of stating the obvious, this is complex and high profile and most importantly, a very large number of users are going to be affected | 21:13 |
| micahg | chrisccoulson: s/days/weeks :) | 21:14 |
| chrisccoulson | ;) | 21:14 |
| jdstrand | on an LTS, these users are very unforgiving of regressions | 21:14 |
| jdstrand | (ie, enterprise users) | 21:14 |
| chrisccoulson | yeah, that's what worries me ;) | 21:14 |
| jdstrand | we need as many people doing functional testing as possible | 21:15 |
| jdstrand | upgrades next | 21:15 |
| micahg | jdstrand: we can always blame upgrades on Lucid :) | 21:15 |
| jdstrand | if ara can dedicate resources to that, great | 21:15 |
| micahg | and fix for .1 | 21:15 |
| jdstrand | I plan to be testing this too, but 2-3 people really does not seem like enough... | 21:15 |
| chrisccoulson | yeah, hardy -> lucid upgrades aren't enabled yet | 21:16 |
| micahg | chrisccoulson: they're not | 21:16 |
| micahg | ? | 21:16 |
| jdstrand | at least not for the quick turnaround that I perceive is desired for publication | 21:16 |
| micahg | chrisccoulson: what upgrade path is there for hardy at this point? | 21:16 |
| chrisccoulson | micahg - not yet (it's not offered by u-m anyway) | 21:16 |
| chrisccoulson | micahg - no upgrade path is offered in u-m atm | 21:16 |
| chrisccoulson | although | 21:16 |
| chrisccoulson | it is technically still possible to upgrade to intrepid, as the archive isn't on oldreleases yet | 21:17 |
| micahg | chrisccoulson: ah, but they can choose it then? is hardy -> intrepid still offered, or was that turned off in u-m? | 21:17 |
| chrisccoulson | i'm going to ask about moving that next week | 21:17 |
| jdstrand | that should be turned off | 21:17 |
| chrisccoulson | micahg - u-m in hardy offers no upgrades atm | 21:17 |
| jdstrand | if it isn't, that is a serious problem | 21:17 |
| micahg | chrisccoulson: ok | 21:17 |
| jdstrand | chrisccoulson: I asked micahg, but I'll ask you. when is the call for testing email going out (micahg said tuesday hopefully)? when do you hope for this to be published to -security? | 21:18 |
| chrisccoulson | jdstrand - yeah, tuesday seems like a good time (although, it will be possible to test the firefox and extension upgrades before then, hopefully) | 21:20 |
| chrisccoulson | i might not go to bed tonight until all the extensions are done ;) | 21:20 |
| micahg | chrisccoulson: yeah, I'll finish those first, so that should be Monday | 21:20 |
| micahg | chrisccoulson: also, we need to make sure any extensions (not addons) work as well | 21:20 |
| micahg | chrisccoulson: have you thought about openjdk backport yet? | 21:20 |
| chrisccoulson | micahg - yeah, that was in the mail i sent to ara earlier (the one pasted in to the wiki) | 21:21 |
| chrisccoulson | i'm not sure about openjdk | 21:21 |
| micahg | chrisccoulson: I'm going to try to grab the commit that built it for xul192 and add that to the hardy version and see if it works | 21:21 |
| chrisccoulson | cooo, thanks | 21:21 |
| chrisccoulson | s/cooo/cool | 21:22 |
| chrisccoulson | lol | 21:22 |
| jdstrand | please consider sun-java too -- it is very popular on hardy | 21:22 |
| asac | ack | 21:22 |
| micahg | jdstrand: I think that might already work | 21:22 |
| fta | http://arstechnica.com/microsoft/news/2010/05/for-some-companies-ie-6s-ineptitude-is-a-feature-not-flaw.ars | 21:22 |
| chrisccoulson | i will install that when i text the next round of extension updates on hardy | 21:23 |
| asac | chrisccoulson: get mvo involved in testing the upgrades | 21:23 |
| micahg | jdstrand: more than that, on hardy sun-java6 is the default java :) | 21:23 |
| asac | chrisccoulson: he has a good infrastrcuture to test various combinations | 21:23 |
| jdstrand | chrisccoulson: ok. let me be blunt. how do you see testing going? keeping in mind monday is a holiday in the US | 21:23 |
| chrisccoulson | jdstrand, monday is also a holiday in the UK | 21:23 |
| jdstrand | micahg: well, it is still multiverse, but yes | 21:23 |
| chrisccoulson | although, tbh, i think i will be working on monday | 21:23 |
| asac | chrisccoulson: the release 3.6.4 is now scheduled june 7/8 :) | 21:26 |
| asac | so another week ;) | 21:26 |
| asac | build6 is going out | 21:26 |
| micahg | asac: I now understand your original worries about this project :) | 21:26 |
| chrisccoulson | asac - that's quite a relief | 21:26 |
| chrisccoulson | asac - build6 is already in the PPA ;) | 21:26 |
| asac | heh ;) | 21:26 |
| micahg | asac: still June 1 on the releases page | 21:27 |
| jdstrand | chrisccoulson: what is the testing plan? if there is a URL, can you point me to it? | 21:27 |
| chrisccoulson | jdstrand - rick started a page here: https://wiki.ubuntu.com/DesktopTeam/Mozilla/FirefoxHardyJaunty | 21:27 |
| jdstrand | yeah, I saw that. it is quite high level | 21:27 |
| chrisccoulson | but i haven't done any proper planning for testing yet (i'm still trying to get everything updated) | 21:28 |
| asac | chrisccoulson: micahg: jdstrand forwarded you that mail | 21:28 |
| chrisccoulson | asac - thanks | 21:29 |
| micahg | asac: k, thanks | 21:29 |
| chrisccoulson | asac - is that a list that anyone can subscribe to? | 21:29 |
| * jdstrand feels like he isn't getting through... | 21:29 | |
| micahg | chrisccoulson: that's the sec list for Ubuntu | 21:30 |
| micahg | chrisccoulson: oops, mozilla | 21:30 |
| asac | chrisccoulson: no. but we should get you subscribed to announce | 21:30 |
| asac | i think its invite only ... could be that it changed | 21:30 |
| chrisccoulson | asac - yeah, would be useful ;) | 21:30 |
| micahg | asac: not on their lists page | 21:31 |
| asac | yeah its invite only | 21:31 |
| chrisccoulson | jdstrand - so, mvo has offered to help out with upgrade tests | 21:31 |
| chrisccoulson | (in #ubuntu-desktop a few moments ago) | 21:32 |
| micahg | chrisccoulson: I think jdstrand is saying we should have a full test plan ready when the call goes out to maximize testing | 21:32 |
| jdstrand | chrisccoulson: I appreciate your continued work in getting the update ready to test. I also appreciate that you know this is a big update that needs testing and you are willing to do it. in the past, testing has largely fallen on the mozilla maintainer and myself. I am extremely concerned that more resources than 2-3 people have not been allocated for testing this huge update | 21:32 |
| jdstrand | many users could be affected. the reputation of Ubuntu's LTS is at stake | 21:33 |
| jdstrand | and not least of which (to me), my name will be on the USN | 21:33 |
| chrisccoulson | jdstrand - that's why ara is involved too. she's going to get people involved with testing as well | 21:33 |
| jdstrand | chrisccoulson: where is the conversation with ara taking place? | 21:34 |
| jdstrand | chrisccoulson: I may be worrying simply because I haven't been part of the testing conversation... | 21:35 |
| chrisccoulson | jdstrand - via e-mail at the moment, although there's not much in the way of details. that's something i need to start pushing along next week though | 21:35 |
| chrisccoulson | i'll make sure you're involved with that | 21:35 |
| jdstrand | chrisccoulson: ok. I very much appreciate it | 21:35 |
| micahg | chrisccoulson: now that we have another week, what's the plan? | 21:36 |
| jdstrand | chrisccoulson: please CC security@ubuntu.com so our whole team can know what is happening (in the case of regressions, any one of us may need to respond) | 21:36 |
| chrisccoulson | jdstrand - ok, will do | 21:36 |
| micahg | chrisccoulson: also, can you CC me on the test stuff also, please | 21:36 |
| chrisccoulson | micahg - the plan is the same (get hardy ready, then work on karmic and jaunty) | 21:36 |
| chrisccoulson | giving priority to getting extensions updated too | 21:37 |
| jdstrand | chrisccoulson: again, I apologize if I am being a pain :) I want what we all want: a smoothe update | 21:37 |
| micahg | chrisccoulson: k, the thing is that extensions compile against xul, so that's a possible issue as well | 21:37 |
| chrisccoulson | jdstrand - no worries ;) | 21:37 |
| jdstrand | I just happen to know what regressions in stable releases feel like, and it is no fun.... avoiding them is *way* better ;) | 21:38 |
| chrisccoulson | i feel your pain - i expect i probably won't sleep much between now and roll-out ;) | 21:38 |
| jdstrand | chrisccoulson: through testing comes confidence | 21:38 |
| asac | chrisccoulson: jdstrand: send a mail to Dan asking you to be added to security-announce ... if you dont get added by mid next week let me know | 21:38 |
| chrisccoulson | asac - thanks | 21:38 |
| jdstrand | if we can be confident that the vast majority of users will have a smooth upgrade, that is a very good thing | 21:39 |
| chrisccoulson | jdstrand - if you want, then you can already start testing hardy | 21:39 |
| jdstrand | chrisccoulson: well, that is how this conversation started | 21:39 |
| chrisccoulson | (firefox and xulrunner are in the PPA, along with ubufox and some other extensions) | 21:39 |
| chrisccoulson | but not everything is there yet | 21:39 |
| jdstrand | chrisccoulson: I was told some 30 things needed to be uploaded still | 21:39 |
| jdstrand | ok | 21:39 |
| chrisccoulson | jdstrand - yeah, that's right. but the major thing is updated already :) | 21:40 |
| micahg | chrisccoulson: firefox is less likely to break in Hardy than the rdepends/extensions(plugins) | 21:41 |
| jdstrand | that may be, but we need to test it all :) | 21:41 |
| micahg | jdstrand: of course | 21:41 |
| * jdstrand goes to download | 21:41 | |
| chrisccoulson | are powerpc and hppa officially supported on hardy? | 21:45 |
| chrisccoulson | oh, that's not too bad. webkit only fails to build on those arch's because the symbols files need updating | 21:46 |
| micahg | chrisccoulson: according to https://wiki.ubuntu.com/UbuntuDevelopment/PackageArchive#Official%20Architectures the answer is no | 21:48 |
| chrisccoulson | micahg - thanks. it doesn't matter too much, as it looks like the webkit build failure is a trivial issue anyway | 21:48 |
| micahg | chrisccoulson: it'll matter is the rdepends FTBFS though :) | 21:48 |
| chrisccoulson | right, i can't build libproxy, enable the gnome component in libsoup and build epiphany until webkit has built ;) | 21:49 |
| micahg | chrisccoulson: I mean the xul rdepends :) | 21:49 |
| chrisccoulson | ah | 21:50 |
| jdstrand | fwiw, imo if it didn't build before don't worry about it. if say, epiphany did build on powerpc and doesn't because say webkit doesn't build, that is a problem | 21:50 |
| jdstrand | s/and doesn't/and doesn't now/ | 21:50 |
| asac | jdstrand: chrisccoulson: you got mail? | 22:01 |
| chrisccoulson | asac - not yet | 22:02 |
| chrisccoulson | asac - oh, actually, i did ;) | 22:02 |
| chrisccoulson | my mail filter moved it in to my mozilla mailing list folder ;) | 22:02 |
| chrisccoulson | i probably should have checked there first | 22:03 |
| asac | cool | 22:03 |
| fta | is it clear enough? http://paste.ubuntu.com/441058/ | 22:05 |
| fta | ..that it will be wiped out at the end of the session | 22:05 |
| BUGabundo | asac: https://edge.launchpad.net/~network-manager/+archive/trunk you lazy arse :) | 22:09 |
| fta | ? | 22:11 |
| asac | BUGabundo: please ping cyphermox | 22:13 |
| asac | that should have been set free already | 22:13 |
| asac | fta: what is the temp provile for? | 22:14 |
| asac | fta: i would say: "--temp-profile" | 22:14 |
| fta | asac, debugging, testing stuff with a fresh profile, .. | 22:15 |
| fta | i wanted to express the idea that it will be removed | 22:15 |
| asac | right. but temp implies that better imo | 22:15 |
| BUGabundo | asac: so you basicly don't do any more work anymore? just delegate ? :D | 22:15 |
| asac | BUGabundo: i want to give others a chance to step up .. | 22:16 |
| BUGabundo | okay | 22:16 |
| fta | asac, http://paste.ubuntu.com/441065/ ? | 22:20 |
| asac | not sure if clean is the right word ... maybe "Start with a new and temporary profile" | 22:21 |
| asac | i think clean is just computer slang | 22:21 |
| asac | or "Start with an empty and temporary profile" | 22:21 |
| asac | but in the end it doesnt matter much | 22:22 |
| asac | i think typical use is: BUGabundo complains that everything crashes; you say: try with --temp-profile ;) | 22:22 |
| BUGabundo | yep | 22:22 |
| BUGabundo | and it works | 22:23 |
| asac | oh ... you already tested. nice. | 22:23 |
| BUGabundo | asac: alias chromiumnew='chromiumdatadir="$(mktemp -d)";chromiumdiskcache="$(mktemp -d)";chromium-browser --user-data-dir=$chromiumdatadir --disk-cache-dir=$chromiumdiskcache' | 22:25 |
| fta | will be unnecessary tomorrow | 22:26 |
| BUGabundo | don't care | 22:27 |
| BUGabundo | wfm | 22:27 |
| BUGabundo | :) | 22:27 |
| BUGabundo | and 'should' always work , with any version of chromium or chrome in pretty much any OS | 22:28 |
| fta | i like jdstrand's idea too: http://penguindroppings.wordpress.com/2010/05/17/browser-profiles-in-chromium/ | 22:28 |
| asac | yeah ... multi profile is good | 22:31 |
| asac | but its perceived as too power userish ;) | 22:31 |
| fta | most users won't even notice, users never read man pages | 22:32 |
| asac | right. and for firefox its just legacy afik ... they officially dont support multiple profiles | 22:35 |
| BUGabundo | they aren't?? | 22:40 |
| BUGabundo | that's a major loss | 22:40 |
| chrisccoulson | bah, we have far too many extensions in the archive ;) | 22:45 |
| asac | chrisccoulson: right. thats what i meant | 23:06 |
| fta | oh my, the new ffmpeg killed mplayer | 23:11 |
| fta | mplayer: relocation error: mplayer: symbol codec_wav_tags, version LIBAVFORMAT_52 not defined in file libavformat.so.52 with link time reference | 23:11 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!