[00:00] wgrant: if you could read my email and tell me which stuff i need to pay attention to, that would be great :/ [00:00] mwhudson: Which email? [00:00] Oh. [00:00] wgrant: my inbox :) [00:00] nm [00:00] Right. Just worked that out. [01:28] rockstar: I'm back if you are around [02:17] hey wgrant, can you try to give me a little help here? :) [02:18] Ursinha: Sure. [02:19] wgrant, I've added a field to the Product table, where on earth should I declare it to access it in translations side? [02:20] Ursinha: Are you getting ForbiddenAttribute exceptions? [02:20] wgrant, yes sir [02:20] * thumper away again [02:21] Ursinha: It depends on how the particular object is set up. In some cases there will be interfaces like IProductPublic, IProductEdit, IProductAdmin which have attributes protected by different permissions. In other cases, lib/lp/SOMETHING/configure.zcml will name a permission for each attribute. [02:21] Let's see which case IProduct is... [02:22] Ursinha: OK, so if you look at lib/lp/registry/configure.zcml around line 1040, you'll see security stuff. [02:24] let me see [02:26] You'll see . That restricts access to the attributes and methods on Bar to those users holding the launchpad.Foo permission. [02:26] restricts setting of attributes 'foo' and 'bar' to launchpad.Foo. [02:37] argh, right, sorry wgrant, got lost here [02:37] Still lost? [02:37] about other bit [02:37] Ah. [02:44] wgrant, thank you very much. [02:44] again. [02:44] np [03:12] sinzui: did you get that QA done? [03:12] I'm checking staging currently but the css issue still seems to be there :/ [03:13] Ah, staging is on r8640... np. [03:56] aargh soyuz is eating my brain [03:57] nom nom nom [03:58] mwhudson: Which bit? [03:58] (it is very effective at that) [03:58] wgrant: the bit you told me to run away from [03:59] Ah, yes. [03:59] Got it to run or do anything yet? [03:59] nah, haven't even tried that [04:02] wgrant: there seems to be a "build style" concept, of which the only actual instance is debian style? [04:02] mwhudson: Correct. [04:02] mwhudson: You can see it registered in daemons/buildd-slave.tac [04:02] (that bit took a while to find) [04:03] wgrant: right [04:52] wgrant: soyuz builders get the sourcepackage bits that need to be built from the librarian right? [04:53] mwhudson: Sometimes. [04:54] mwhudson: Unless the files are private. [04:54] oh right [04:54] In that case, they live in the restricted librarian and cannot be retrieved. [04:54] so how does it work then, do you know? [04:54] buildd-manager instead gives them a private-ppa.launchpad.net URL with credentials, IIRC. [04:55] oh ok [04:55] wgrant: the chroots are in the librarian too? [04:55] Yup. [04:57] mwhudson: You can probably reuse the same chroots, and just install bzr-buildpackage inside them during the build. [04:58] wgrant: yeah, not having to open that bear pit would be nice [04:58] It will make $UOSA happy. [04:59] wgrant: unlikely. he's a grumpy old bugger. [05:00] :( [05:00] I must try to catch him tomorrow morning, since I failed today. [05:00] spm: s/happy/less grumpy/ [05:01] tbh, I don't think I've ever seen lamont unhappy. irritated yes, unhappy, no. [05:01] I'm sure that maintaining an extra set of buildd chroots would tip him just over into unhappy :P [05:01] spm: the baseline for admins is not especially happy after all [05:02] mwhudson: gah!!!! do you mean to imply that I'm -EHAPPY ???? [05:03] wgrant: if you present them as hpia mk2's you could swing it with him being non the wiser? :-P [05:04] Ooh, true. [05:05] mwhudson: Deciphered the beast yet? [05:06] wgrant: well, making progress [05:12] wgrant: is there a difference between the slave scanner and the buildd master, or are they different names for the same thing? [05:13] mwhudson: slave-scanner was scaling badly, so was reincarnated as buildd-manager. Both are builddmasters. [05:13] slave-scanner is, I believe, only of historical value. [05:13] wgrant: ahh ok [05:17] wgrant: and aiui build scheduling is done be yet another process that manipulates the buildqueue table which is then consulted by builddmanager? [05:17] mwhudson: mrrrrrrh sort of but not really [05:17] mwhudson: There is this queue-builder thing. [05:17] It is run occasionally. [05:18] It only really needs to be run when you add a new arch or distroseries. [05:18] It creates new builds, and creates buildqueues for those that are missing. [05:18] But in the normal case, the Builds and BuildQueues are created and given a score at upload-time. [05:19] does it need to be run when a build is rescored for example? [05:19] No. [05:19] Rescoring directly updates the BuildQueue, which findBuildCandidate then sees. [05:19] or not, because the score is the output of the at-upload computation and rescoring overrides the result [05:19] Right. [05:20] queue-builder has a rescoring mode, where it will recalculate all scores. [05:20] But I don't think queue-builder is run regularly in production at the moment. [05:28] are bug subscriber sprites (not those in the "Also notified" section, though) slightly raised for anybody else)? === spm changed the topic of #launchpad-dev to: LP down 0900 UTC - 1030 | This is Launchpad Development Channel | Week 4 of 3.1.10 | PQM is release-critical only - release manager is noodles775 (flacoste, BjornT) | I am Zero OOPS and So Can You! http://is.gd/4fkLl | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting | On-call review in #launchpad-reviews | Use http://paste.ubuntu.com/ for pastes | This channel is logged: http://irclogs.ubuntu.com/ [05:35] wgrant: sorry, i'll stop this soon [05:36] wgrant: does the buildd-manager suck the built files off the slaves too? [05:36] mwhudson: Hey, it's more interesting than Java revision.. [05:36] It does. [05:36] There's a handler in BuildQueue to do that, I think. [05:37] (the slave will report its state as WAITING when it is complete) [05:37] Hm, no, not in buildqueue. [05:37] But it's somewhere, anyway. [05:38] buildd-manager will grab the files, tell the slave to clean up, and then process the binary upload. [05:38] wgrant: lp.buildmaster.buildergroup it seems [05:38] Something like that. [05:40] # XXX cprov 2007-07-11 bug=129487: untested code path. [05:40] yay! [05:41] This is Soyuz. [05:42] There are altogether too many XXXs in that method. [05:42] But bigjools wants to offload most of the processing, anyway. [08:32] good morning [09:10] noodles775: What's wrong with the librarian? Disk upgrades? [09:18] mo disk IIRC [09:18] Makes sense. [09:19] bigjools: I wonder how much librarian disk would be saved if daily PPAs were configured to cull binaries after just a few days. [09:20] lots! [09:20] I might chop the current 28 day limit to much less [09:20] across the board [09:21] however the biggest culrprit by far is Ubuntu [09:21] Morning [09:21] eyup mrevell [10:14] 10:11 < SteveA> the launchpad offline page is really really annoying [10:14] 10:11 < SteveA> instead of serving a page up at the URL I want, but serving the offline page with a 5xx error [10:14] 10:12 < SteveA> it redirects me to an "offline" page [10:14] 10:12 < SteveA> the problem with this is, I lose any URLs I have in my browser when I refresh the URLs [10:14] 10:12 < SteveA> such as when I restart my firefox session [10:14] 10:12 < SteveA> this is a significant loss of state. [10:14] 10:13 < SteveA> the offline page is even a 200 OK [10:14] 10:13 < SteveA> it should be a 5xx [10:25] I guess if haproxy doesn't support what we want, we can open a bug and maybe submit a patch. === mthaddon changed the topic of #launchpad-dev to: This is Launchpad Development Channel | Week 4 of 3.1.10 | PQM is release-critical only - release manager is noodles775 (flacoste, BjornT) | I am Zero OOPS and So Can You! http://is.gd/4fkLl | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting | On-call review in #launchpad-reviews | Use http://paste.ubuntu.com/ for pastes | This channel is logged: http://irclogs.ubuntu.com/ [11:01] Morning, all. === danilo-afk is now known as danilos === matsubara-afk is now known as matsubara [11:32] james_w: fancy a trip to NZ in january? [11:33] I wouldn't say no :-) [11:33] james_w: we're holding the build from branch sprint in Wellington [11:34] nice [11:34] so it's up to you, it will be a mostly coding sprint I think [11:34] LCA is the week after, which is why we're doing it there [11:34] thought so [11:48] jtv: around? [11:48] bigjools: around [11:49] bigjools: what's up? [11:49] jtv: howdy - I was wondering if you'd like to get involved in documenting how our build farm works, so you can learn a bit about it in preparation for the work we're doing to support translations jobs [11:49] bigjools: sounds like a good idea [11:50] bigjools: got any concrete plans for it? [11:50] al-maisan and noodles775 are looking into it as well, can you join up with them [11:50] mwhudson had fun with lp-buildd today, too! [11:50] I hear :) [11:51] al-maisan, noodles775: how are you going about it? (Or rather, how should I :-) [11:52] jtv: I haven't yet (still following the release issues), but plan to install soyuz locally using wgrant's instructions (I've never had time to do it yet) [11:53] noodles775: I rewrote them to be more useful this evening, but haven't uploaded them anywhere yet. [11:53] jtv: I am also a bit busy but thought of reading the code and playing with the Soyuz dogfood system. [11:53] noodles775: oh, I hadn't even realized that there was an additional installation procedure [11:53] wgrant: ah, please let me know when you do... [11:54] jtv: I'll include you on an email I am sending about the db schema [11:54] jtv: for actually running the build infrastructure locally. [11:54] jtv: you need to get a local buildd running [11:54] wgrant: as you've guessed by now, please add me to the distribution list. :-) [11:55] bigjools, noodles775: were you deliberately completing each other's sentences there, or should I read you as two separate input channels? [11:55] heh, separate :) [11:56] prophecy - Guns n Roses "Back off Bitch" just started on my random playlist [11:58] noodles775, jtv: http://williamgrant.id.au/f/1/2009/running-soyuz.html is the new version. [11:58] wgrant: thanks! [12:01] nice [12:02] wgrant: it would be good if you can put that on dev.launchpad.net [12:02] bigjools: I've been planning to. [12:03] coolio [12:03] Any hints as to where it should go? [12:03] somwhere under https://dev.launchpad.net/Soyuz [12:04] * wgrant hunts out the Codehosting equivalent. [12:05] Thanks wgrant === barry` is now known as barry [12:32] https://dev.launchpad.net/Soyuz/HowToUseSoyuzLocally [12:32] \o/ [12:32] Nice :) [12:32] thanks muchly [12:33] * wgrant works out how to attach those two helper scripts. [12:36] Alright, that looks reasonable. [12:37] wgrant, btw, I think it'd be great to have sampledata on launchpad_dev updated to reflect what's there after update-ubuntu-sane.py script [12:37] danilos: Probably, now that it's separate from ftest. [12:37] wgrant, exactly :) [12:37] Amusingly, most of the tests use the 'ubuntutest' distro anyway, because the 'ubuntu' distro is too screwed. [12:38] wgrant, yeah, we're also trying to use more and more of in-place object creation instead of relying on sampledata [12:40] but the jury's out on whether that's quicker than resetting the db each test [12:41] I should also probably get the archive.launchpad.dev stuff added to rf-setup, like I did with ppa.launchpad.dev a couple of months back. === mrevell is now known as mrevell-lunch [12:44] bigjools: The external archive dependencies feature is very handy. [12:44] wgrant: that's its problem :) [12:57] I never like to see 'NIGHTMARE.' in specs. [12:58] ooi, when does the db-stable->devel merge take place? Is it scheduled, or somewhat adhoc? [12:59] * wgrant sadly compares the Codehosting and Soyuz local instance pages. [13:00] wgrant: NIGHTMARE is honesty :/ [13:00] bigjools: Yep. [13:00] wgrant: which ones do you like better? [13:00] ;) === mrevell-lunch is now known as mrevell [13:57] bigjools: ping [14:13] sidnei, around? [14:33] EdwinGrubbs: hey, wassup [14:35] bigjools: you mentioned in bug 196322 that the +allpackages could be removed. Will this prevent google from crawling the source packages pages, and is that a problem? [14:35] Bug #196322: "All source packages" page empty when distribution doesn't use Launchpad for packaging [14:36] let me check [14:38] EdwinGrubbs: blow it away [14:46] matsubara, meeting in 15 mins? [14:46] stub, Chex, gary_poster, rockstar, bigjools, danilos, sinzui, allenap: LP production meeting in 14 min @ #launchpad-meeting [14:47] matsubara, thanks for the reminder, I might not be a good participant (along with bigjools, gary_poster and sinzui) since we have a TL call at the same time [14:47] damn DST [14:48] matsubara, oops summaries won't be completed in time as well [14:48] noodles775: can you attend the production meeting instead of me please! [14:53] danilos, I can cover for gary_poster [14:54] bigjools, that's why I had complaints about moving the call last week, but forgot about the timing for this week :) [14:54] matsubara: thank you. what do you want to do about this longer term? Presumably we hould move one or the other [14:54] I was wondering if the team lead meeting on Thursday should be moved to the same time as Wednesday during DST [14:55] gary_poster, not sure what do about it in the longer term. we can try to re-schedule or send other people as QA contacts for each of the teams when the TL is not available [14:56] matsubara: ok. I'm sure we'll bring up the conflict on the TL call; I or someone else will then check up with you and Ursinha [14:57] salgado: sorta [14:57] gary_poster, all right. thanks [14:58] sidnei, hi there. I was wondering if you could help me with that lazr-js bug I mentioned yesterday. would you have some time for that today? [14:59] salgado: today is looking nasty, but if you have a reproducible case, let's schedule it for 4pm BRT [15:00] sidnei, I do have one, yes. thanks a lot! [15:00] salgado: deal. thanks! === salgado is now known as salgado-lunch [15:06] mrevell, I just realized I never replied to your tour email (it dropped off my unread emails) [15:07] hey beuno, there's no rush. I had a meeting with Iain Farrell yesterday, as I happened to be at Millbank for some other meetings. I'll CC you on the writeup of what we spoke about. [15:08] mrevell, thenks [15:28] abentley: around? [15:28] rockstar: around? [15:28] flacoste: Yes. [15:29] abentley: we have a problem with codehosting can you help? [15:29] If it's the fact that http access doesn't work, I'm already looking into that. [15:29] ok, what is the situation? [15:29] do you need any help? [15:30] I don't know what's causing it. It's hard to reproduce locally, but I'm getting there. [15:38] abentley: what info do we have on the issue? [15:38] The best info I have is that some files which should be there are 404ing and others are not. [15:39] abentley: btw, mthaddon was looking for you on #launchpad-code [15:40] flacoste, mthaddon: I lost my connection to irc.canonica.com, and it doesn't want to reconnect. [15:40] abentley: hmm, weird - well we can work here if you like [15:41] mthaddon: jam is reporting that not all files are 404ing, but I'm getting "connection timed out" when I try to access bazaar.launchpad.net. [15:41] mthaddon: This URL is reported to not 404: http://bazaar.launchpad.net/~bzr-pqm/bzr/bzr.dev/.bzr/branch-format [15:42] yeah, I don't get a 404 for that one [15:42] mthaddon: Is that possibly a caching issue? [15:42] and http://bazaar.launchpad.net/~bzr-pqm/bzr/bzr.dev/changes works for me too [15:43] I can't see any caching options in the apache configs [15:43] mthaddon: I guess it can always be a caching issue, with the prevalence of intercepting proxies. [15:44] mthaddon: But I tried it on devpad and it worked there too. [15:45] https://bazaar.launchpad.net/~launchpad-pqm/launchpad/production-devel/changes works for me too [15:45] so who is reporting problems, and what kind of problems? [15:45] mthaddon: Sure, but AIUI, we're worried about http codehosting, not codebrowse. [15:46] mthaddon: jam is one of the people noticing problems. [15:47] ok, so what specific bit is failing? [15:52] BjornT, ping re: lazr-js and buildout [15:52] hi mars [15:52] hi BjornT. Did your lazr-js and LP buildout changes land successfully? [15:53] mars: yes, it landed quite a while ago [15:54] BjornT, ok, so is it safe to bump the version number on the lazr-js trunk, and start landing sidnei's branches? [15:54] mars: yes [15:54] \o/ [15:54] BjornT, awesome, thank you [15:54] now, how do i land them? [15:54] ... [15:55] bzr pqm-submit? [15:55] sidnei, yes, that should work [15:55] it Works For Me [15:55] it didn't last time. maybe i have to be added to some acl somewhere [15:55] BjornT, ^ ? [15:56] sidnei: can you try again? or show me which error you got? [15:56] * mars wonders who is the scrollkeeper for the PQM system, the repository of such arcane knowledge? [15:57] the error i got was [] [15:57] Command failed! [15:57] All lines of log output:"Branch ['/home/pqm/archives/rocketfuel/lazr-js/toolchain', 'bzr+ssh://bazaar.launchpad.net/~launchpad-pqm/lazr-js/toolchain'] not found in config" [15:57] flacoste: We are seeing a gradual recovery. Do you have any examples of branches that are still failing? [15:58] sidnei: to which pqm address did you send the request to? [15:58] abentley: not really, mthaddon: do you have any example HTTP branches still failing? [15:59] BjornT: launchpad@pqm.canonical.com apparently [15:59] BjornT: rockstar pasted me the config [15:59] sidnei: i would try again [16:00] BjornT, how much longer are you on for? [16:00] Am I reading correctly the codehosting issue is slowly recovering? My commit from 2 hours ago is still not available via launchpad yet. [16:01] mars: no idea. i'll probably be on and off the whole evening, and night since i need to prepare for my trip tomorrow [16:01] ok [16:01] Ursinha, sorry I missed the meeting. I forgot my timezone changed. [16:03] rockstar, btw, we are working on YUI3 for LP, devmode is good, but the production JS build is broken. If I can fix that before the sprint, then we can land it. [16:03] but we do have an integration branch for people to use. [16:04] thumper could try that branch if he is feeling adventurous :) [16:05] abentley: http://bazaar.launchpad.net/~launchpad-pqm/lazr-js/toolchain/ [16:08] rockstar, okay [16:09] mars, let me know how I can help. === beuno is now known as beuno-lunch [16:11] uhm, my submission didn't even show up in pqm, no reply even. maybe it's just a delay [16:12] sidnei, there's funkiness in the web display recently. [16:12] BjornT: Are you able to view bzr+ssh://bazaar.launchpad.net/~launchpad-pqm/lazr-js/toolchain/ ? [16:13] rockstar, it was asked someone from code to comment on https://bugs.edge.launchpad.net/launchpad-code/+bug/475394 [16:14] Bug #475394: bzr http access broken after 3.1.10 rollout [16:14] if you or abentley could do that it would be nice [16:14] thanks [16:14] Ursinha, I'm trying to find out what's going on by reading the backchat. [16:14] abentley, https://code.edge.launchpad.net/~launchpad-pqm/lazr-js/toolchain works for me [16:14] Ursinha: We don't know what's wrong. === matsubara is now known as matsubara-lunch [16:16] barry: EdwinGrubbs: bac: can you close your 3.1.10 registry bugs? [16:16] abentley, shall we have a call to catch me up? [16:16] rockstar: Okay [16:16] sinzui: i just ran close-my-bugs :) [16:16] sinzui: sure [16:16] fab [16:17] Maybe I will find and fix the insane karma action for that action so people can safely close other people's bugs [16:17] barry: Do you think the rule is: use the assignee or the user doing the action? [16:18] sinzui: i think the person who sets fix committed should get the karma [16:18] fscking rich-roots kills me again! [16:19] I disagree. I find lots of bugs that you have indirectly fixed a long time ago. I assign it to you and mark it fix committed. you should get the credit === salgado-lunch is now known as salgado [16:19] sinzui: oic. assignee then [16:19] rockstar: ringing... [16:19] barry: I think I gave you and edwin credit for about 30 bugs I discovered in foundations last month. [16:19] uhm, now it shows up, but it's out of order. how can i remove a script from pqm? [16:20] sinzui: awesome. i can go out to lunch now! have you seen my new site www.karmaforgold.com? [16:21] \o/ [16:21] losas, does someone have a moment to help sidnei with his PQM submission? [16:23] abentley: yes, the bzr+ssh url works, but not http [16:26] BjornT: You don't have write access to that branch, do you? [16:31] bigjools, can I mark bugs 470150, 470411, 472326, 472608, 472929 as Fix Released? [16:31] hmm [16:32] Ursinha: some URLs would be nice :) [16:32] bug 470150, bug 470411, bug 472326, bug 472608, bug 472929 [16:32] Bug #470150: package set based archive permission checks need to consider the distro series [16:32] Bug #470411: please change the default source suite for Debian to 'testing' in _syncorigins.py [16:32] Bug #472326: When adding and removing package subsets the distro series needs to be observed [16:32] Bug #472608: IPackagesetSet.getByName() is missing LP API metadata for the newly added 'distroseries' param. [16:32] Bug #472929: Archive +upload URLs for package sets need to be distro series aware [16:32] :) [16:32] lol [16:32] Ursinha: well if you want to steal al-maisan's karma [16:33] yeah, all of these were rolled out. [16:33] bigjools, hm, I thought the assignee gets the karma [16:33] Ursinha: I think the person who marks it does [16:33] bigjools, this kinda sucks [16:33] yes indeed. [16:33] Ursinha: patches accepted I'm sure ;) [16:33] bigjools, I'm learning, hopefully will be able to do that soon :P [16:34] \o/ [16:34] * bigjools closes his bugs [16:35] yay for the API [16:37] noodles775, are we having a second rollout? [16:38] bigjools, so I won't close al-maisan's bugs so I won't steal his karma [16:38] rockstar: noodles775 has left for the day, he was up at 3am his time [16:38] I can close them later if that suits [16:38] al-maisan: I have a script that will close your own bugs in 10 seconds [16:39] bigjools, I don't remember that part being in the Release Manager job description... [16:39] bigjools: that's very handy. [16:39] mars: yeah it kinda sucks but he's a diligent man [16:40] * beuno-lunch hugs barry for filing the bzr bug [16:46] abentley: no, i don't think i have write access to that branch. however, branches that i have write access to fail as well, i tink [16:47] abentley: for example http://bazaar.launchpad.net/~bjornt/garmin-sync/devel [16:47] BjornT: Yes, it's just that as a debugging tool, we want to make sure the mirrored copy is alright, and the you see the mirrored copy over bzr+ssh if you don't have write access. [16:47] gary_poster, are you working on bug 475550 for the reroll? [16:47] Bug #475550: _pythonpath whitelist of clean_modules is too fragile [16:47] Ursinha: yes [16:47] gary_poster, thanks [16:48] Ursinha: np [16:55] BjornT, mars: pqm finally failed my branch, with the same error as before. i don't think it's because of release-critical though. [16:56] sinzui, I'm sorry, I think I forgot asking in the meeting, you said bug 475433 was a good candidate for CP. you meant CP or reroll? [16:56] Bug #475433: oops leaving milestone code_name empty [16:56] I meant CP, but now that I see the issue is lower, I am working on an RC [16:57] Ursinha: we may see many oops that end with: [16:57] AttributeError: 'NoneType' object has no attribute 'str [16:58] Ursinha: This may be happening with an not-required StrippedTextField [16:58] s/an/any/ [16:58] sidnei, it looks like dealing with that will have to wait until after they get codehosting back to full speed :( [16:58] mars: thats fine. i assume lazr-js will be taken out of pqm as well? [16:59] sidnei, that I do not know [17:01] beuno-lunch: ;) [17:02] rockstar: are we getting your magic migrate-my-branch button this cycle or the next? [17:03] * Ursinha stabs oops-tools [17:03] sidnei, probably the beginning of this next cycle, on edge. The infrastructure got rolled out with the rollout, but the UI didnt'. [17:03] rockstar: awesome, thanks! [17:06] sidnei: you'll have to ask a losa about that error. we will only take lazr-js out of pqm if we set up a buildbot for it. but it might be better to run the tests before commit, rather than after, assuming it won't take long to run them. === matsubara-lunch is now known as matsubara === beuno-lunch is now known as beuno [17:50] anyone else able to file bugs in lp? [17:50] hi deryck, a user filed bug 475498 and I'm also getting consistently timeouts trying to do what he was doing [17:51] deryck, I initially thought about another bug on dupefinder with long sentences but even a few words trigger the timeout [17:52] Ursinha, ok, I'll take a look at it. [17:52] thanks deryck [18:08] Night all [18:34] salgado: just got out of a meeting, and have to leave in 20, can you email me instead? [18:34] sidnei, sure, I'll do that [18:35] salgado: thanks! [18:36] sidnei, thank you [19:14] barry: close this bug I rediscovered and get the karma: https://bugs.edge.launchpad.net/launchpad-registry/+bug/102273 [19:14] Bug #102273: product search shows description instead of summary [19:16] yay karma! [19:29] EdwinGrubbs: I wrongly closed the bug I was looking for: bug 263223 relates to the layout you were looking at [19:29] Bug #263223: "Working on" and "Latest Memberships" inconsistencies on person page