/srv/irclogs.ubuntu.com/2012/07/03/#ubuntuone.txt

=== facundobatista_ is now known as facubar
=== zumbi is now known as Guest42808
JamesTaitGood morning all! :-D08:32
ryeaway08:42
=== yofel_ is now known as yofel
=== zyga_ is now known as zyga
=== Guest42808 is now known as zumbi
gatoxgood morning!11:04
daveonearthCan anyone tell me why the Ubuntu One logo is not in place for the iphone app?12:10
alecutop of the morning to y'all!12:16
alecudaveonearth: how is it not in place? can you take a screenshot?12:17
gatoxalecu, hi12:20
daveonearthHi alecu12:23
daveonearthIt is just not the new Logo12:23
daveonearthits an orange file with a standard 'synch' round arrow12:24
ralsinagood morning!12:25
gatoxralsina, hi12:30
* briancurtin coffee, back in a few minutes14:22
alecummcc: this branch is not being merged by tarmac because it lacks "commit message": https://code.launchpad.net/~mikemc/ubuntuone-control-panel/fix-1018614-darwin-raise/+merge/11249314:24
dobeybrb14:35
* briancurtin back14:37
thisfredme15:00
gatoxme15:00
briancurtinme15:00
gatoxralsina, alecu mmcc dobey standup?15:04
gatoxdid i forget anyone?15:04
ralsiname15:04
gatoxralsina, no, i did'nt forget you :P15:04
gatoxjejejeje15:04
thisfredDONE: https://code.launchpad.net/~thisfred/u1db/unconfuse_sync_info/+merge/113214  bug #1019333 TODO: bug #1019333, BLOCKED: no NEXT: gatox15:05
ubot5Launchpad bug 1019333 in U1DB "When InvalidGeneration is raised proceed with sync anyway" [High,Confirmed] https://launchpad.net/bugs/101933315:05
dobeyme15:05
gatoxDONE:15:05
gatoxReviews, and refactoring (finally is looking good)15:05
gatoxTODO:15:05
gatoxFinish with refactoring TODAY!15:05
gatoxBLOCKED:15:05
gatoxNo15:05
gatoxbriancurtin, go15:05
briancurtinDONE: fixed up stringio branch, started on py3-unicode part 215:05
briancurtinTODO: work on py3-unicode part 215:05
briancurtinBLOCKED: no15:05
briancurtinNEXT: ralsina15:05
alecume (writing notes)15:06
mmcc`me15:06
dobeyralsina: ?15:06
ralsinaDONE: tech leads, mgmt call, 1-1s cmake tweaking, canonicaladmin TODO: cmake tweaking if needed, bug triaging (are late there) BLOCKED: no15:06
ralsinasorry, was otp15:07
dobeyDONE: 3.0.2 SRU uploads,15:07
dobeyTODO: finish cmake3 review, bug poking15:07
dobeyBLCK: None.15:07
dobeyalecu: go15:07
alecuDONE: some reviews, struggled to get a running buildout on mac, techleads mumble15:08
alecuTODO: catch up with briancurtin before his holidays15:08
alecuBLOCKED: no15:08
alecuNEXT: mmcc15:08
mmcc`DONE: path fixes,investigated how we launch SD on windows15:08
mmcc`TODO: improve fix for change to shebangs15:08
mmcc`BLCK: none15:08
mmcc`NEXT: ?15:08
mmcc`NOTE: need to be afk a bit today so only partial holiday tomorrow15:09
ralsinaComments?15:09
alecummcc`: this branch is not being merged by tarmac because it lacks "commit message": https://code.launchpad.net/~mikemc/ubuntuone-control-panel/fix-1018614-darwin-raise/+merge/11249315:10
ralsinaTomorrow is US independence day, monday is Argentina's15:10
dobeyindeed15:10
mmcc`alecu, ok, I'll go fix that. thanks15:10
alecuralsina: and weekend is pycamp!15:10
ralsinaplus, most of the .ar gang is taking friday off15:10
* dobey will be off tomorrow15:10
ralsinaso, reviews are going to be hard to find. Ping mercilessly.15:10
ralsinaEOM?15:12
gatoxeom!15:12
gatoxand start of lunch15:13
=== gatox is now known as gatox_lunch
mmcc`So, we have some apparently dead code on windows: in u1client, platform/tools/windows.py uses the wrong registry key to find the path to syncdaemon in get_sd_install_path(), which will (I think) never get called because checking is_already_running causes the tcpactivation code to spawn SD using a path it gets get_activation_cmdline, imported from SSO...15:17
mmcc`had some fun with this last night :) but in order to make sure SSO client doesn't need to know the path to syncdaemon, I wanted to make u1client use get_sd_install_path everywhere and so I needed to know which reg key was right15:19
briancurtinmmcc`: i think the tcpactivation and get_activation_cmdline stuff is newer, or at least more recently touched, so there probably is some older stuff remaining that could possibly be removed15:21
briancurtini believe that was made smarter or bulked up or something more recently around proxy work15:21
mmcc`briancurtin: yes, it is newer. So, there's no way any reg keys would be getting set aside from that installer XML file?15:22
briancurtinmmcc`: not that i know of off the top of my head15:22
mmcc`ok, good15:22
briancurtinmmcc`: a quick check for anything registry related would be to grep for winreg and you'll find everything15:23
briancurtinand everything is not much :)15:24
mmcc`briancurtin: yeah, I did that once, will double-check15:24
dobeyok, need to get lunch, bbiab15:44
=== zyga is now known as zyga-food
=== gatox_lunch is now known as gatox
=== mmcc_ is now known as mmcc
ryeping ralsina, if SD stops working right after 2012-06-17 22:48:57,953 - ubuntuone.SyncDaemon.OffloadQueue - DEBUG - Using temporary file: 'c:\\users\\username\\appdata\\local\\temp\\tmp3dttxi' - what can I do to get more information about this - CP can't contact SD and SD seems to have reached nirvana, but not in a good way16:45
briancurtinrye: does that file exist?16:48
briancurtin(i don't know the answer to the general question, just looking through the code)16:48
alecurye: what is the version of u1 that's giving that error?16:51
ryealecu: 3.0.2 around 2012-06-19, ralsina by the way, is current 3.0.2 writing 3.0.2 in the log file?16:54
ryebriancurtin: that file is not supposed to exist - that's a temporary one for offloading the in-memory request queue16:54
=== m_conley_away is now known as m_conley
ryeverterok: it looks like tritcask usage can trigger a bug in eCryptfs causing the metadata item to become unreadable17:01
ralsinarye: AFAIK, having it say 3.0.2 is in the pre-release checklist :-)17:02
verterokrye: hmm, is there something we can do about it?17:06
ryeverterok: ask ecryptfs not to cause data loss?17:08
verterokrye: jaja17:08
verterokrye: do you know what action by tritcask is causing ecryptfs to loss data?17:09
mandel_alecu, ping!17:09
mandel_gatox, any reviews I can do while I listen to bad very bad italian tv?17:09
ralsinamandel_: is there any other kind of italian tv?17:09
gatoxmandel_, not right now.... i'm finishing with the refactor TODAY... so i'm fully focus on that and nothing new17:10
ryeralsina: yet the logs have 3.0.0, please add this to check list again :)17:10
mandel_gatox, still refactoring! wow!17:10
ralsinabriancurtin: see what rye said ^17:11
mandel_ralsina, I don't think so.. but if there is, lets called x :)17:11
briancurtinis this only a windows thing?17:11
alecumandel_: pong!17:11
gatoxmandel_, i needed to restart the test refactor for darwin4.... everything was broken17:11
ralsinabriancurtin: yes, on linux, clientdefs.py is generated by autotools, on windows it has to be donemanually17:11
ryeverterok: not really, in all cases people weren't doing much, and then SUDDENLY ecryptfs corrupts the lower file17:11
briancurtinok17:11
* rye relocates again17:11
mandel_alecu, did you have the time to look at my fsevents branch?17:11
verterokrye: oh, that's nice (not) :-(17:12
alecumandel_: no. Get back to europython, get out in florence, meet tourist chicks. You'll have your review when you get back :-)17:13
alecumandel_: how's europython so far, btw?17:13
mandel_alecu, I can multi task hehe :)17:14
mandel_alecu, but ok, I'll be around if you need me17:14
mandel_alecu, I went to a couple of advance talks that I found very interesting, the keynotes where very similar form the pycon ones so I did find them more boring17:15
mandel_alecu, lots of pypy happening which is really great to hear about17:15
briancurtini'll be back in a bit. combined lunch + bike shop run17:18
alecumandel_: more pypy is awesome, yes.17:20
mmcchi mandel_ - hope you're having a good trip!17:21
mandel_mmcc, hello!17:22
mmccmandel_ this path change is more fun than we expected - did you know we had functions for finding the path to syncdaemon in both u1 client and sso client? and only the one in sso client actually works on windows?17:22
mmcc(if you don't know, now you know… )17:22
mandel_mmcc, no, I had no idea.. oh my god!17:23
mandel_mmcc, that is terrible (that it happens) is great that you found it :)17:23
mmccheh. not sure I'd go so far as terrible, but yeah it's good to get it cleaned up17:24
mmccso i'm working on a branch that makes u1client find its own path to sd, so SSO doesn't need to know about where sd is installed17:24
mmccand it uses the right registry key on windows17:24
mmccI have found many skeletons on this journey17:24
mandel_mmcc, was it using the wrong registry key?17:26
mandel_mmcc, what the.. how did we even work on windows?17:26
mmccmandel_ - the one in ubuntuone/platform/tools/windows.py was.17:26
mmccthat function was called get_sd_install_path17:26
mmccthe function get_activation_cmdline in SSO uses the right one17:26
mandel_mmcc, but is using SD_INSTALL_PATH, is not that correct?17:27
mmccno17:28
mandel_mmcc, and what is that?17:29
mmccthe one that we set in the windows installer is "path-ubuntuone-syncdaemon"17:29
mmccthe reg key "SyncDaemonInstallPath" is never set anywhere17:29
mandel_mmcc, that looks like someone changed it in sso and did not bother to change it in the tools from ubuntuone.. I hate that we have that much until code in sso17:30
mandel_mmcc, if you need reviews let me know, now I'm going to tr and find pizza, it cannot be hard in italy :)17:30
mmccthe only way this works is that tcpactivation is starting the service using the command line from its description, and the start() method in tools/perspective_broker.py is never called17:30
mmccmandel_ yes it shouldn't be in sso at all, I'm moving it back into u1client17:31
mandel_mmcc, superb, that is a much better idea, let me know if you need any help17:31
mandel_ok, I'm off to visit florence...17:32
mandel_all, laters!17:32
mmccmandel_ I would like your confirmation that I'm understanding tcpactivation correctly,17:32
mmccwhoops17:32
mmccI'll check later(s)17:32
briancurtinralsina: 1-1?18:24
ralsinabriancurtin: sure, let's do it on IRC18:24
ralsinabrb errands19:18
gatoxalecu, ping19:33
alecugatox: pong19:36
gatoxalecu, i finish with the refactor, and everything is working..... the problem is that it says that the diff is 2420 lines (+874/-1003) 7 files modified19:37
gatoxbut..... almost all of them are deletions19:37
alecugatox: are those the test being refactored?19:40
gatoxalecu, yes19:41
gatoxthe original branch without the refactoring was 839 lines19:41
gatoxalecu, so.... is that ok?? or do you have any magic idea?19:44
gatoxalecu, this is the branch: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin4-fsevents/+merge/11191419:44
alecugatox: what were the changes added before refactoring the tests?19:49
alecugatox: can't you split those things in two?19:49
alecugatox: the 839 lines first, then the refactoring of the tests.19:49
gatoxalecu, yes..... but it would be a lot of code...... without tests...... and then another branch with the tests19:50
gatoxahhhhhhh19:50
gatoxsooooooo..... you want me to do the refactoring in another branch19:50
gatox?19:50
alecugatox: I thought that was what we discussed with you and mandel!19:50
alecugatox: or perhaps I got that backwards?19:50
gatoxno..... we discuss that i should do the refactor in the same branch to not land ugly code19:51
gatoxok..... i'll propose 2 branches..... give me a couple of minutes19:51
* gatox reverting.....19:51
gatoxagain....... :S19:51
alecugatox: I thought that was at -3... and you and mandel  convinced me of doing all the refactoring fo the tests later19:52
gatoxi would prefer to do it in a diff branch :P19:52
alecugatox: you should *demand* mandel to give the bzr pipeline talk again.19:52
gatoxalecu, i don't trust in mandel19:53
gatoxjejejeje19:53
alecugatox: neither do I !!!19:53
gatoxalecu, so, now this is without the refactoring: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin4-fsevents/+merge/11191419:54
gatoxnow i'll propose another one with the refactoring19:54
gatoxalecu, you'll need to wait that launchpad updated19:54
gatoxupdates19:54
gatoxalecu, ok..... so this one is the final one for filesystem notifications: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin4-fsevents/+merge/111914    ....... and this one is for the tests refactoring: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-tests-refactoring/+merge/11328920:02
gatoxthe last one has some conflicts..... i'm fixing it20:04
gatoxalecu, but it also shows a lot of lines in the diff: 2654 lines (+672/-1489) 3 files modified20:06
gatoxbutttttttt..... 1489 are deletions20:07
gatoxalecu, there are the two branches.... now i'll email mandel about this (he says he like to review branches while he listen to talks :P)20:08
gatoxok..... email sent..... and EOD for me!20:10
alecugatox: ???? how comes the second branch is bigger than when they were merged?20:10
gatoxalecu, magic20:11
gatoxalecu, probably..... because some of the lines in darwin 4 are tests that i'm adding..... and i the next one i'm removing those lines to refactor them... or something like that20:12
alecuack20:13
gatoxalecu, i already sent an email to mandel.... so.... i'm off for today (and happy! :D)20:13
gatoxalecu, let me know how the review goes :P20:13
dobeyneed to run. later all.20:46
alecubriancurtin: a couple more lint issues here: https://code.launchpad.net/~brian.curtin/ubuntu-sso-client/py3-urllib/+merge/11242921:00
briancurtini wish lint was actually useful, or it worked on windows21:01
briancurtinor something21:01
alecubriancurtin: sorry about that...21:04
briancurtinalecu: it's not your fault! none of these changes show lint notices on my box, which is weird21:05
briancurtinso i think everything's fine when i push the branch21:05
alecubriancurtin: I usually run my windows virtualbox from ubuntu, and I share the source tree from the host to every devel guest.21:05
alecubriancurtin: perhaps that can help you with this.21:05
alecubriancurtin: I'm using virtualbox's "shared folders" but surely vmware has something similar.21:05
briancurtini run directly on windows. now that i have a better VMware setup i should probably try again to get a dev environment running on ubuntu21:06
alecubriancurtin: don't worry about that just now... but I can give you a hand with that after your holiday if you need.21:07
briancurtinlast time i tried to set it up, something about LXC didn't work or there was a version mismatch...ended up never figuring it out21:07
briancurtinbut that was months ago21:07
alecubriancurtin: lxc as in "linux containers" ? weird.21:08
briancurtinyeah, that was a part of the directions and someone helped me try to set it up, but we couldnt get it running21:08
alecubriancurtin: that's a bit weird. lxc is a way to do lightweight virtualization, to have linux running on top of linux, and it's mostly used instead of vmware... I do not use lxc at all right now, just heavyweight VMs with windows and all ubuntu versions running on top of a single Precise.21:13
mmccalecu, I'm wondering if there's a HOWTO somewhere for using xbox shared folders with source trees - I tried to set that up a while ago but ran into permissions issues in the goest...21:16
mmccs/guest/goest/21:16
mmccer. you know… guest21:16
alecummcc: right. And virtualbox, right?21:16
alecu:-)21:16
mmccyes. hah!21:16
mmccI typed v box, and os x 10.7 autocorrects like iPhones do :(21:17
alecummcc: hahahahahaah21:17
mmccearlier today I typed 'd obey' and it wrote 'dopey'21:17
mmcc:(21:17
aleculolz21:17
alecummcc: I don't have a HOWTO, but here's what I do:21:18
alecummcc: I just mount my ~/canonical folder from my host in my ~/canonical folder in the guest21:18
alecummcc: and I always run bzr from the host21:19
alecummcc: if I run it on the guest, then the guest would say that every file was added the execute bit21:19
alecummcc: so my rule is to never use it there.21:19
alecummcc: but I can edit from any guest though.21:20
mmccalecu, interesting. I'll have to give that a try21:20
briancurtinalecu: py3-urllib should be corrected, at least for the lint warnings related to my stuff (the TODO are others)21:26
alecubriancurtin: great. Yes, the other ones are there in trunk too.21:26
alecubriancurtin: in py3-StringIO there seems to be an unneeded "import sys"21:27
briancurtinalecu: where at?21:28
briancurtinalecu: ah i see21:28
alecubriancurtin: +1 to py3-urllib21:36
briancurtinalecu: thanks!21:36
mmccOk I have to run again, sorry I've been out of touch a bit today. I'm going to work more tonight, should have branches for this path setup problem tested and ready tomorrow. I'll be working tomorrow morning.21:46
mmccs/this problem/all these problems/21:46
alecubriancurtin: I think that while replacing u"..." -> "" in any given file, we should also replace any "" -> "".encode("utf-8").22:08
alecubriancurtin: for instance, in one of the test files we are checking mixing both types of strings, and replacing only one makes no sense...22:09
alecubriancurtin: let me find it.22:09
briancurtincould be. unicode/bytes/str isn't something i've ever had to care about so weird stuff probably slipped in22:10
alecubriancurtin: for instance, test_get_simple_token_name and related.22:10
alecubriancurtin: btw: I think that "compat.text_type" looks weird, but I think we'll get used to that :P22:11
alecubriancurtin: it beats having to guess whether the file was updated or not.22:12
briancurtinyeah it's not the greatest22:12
alecubriancurtin: hhmmm... I seem that we *can't* blindly replace "" -> "".encode("utf-8")22:15
alecubriancurtin: either22:15
alecubriancurtin: it looks like some tests already assume the implicit unicode to bytes comparison as in py2.22:15
briancurtinalecu: here is the remaining diff of parts that i haven't covered - https://pastebin.canonical.com/69391/22:18
briancurtini haven't looked into this part yet, but i dont think we need to do anything about the filter/list-comp change since we just use the result as an iterator anyway22:19
briancurtinthe self.next() vs next(self) part looks like a 2to3 confusion, since i think that next() call is something inside PyQt IIRC. we will probably have to tackle that part once we can actually run with PyQt on Py322:20
alecubriancurtin: right.22:20
briancurtini don't know a good strategy to port the metaclass stuff, haven't really looked into that yet22:20
briancurtinand with that, it's EOD. i would like to stick around but i have to go do a bunch of last minute shopping before the vacation22:22
alecubriancurtin: have a great time!22:22
briancurtinalecu: will do. see you soon!22:22

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!