[01:11] and support type people around? [01:26] thumper: not really, but maybe I can help you [01:26] ralsina: hi [01:26] ralsina: my server has stopped syncing [01:26] the U1 window says "File Sync starting..." [01:26] but nothing changes [01:27] where should I look? [01:27] thumper: did you recently add a huge number of files? Like, dozens of thousands? [01:27] yes [01:27] well... [01:27] not that many [01:27] 500 [01:27] ok, what version of ubuntuone-client do you have installed? [01:27] 500 should not be a problem :-) [01:28] how should I find the version? [01:29] 3.0.1-0ubuntu1.0.1 [01:29] from apt-cache policy [01:30] from precise-updates [01:32] ok, I suggest you file a support request and attach your logfiles [01:32] because we don't have any known problems that sound like it [01:32] thumper: logs are in ~/.cache/ubuntuone/log/ [01:33] * thumper looks === wgrant_ is now known as wgrant === zyga-afk is now known as zyga [08:57] Good morning all! :) === JamesTai1 is now known as JamesTait [11:18] good morning [11:24] gatox, morning! [11:24] mandel, hi [11:25] gatox, have you seen the comments here: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-fsevents-1/+merge/110382 [11:26] gatox, I'd love to land the refactoring you did regarding the pyinotify stuff etc (specially so that I merge my branch with that) but we are blocked my the refactoring you did with the tests :( [11:26] as alecu mentioned [11:27] didn't understand this part: "but we are blocked my the refactoring you did with the tests :(" [11:28] gatox, that I cannot land that until you fix that skip decorators work :) [11:28] mandel, ahhhhh yes..... also...... did you read the comment of alecu in darwin-fsevents-2? [11:28] gatox, so, I would have done the refactoring around and done a diff branch for the tests, but we know have to wait for those to things [11:29] gatox, yes, we have to talk :) [11:29] mandel, i'm reading alecu's mail [11:30] gatox, I think is better to not do everything in one step, copy if needed, refactor once it works and passes tests [11:30] one step at a time [11:31] mandel, ok..... i've read all the emails [11:32] * gatox thought he was almost done with fsevents..... so young and innocent [11:32] jejee [11:33] gatox, yes + code completion hehehehe [11:33] gatox, how are you doing that? [11:33] gatox, I mean, code completion :) [11:34] mandel, take a look: http://youtu.be/xCqiOEXfZd4 :P [11:34] we arre getting there [11:34] gatox, ok, but the backend, how do you store the info? [11:35] gatox, specially for a project [11:38] mandel, at this moment i'm saving the info only in memory for the opened files, i'm just starting with the part of the completion daemon..... where it resolves the symbols that couldn't be resolved in the first analysis of the code (because of imports, or variables pointing to symbols in the same file, etc), so the daemon explores the different modules, looking at everyone symbols based on the imports, and who call each function, etc..... [11:38] and try to resolve those cases..... then i'm going to stored the info collected with sqlite, using the paths of the files as keys.... and only update them and access them when necessary, so you don't have a lot of things being processed, or the memory full of stuff [11:39] and i'm doing alllllll the analysis of the code using a lot of tokenizer and ast [11:40] but i think i can reach a level where code completion will be really accurate [11:41] gatox, so, what you could do is not use the path of the files as keys but store in the db the info per module using the module as the key, for example, everywhere you use threading the info is going to be the same [11:41] gatox, what you can do is divide the data in two, that common to everyone (os, threading, etc..) and that just for the project [11:42] gatox, common info can be build when imported or you can even build/deploy the common lib info since should be present always [11:43] mandel, yes, the "storing" part i haven't resolve it yet... i'm still thinking about that..... saving it as module has sense...... but also i'm saving mostly information about the projects, not threading, because that is really cheap to calculate [11:44] not threading, os, etc i mean [11:44] gatox, the problem would be the index.. so you can on top of the normal autocomplete create a 'first hit cache' with the most common used methods/objects by the current user [11:45] gatox, oh, but being cheap to calculate I don't care.. I want it to be fast.. also, how do you deal with the objects being dynamic? if I do t = Thread() t = '' ? [11:45] mandel, i was thinking of load in memory the "completion info" of the file being currently edited in the editor [11:46] gatox, you must add a limit to that, you don't want the IDE to use a lot of memory, I think visual studio stores eveything on disk (but they cheat, they have compiled langs) [11:46] mandel, for that i'm calculating the distance going up based on the line number, and analysing the scope of the variable...... also i know is not going to be always 100% accurate..... and if it founds that is confuse between 2 values.... is going to suggest first the possible object.... and then the symbols for that one....... BUT THIS IIS NOT COMPLETE YET!! [11:47] gatox, just poking your brain, I know is not done yet :) [11:48] mandel, and it's the only part of ninja that has tests [11:48] jejeje [11:48] lol [11:48] mandel, ok.... going back to filesystem notifications [11:48] mandel, do you want to talk about something? [11:48] or should i start looking at the code and see how to refactor everything to create the sauron ring? [11:49] the good news..... is that based on the test, mac is working... the bad news..... is that we need to change all the implementations [11:50] gatox, uh... I'm fixing my bits to work correctly, when I do a move from a watch path to a not watch path is not currently converted in a delete (I think is the only bug) [11:50] gatox, I do have a small review request (which is useful for you too) :) [11:50] let me find the MP [11:51] gatox, here: https://code.launchpad.net/~mandel/ubuntuone-client/sys-init-done-darwin/+merge/110967 [11:51] gatox, that was harder to find than expected hehe [11:51] mandel, ok.... so i can do the reviews...... look at the code and the possible issues of the refactoring...... and we can wait for alecu to have a mumble with him about fsevents....... what do you think? [11:51] gatox, sounds perfect :) [11:52] mandel, okkkkkkkkk [11:52] mandel, so you are going to europython...... lucky bastard! [11:52] jejee [11:53] gatox, yes :) [11:53] gatox, oh, if you have the mac around, I'd appreciate reviews for the fsevents daemon.. I have lots left :( [11:53] gatox, and will propose a talk for pyconar :) [11:53] mandel, https://code.launchpad.net/~mandel/ubuntuone-client/sys-init-done-darwin/+merge/110967 JAJAJAAJ [11:54] mandel, +1...... i saw the talk yesterday.... and it has sense, because that has to be done in the package [11:54] mandel, yes? what are you going to propose? [11:55] gatox Como hacer programacion asincrona con coroutinas (greenlets) antes de que los deferreds de twisted te hagan hablar en gallego ;) [11:55] jejejejeje [11:55] mandel, start sending your branches for review in this direction [11:56] gatox, ok, let me make a list with the correct order :) [11:56] mandel, i know you want to come to pyconar because you can't keep away of argentina for so long [11:56] jejejee [11:57] lol [12:03] gatox, here is the list in the correct order: http://paste.ubuntu.com/1048968/ [12:03] mandel, roger that [12:03] mandel, start reviewing now, waiting for alecu [12:04] gatox, after the addition of the integration tests you can do sudo ./run-integration-tests which will start the fsevents-daemon and run tests that send events from the fs to python :) [12:04] mandel, it says that in the description of the branches? [12:04] gatox, I'll check [12:31] gatox, looks like it works, I just deleted all my u1 files :( [12:31] mandel, on purpose? [12:32] gatox, accident.. I used my account, not big deal I do have a back up :) [12:33] gatox: mandel, hello all! [12:33] alecu, hello! [12:34] alecu, hi! [12:34] * alecu catches up with the backlog [12:34] alecu, I have no picts but yes it works! [12:34] mandel: orsom! [12:34] alecu, I have logs to prove it :) [12:35] oh, and the inhibitor on mac has a bad stub.. I'll have to fix that I see an exception in the logs due to that [12:36] mandel, could it be that your branches depends of something of u1-client that is not yet in trunk? [12:37] gatox, why, are you getting errors? [12:37] mandel, ImportError: no module named txsocketserver in u1-dev-tools [12:37] gatox, is in trunk, you might not have the u1-dev-tool up to date in the mac [12:38] mandel, i just updated...... weird.... i'll try to delete it and branch it again [12:38] gatox, ok [12:38] alecu, gatox I'm off to have lunch, mumble with ralsina when I get back? [12:38] mandel, ralsina or alecu? [12:39] gatox, both, je ne se pa! [12:39] mandel, okok [12:41] alecu, gatox, I left sd syncing all my music back to the cloud to see if it fully works [12:41] mandel, ack [12:41] alecu, gatox, is from a samba share to the cloud, 14 gb.. shall be interesting to see the results [12:42] * mandel is scared it seems to work and is fast! [12:42] hmm [12:42] cloud is just a synonym for /dev/null [12:43] dobey, I have back ups for that.. and cloud == weather, right ? [12:43] hehe [12:43] ok, I'm off to have lunch [12:43] cloud == rain. that's what looking outside shows anyway [12:48] * alecu bbaiab [12:54] * alecu is back [13:00] mandel, i needed to change your script to run the python tests.... are you creating the script taking into account the buildout?? or there was any change in the buildout regarding u1lint and u1trial and i wasn't notify? :P [13:01] good morning! [13:01] hola boss! [13:02] oi [13:03] ralsina, hi [13:03] aloha *! === hito_jp0 is now known as hito_jp [13:12] mandel, PING PING! (when you are back :P) [13:14] i wonder if there is any way we can run qt tests without using qt4reactor [13:15] well, i guess "rewrite u1trial to not use twisted/trial" would be how :-/ [13:17] alecu, do you have time to mumble when mandel get back? [13:19] dobey: a test runner that did what we want and didn't require twisted would be a great thing to have. [13:19] yes [13:19] gatox: sure! [13:20] but it basically involves rewriting trial to not use twisted :-/ [13:20] alecu, cool [13:20] dobey: is there *no* test runner that takes deferreds from tests and doesn't use twisted? [13:21] it's not the deferreds that are the problem [13:22] dobey: if it's a short explanation, what is? [13:24] the problem is that all the infrastructure to do test running is in twisted trial. u1trial is just a very thin wrapper on top of that to do some extra stuff we need [13:24] so there's reactors which we need to deal with for various tests because our code is using twisted, and thus needs a reactor [13:24] and there's mainloop integration stuff, so we have to use certain reactors for certain tests [13:25] for some projects we probably don't really need a reactor, but for some we still do [13:27] fixing u1trial to not rely on twisted trial is no small/trivial thing. :-/ === zyga is now known as zyga-afk [13:50] * mandel back [13:50] gatox, I'm back! [13:50] mandel, did you read my comment about the tests in your branches? [13:51] gatox, it is present, but in one of the later branches.. dont know which one [13:51] gatox, I really hate that $u1trial solution.. [13:51] mandel, ahhhh.... so you change it to $u1trial and $u1lint later? [13:51] gatox, yes, it was after one of the comments from mmcc [13:52] buildout really needs to install dev-tools and dirspec from trunk, rather than tarblals [13:52] mandel, ahhhhh ok...... don't lie to me!! i'll check jejejeje [13:52] tarblahs [13:52] gatox, is there, I promise [13:56] gatox, where are the control panel logs written? [13:56] nothing, found them [13:56] ack [13:58] how is s/raise Exception(...)/# TODO: .../ a bug fix exactly? :) [13:58] ralsina, alecu, mmcc, gatox: I have sd running (it has uploading 3,5 gb with no problem) tested move and delete and everything works. Control panel does not start because it cannot find the binaries to execute, so we are in a situation where we need to start packaging this things asap to have it running [13:58] once we solve the binary problems we can move to have control panel working [13:58] mandel: yay! [13:58] atm, sd with my implementation works [13:59] mandel: awesome! Now, urgently, stop moving upwards and let's get all this landed! :-) [13:59] mandel: mmcc is not here yet, so let's bring it back when he arrives [13:59] alecu, I agree, landing all the code this week must be a priority [14:00] mandel: right. Also, be aware that there's a holiday in .AR tomorrow. [14:00] alecu, gatox, shall we mumble about the comments in the fsevents branches? [14:00] alecu, really.. ok had no idea.. [14:00] mandel, stop showing off! i also think that my implementation works.... at least all the tests are ok ejjee [14:00] haven't tested it yet :P [14:01] mandel, mumble now sounds good to me [14:01] alecu, ? [14:02] gatox: already there! [14:02] going [14:32] man, triaging bugs is not fun. nor his having to create these pretty commit messages for merging a bunch of changes into stable branches === zyga-afk is now known as zyga [14:33] also, that powerbar energy gel thing i just had is making me tire === salgado is now known as salgado-afk [14:41] Hi Folks [14:47] hi mmcc [14:49] hi dobey. just caught up on the logs [14:50] mandel, for your set_application_name fix, it shouldn't be a TODO. It should just be a no-op, since the app name is set in the info.plist on macos [14:50] mmcc, sure, I can remove the comment [14:51] mandel: cool. also, send reviews my way today [14:52] ooh [14:52] * dobey suddenly remebers how cool bzr log -r -1 -n0 is [14:53] mmcc, sweet, hereis a list with all the reviews I need in order: http://paste.ubuntu.com/1048968/ [14:53] mmcc, some might have already the +1 but it should not be the case [14:54] ok mandel [14:54] btw mandel, you said controlpanel doesn't start because it doesn't find the binaries to execute, but I'm not seeing that problem (or maybe I am and I don't know it..) - can you explain more? which binaries? On my system, it finds sso-login correctly, but freezes later. However I didn't see any evidence of trying to find another binary - ie, no backtrace [14:57] mmcc, TODO removed :) [14:58] mmcc, when I look at the logs I'm getting that it cannot find the file to execute the sso. I'm not surprised since I have sso in a diff location etc.. so the problem I have is just that the code is not packaged and when it tries to run the backend/sso things go wrong [14:59] mmcc, is expected when I just executing things from source without setting the paths etc.. [15:00] meh [15:00] me [15:00] me [15:00] ah ok. yeah, when I'm testing I'm setting PYTHONPATH on the commandline. I have a branch pending to get the right paths when packaged... [15:00] me [15:00] mmcc, exactly, so is a no issue in the current state [15:00] me [15:00] alecu, thisfred, gatox? [15:00] me [15:00] me [15:01] me [15:01] λ DONE: tarmac tweakery, uploads, bug #1009786 (u1sync), bug #1014772, got MRE approval [15:01] Launchpad bug 1009786 in u1sync "Does not connect: exceptions.TypeError: get_ssl_context() takes exactly 2 arguments (1 given)" [Undecided,Fix released] https://launchpad.net/bugs/1009786 [15:01] λ TODO: finish 3.99.0 triage/releases/uploads [15:01] Launchpad bug 1014772 in ubuntu-sso-client (Ubuntu Quantal) "Tests fail with GTK+ 3.5" [Undecided,Triaged] https://launchpad.net/bugs/1014772 [15:01] mandel [15:01] λ BLCK: None. [15:01] DONE: fixes for bug 1014752 bug 1015053. Got syncdaemon running on mac os x, synced 3,5 gb so far out of 15 gb. Tested cretion, deletion and rename of files. [15:01] TODO: Get the fsevents-branches landed. Stress tests the sd on mac os x. [15:01] BLOCKED: no [15:01] briancurtin, please [15:01] Launchpad bug 1014752 in Ubuntu One FsEvents daemon "An error occur when a path is added twice" [Medium,Triaged] https://launchpad.net/bugs/1014752 [15:01] Launchpad bug 1015053 in Ubuntu One FsEvents daemon "And error occures when a path that does not exist is removed" [Medium,Triaged] https://launchpad.net/bugs/1015053 [15:01] DONE: jenkins stuff, reviews, created installer for signing, but that's screwed up [15:01] TODO: do more testing on this installer to make sure it was created properly, then go back through the signing process [15:01] BLOCKED: no [15:01] NEXT: mmcc [15:02] DONE: poking at controlpanel UI all day [15:02] TODO: review mandel's 6 branches, other reviews, awaken controlpanel [15:02] BLOCK: need help with controlpanel UI freeze [15:02] NEXT: ralsina [15:02] DONE: reviews, mgmt call, misc stuff TODO: get a Q VM up because u1cp iscompletely broken there, help in u1db build system BLOCKED: no NEXT gatox [15:02] DONE: [15:02] Reviews, refactoring, mumble with alecu and mandle [15:02] TODO: [15:02] Refactor windows and darwin implementation. [15:02] BLOCKED: [15:02] No [15:02] alecu, go [15:02] DONE: leveled up to undead-dog cough, took a sick day, mumbled with mandel and gatox about reviews and refactorings [15:02] TODO: do lots of reviews for the mac team [15:02] BLOCKED: no [15:02] NEXT: thisfred [15:02] DONE: bug #1006872 TODO: bug #1006872 BLOCKED: no NEXT: NULL [15:02] Launchpad bug 1006872 in U1DB "sync_exchange should transmit the txid it thinks the target was at" [High,In progress] https://launchpad.net/bugs/1006872 [15:04] comments? [15:05] n/c [15:05] ac/dc [15:05] i can hear the sine wave resonating in the cyrstals of one of my LCDs [15:06] (newtons/coulomb) [15:06] ok, EOM [15:06] comment: I should be busy with reviews today, but if anyone has tips on debugging hung PyQt UI, I'd appreciate it for after [15:06] which is weird, since it should be DC at that point [15:07] so probably the power supply is slowly going the way of kaboom [15:07] mmcc: if you point me at the branch I can try running it. [15:07] ie, debug Env vars or stuff... [15:07] alecu, ok - let me check that I have everything relevant pushed and I'll ping you in a sec [15:07] mmcc, let me fix a branch I have left, deal with email and I'll poke you to see if I can give a hand [15:08] mmcc: btw: is this only inside the py2appd control panel, or it happens when running from trunk too? [15:08] ok, i need to get lunch and deal with a couple things, before i end up taking a nap on the futon [15:08] thx mandel. I'll paste the info and link it here for both you guys [15:08] bbiab [15:08] alecu, only testing from source for now, I haven't tried it from the .app [15:09] "+TODO: nap" [15:09] sounds like a plan :-) [15:13] * gatox lunch === gatox is now known as gatox_lunch [15:14] alecu: you and me both: My coffee pot broke, so I'm going through serious withdrawal [15:22] thisfred: I recommend always having three or four backup coffee pots. Two french presses, and a stovetop espresso maker, at a minimum. Engineer some redundancy! [15:23] mmcc: you are of course correct (this was a french press, actually, which I really love). I suggested this morning that it was no reason for panic, and my wife damn near killed me. I'm starting to come around to her point of view. [15:24] next day shipping a replacement part from amazon [15:24] I think I'll walk to a coffee place over lunch to get my fix [15:29] all with a mac machine, I have sent you a mail on how to get sd runnig on it for you to tests the current status and crush my happiness, let me know if it works :) [15:31] mandel: you left out the explanation about how to get the whole environment setup :-) [15:31] ralsina, meh, I have high expectations with you :) [15:32] mandel: but you did not mail *me* you mailed ubunet-discuss! [15:32] ralsina, as cc, those are second class citizens.. [15:32] * mandel writes a second email pointing to the google doc [15:38] alecu, mandel, *, here is a paste about the UI issues I'm having with control-panel: http://paste.ubuntu.com/1049301/ [15:38] * mandel looks [15:47] hi [15:48] rye: are there any news about my double account problem? === salgado-afk is now known as salgado [15:51] Moscherkobold: hi, matiasb is working on mergint the accounts now and this is still in progress [15:52] ok thank you [16:14] dobey, I fixed the lint issues for https://code.launchpad.net/~mandel/ubuntu-sso-client/fix-tcp-activation/+merge/109367 can it be set to approved? === zyga is now known as zyga-afk [16:21] mmcc: http://stackoverflow.com/a/1745965/1098199 [16:22] alecu: handy, thanks! === gatox_lunch is now known as gatox [16:26] mandel, your python test suite for fsevents-daemon still doesn't run for me without tweaking... [16:26] oh man, we really need to fix this buildout install [16:26] mandel, which branch? [16:26] s/install/situation on macos/ [16:26] herb, soy tonto [16:26] mmcc, which branch? [16:26] mmcc, and yes, we have to fix that [16:26] mandel, mandel, the reduce-json-tools branch [16:27] yeah, i have two different dev-tools branches in parts/ so now when I run env-mac, it defines $u1trial to be BOTH of them [16:27] >_< [16:27] mmcc, yes, sorry I added that in the following one.. but yes, we ought to fix that asap [16:28] mandel: i set it so [16:29] dobey, superb, thx I also updated https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/fsevent-is-directory/+merge/109826 following your comment, is not a constant in a header that I know so I created one [16:30] mandel, ok I see the change in the next branch, but I'm still wondering how you set your environment up. what's your PYTHONPATH when you run those tests? [16:30] mmcc, I set everything manually 'cause I hate the way is done atm.. so I'm not a good citizen.. [16:31] * mmcc is not happy about how fast I'm getting at typing 'PYTHONPATH' [16:31] mandel: where did you get those hex values from originally? [16:31] mandel: ok, well I'm inclined to approve the branch, since I can make the tests work and they do pass. but we need to fix the buildout env setup [16:32] mandel, ping me when you and dobey are done - I have a question and don't want to weave two conversations together [16:32] mmcc: you can make a script called PYTHONPATH, that runs "PYTHONPATH=blahblahblah $@" then you can just tab-complete it! :) [16:33] urgh [16:33] haha [16:35] dobey, from here: http://osxbook.com/software/fslogger/ [16:36] i wonder where he got them from then :) [16:39] mandel: I approved your json-size branch with a note about post-ARC refcounting changes we might need to make. Maybe you've already addressed them in later branches [16:40] mmcc, I have not looked at remove arc just yet, I was focused on getting the thing in u1-client [16:40] dobey, no idea.. [16:40] ok, sounds good [16:42] mmcc, I'll be away for pycon in a few weeks and then holidays so I want you to have the thing working from point to point and arc is not that necessary right now [16:43] i need to take some holidays [16:44] mandel: that's fine. I will be getting a laptop with 10.7 on it shortly anyway [16:47] mmcc, ok, nevertheless, I don't want to go and, as spaniards say, dejarte en bragas [16:48] "dejarte en bragas" -> google translate -> "leave in panties" [16:48] I TOTALLY UNDERSTAND [16:51] hehehe [17:03] mandel: ok I read through the add-integration-tests branch and it looks good, but I can't actually run the tests, because I can't build the daemon. can we get someone else to run them? I'm guessing dobey couldn't, and he's the other reviewer on there now [17:03] mmcc, yes, add a +1 but do not approve and gatox can run them :) [17:03] i can't run the tests, no [17:04] ok [17:05] mandel, i'm going to keep reviewing your branches now.... and keep refactoring later [17:05] ok, marked and requested gatox [17:05] hi gatox [17:05] :) [17:05] mmcc, hi [17:06] mandel, btw did you get a chance to run control-panel as I pasted earlier? notice anything? [17:07] mmcc, I'm doing it right now, just finished with the old branches.. [17:07] great, thanks mandel [17:07] mandel: I agree with dobey: what's with the "0xfffff000" for the type? [17:08] mandel: it makes no sense that the last nibble of the va_type is also the first nibble of the va_mode. [17:09] mmcc: I'll run the tests [17:09] alecu, the kea->data.mode is the combination of the type and the mode [17:09] mandel: right. But the code is using "and" to split them. [17:09] alecu, why... no idea, I gathered that form the fslogger code which has no doc what so ever.. [17:10] mandel: and it makes no sense the way it is now, because it's not exactly splitting them. [17:10] mandel: can you point me at that line in fslogger? [17:11] mandel: don't worry, I've found it. [17:11] alecu, http://osxbook.com/software/fslogger/download/fslogger.c [17:12] mandel, run-integration-tests need to run it with sudo or something? [17:12] gatox, sudo, if you don't it will tell you to do so :) [17:12] mandel, ack [17:13] mmcc, I get the following paste.ubuntu.com/1049486 [17:14] mandel, http://paste.ubuntu.com/1049488/ [17:14] mandel, that looks familiar. do you also see the UI with the frozen overlay? [17:15] mmcc, yes, gets there does nothing [17:15] yep that's what I see. glad it's not just me [17:15] btw also mandel, quick typo needs-fixing on https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/fsevent-is-directory/+merge/109826 [17:15] mmcc, ok, I'm EOD I'll fix tom morning [17:16] mandel, before you leave [17:16] gatox, u1trial does have a --reactor option 100% sure, you env is funny [17:16] mandel, cra [17:16] crap [17:16] gatox, if you have the latests version you can remove the --reactor because the default in os x is twisted [17:17] gatox, but yes, your env is funky [17:17] mandel, i'll try [17:18] mmcc, I'm EOD, I'll look into control panel tom once I have fixed some little bugs in the u1-client/fsevents integration (not major, nedd to ignore -spotligh files) [17:18] mmcc, thisfred: sanity check for https://code.launchpad.net/~dobey/ubuntuone-client/update-4-0/+merge/111063 ? [17:19] mandel: ok, sounds great. consider controlpanel low priority, [17:19] dobey, will do. [17:20] mmcc, ack [17:20] all, I'm off, laters! [17:22] mandel, bye [17:23] dobey: +1 [17:23] gatox, send me an email if I need to review your branches tom morning, ok? [17:23] thanks thisfred [17:24] mandel, ack [17:24] alecu, I unzipped the linux source code (kernel) in the u1 folder, nothing has happened yet because is processing the music lib I added (5 gb out of 15 gb) I'll let you know how it went tom [17:25] mandel: great. [17:25] mandel: I'm interested in the memory consumption on the python side, and if "DROPPED_EVENTS" shows up. [17:27] alecu, ok, I can give you a num of sd at the moment: Ream mem size 26.5 MB, Virtual Mem 2.42 GB (normal with all those files??) Shared Meme 216 kb Private Mem: 24.6 MB Virtual Private Mem: 111.7 MB [17:28] alecu, cpu % is 7% ~ 9% [17:28] mandel: re: "Virtual Mem", check the size of the tritcask files [17:29] mandel: (regarding the 2gig size) [17:31] alecu, not a single file from the tritcask folder is bigger than 9.2 MB [17:33] alecu, the fsevents daemon virtual meme is also 2.4 GB so the domain.. [17:33] alecu, lets look closer tom [17:33] mandel: sure. [17:33] alecu, the systems seems stable but certainly those numbers do not look good [17:34] mandel: it's awesome that's stable, now we can work on making it smaller. [17:36] alecu, yes, is not the worst to have an alpha version that is memory hungry [17:36] mandel: btw: I seem to be missing ocmock... [17:36] mandel, i can't run your testes..... i remove the --reactor thing but there is another problem now..... i'll try to debug it or submit the need fixing in the MP [17:37] mandel: should we install that from the buildout scripts? [17:37] alecu, sorry, yes, the url is given in the description as well as how to add it to the path [17:37] alecu, I don't know.. but I'd say is too much work for what it brings to add it to the buildout [17:37] mandel: probably not on this branch :-) [17:38] mandel: I mean, it's not on the description of this branch. [17:38] alecu, oh, let me find the one that gave it :) [17:39] alecu, here you have it: https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/add-dispatcher-tests/+merge/107766 [17:39] mandel: thanks. Let's add that bit at least to some readme or wikipage. Tom, of cource. [17:39] *couse. [17:40] alecu, yes, will do [17:40] alecu, added to that branch description too [17:40] sigh...new binaries uploaded for signing, my hope that we could skip that step turned out to be wrong [17:40] *course, dam it. (said the beaver) [17:40] alecu, the 11.04 mini specs share from cparrino are a good tests to see if we support : on mac :) [17:41] mandel: awesome. Do we? [17:41] gatox, mmcc, alecu, I should expect a file with the name 'Foundations 11: Faster Uploads' fail, right? [17:42] alecu, I don't think we do atm, will be a pita since we use the linux methods which do not have the translate decorators.. [17:42] but lets not support that just yet [17:42] mandel, fundations file? [17:43] gatox, yes, look at the evil ':' in the name :) [17:43] evil evil.. [17:43] ahhhhhhhhhhhhhh [17:43] gatox, I remember that share 'cuase it gave me nightmares in the windows port [17:43]