[08:16] Good morning all! :-D [08:22] morning! [09:59] is there an Ubuntu-One extension/Script/Plug-in for Thunar? [10:40] hi all. I was looking at bug 1063170 and I can't figure out where the "Purchased from Ubuntu One" string comes from in the plugin code. Could someone give me a hand? [10:40] Launchpad bug 1063170 in Ubuntu Translations "Not translatable string "Purchased from Ubuntu One" in Rhythmbox" [High,Triaged] https://launchpad.net/bugs/1063170 [10:47] dpm, hm.. let me take a look for you [10:53] thanks [11:05] dpm, AFAIK that is in that name because the folder name of the music that has been purchased from ubuntuone is named like that [11:06] good morning [11:06] dpm, so, ideally you will want us to make the folder path translatable [11:06] mandel, bummer, not translatable, then [11:06] gatox, buenas o/ [11:06] mandel, yes [11:06] mandel, hi [11:06] dpm, I guess it can be translated, yes, because is a constant and we should be able to name it ~/.ubuntuone/Paquito\ el \ chocolatero if we wanted [11:07] mandel, exactly, that was also the song I had just in mind [11:07] I was doing the dance too [11:07] dpm, problem is, if you use a diff folder name in one machine and then you move to a machine that is in eng it brakes [11:08] dpm, as in, first time the udf is created using the spanish name, then you go to a machine in eng and then you have the spanish name synced and shows in spanish [11:08] dpm, so it does not work very well, I think the less damaging thing we could do is remove adj, that is, remove 'purchased' and leave it with the name which is 'international' [11:09] dpm, do I make sense? [11:09] mandel, gotcha. What about leaving the folder name as it is and make what is shown on Rhythmbox translatable? [11:09] dpm, that means patching rhythmbox because is part of the code using by it to monitor folders [11:10] hm, not really a reusable plugin architecture, then [11:10] ok, let me add this info to the bug for now, thanks for the help! [11:10] dpm, no problem :) [11:11] dpm, do explain that changing the folder name is a no no due to sync between machines with diff names [11:11] dpm, I wonder what would happen if we use a symlink which can be translated.. [11:12] dpm, udf folder is always the same, we do a ln -s to a translated version [11:12] I don't know, I wonder too if translatable XDG directory names could be used [11:12] dpm, you can try looking at how to add diff sources: http://library.gnome.org/users/rhythmbox/stable/customize.html.en [11:13] dpm, if we can add a level of indirection between the udf and the player (symlink for example) we are fine [11:13] dpm, as long as we don't change the actual udf name we are fine [11:16] brb.... machine die [12:13] hello, all! [12:14] alecu, hi [12:15] alecu, mandel any of you want to start the morning with a review? (not much morning for you mandel) https://code.launchpad.net/~diegosarmentero/ubuntuone-client/ubuntuone-client-publishapi/+merge/128312 [12:15] gatox, I can start the afternoon :) [12:16] mandel, sounds good :P [12:17] good morning! [12:17] ralsina, hi [12:18] hola gatox [12:18] gatox, looking at your branch, I think you might abusing lambdas a little to much, I mean, you are creating a new lambda each time you add a new handler when you could simply appending to the collection, right? [12:21] gatox, is the lambda really needed or could be have a class private method that does it? [12:26] mandel, if you use a private method, you will need a lambda anyway, because you will need to do something as: lambda *args, **kw: _method(signal_name) [12:26] mandel, if you use a private method, you will need a lambda anyway, because you will need to do something as: lambda *args, **kw: _method(signal_name, *args, **kw) === muffinresearch_ is now known as muffinresearch [12:31] gatox, lets look at it step by step: first improvement I see => http://paste.ubuntu.com/1269167/ [12:31] gatox, there is no need to create a new lambda each time you add a handler, right? [12:32] mandel, we will find out when we run the tests..... i would say no..... but i'm not sure if maybe that binds to the value of the list at that moment [12:34] mandel, ok........... it was just a precaution..... this works tooooo [12:35] gatox, next improvement => http://paste.ubuntu.com/1269176/ [12:36] gatox, that will at least allow us to test that all handlers are called, improves testability, which I think is an improvement.. [12:36] mandel, we can test that already [12:38] mandel, actually, we are doing it in the tests [12:42] mandel, i can change that if it's necessary..... maybe we can ask someone else, because i actually prefer the lambda solution [12:43] gatox, do think about memory etc.. creating a new lambda object for every new handler that has to be gc is an overhead that is not really needed [12:43] * ralsina looks around [12:43] mandel, gatox: you guys asked for a 3rd opinion? [12:44] mandel, mmmmm.... i see [12:44] mandel, now i'm more convinced [12:44] ralsina, i think that is not necessary anymore [12:44] ralsina, mandel has presented a new argument that is more valid to me [12:44] * ralsina points out that 500 lambdas are less than 20KB or RAM and take a microsecond to gc [12:44] * ralsina is great at making up numbers [12:45] I would prefer "pretty, readable code" but I leave it to you gatox [12:45] ralsina, i'll change it to the private function [12:46] gatox: cool [12:46] i only needed a good reason :P [12:47] hehe [12:50] gatox, ralsina, just so that I can add a new diff way to do it, you can use something like this: http://paste.ubuntu.com/1269200/ [12:50] gatox, ralsina, add a new handler list if not present, and append the callbacks, then the list can be called with the *args and **kwargs and will do the right thing.. [12:50] but might be too evil :) [12:51] gatox, ralsina, you reduce the need of a lambda, you use the same amount of memory as a single list and you simple add a new type definition [12:52] but I think I have been doing too much c/c++/golan recently and I have started thinking about memory too much :P [12:52] mandel: that's a list of lambdas disguised as a class [12:52] mandel: disguised as a functiokn [12:52] ralsina, no, is not, in the other you a list with with callbacks and per callback that you added to the list you were creating a new lambda [12:53] ralsina, so, in the second handler added, you had the old lambda (first handle) the 2 handlers (2 lambdas) then create the new lambda (2 handler lambda) and hope that the first lambda was gc at some point [12:53] mandel: I really really hate "for cb in self" :-) [12:53] ralsina, in this case you have a list with the callbacks and there is no need for the generic lambda or the new creation of a lambda [12:54] ralsina, as I said, a little evil :) [12:54] mandel: I would just keep a list of functions and iterate. Explicit. [12:54] but anyway, it's not worth discussing this much, I think. Clear code is best. [12:55] ralsina, yes, I was just having fun by stretching the subclassing model and wondering how little code I could write to do the job and use the less amount of objects hehehe [12:55] ralsina, lets call it brainexercise [12:55] ok, lunch time here :) [12:55] * gatox is not convinced again..... [12:55] haha [12:55] gatox, I'll do a second review when i'm back :) [12:56] gatox: try to write it as dumb as possible [12:56] gatox: remember I may have to read it later! [12:56] ralsina, for mandel? [12:56] jejjee [12:56] write for the common denominator :) [12:56] i'll do it with a private method....... [12:56] * gatox sacrifice his love for list comprehensions [12:57] gatox: if it needs a \ it's a bad list comprehension :-) [12:57] ralsina, it didn't need that....... because i was splitting the line inn the right places jejejeje [12:57] gatox: oh, but then it had a \n ;-) [12:58] :P [12:58] * ralsina goes track when we started having a double spinner [12:58] mandel, branch updated [12:59] ralsina, do you have any special task for me?? or i should try to start killing u1-cp bugs?.... i might found a weird bug in u1-client too [12:59] might not....... i found [13:00] gatox: yes, check https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0AiE68jdtyfBcdG1RTkpEYl8yNV9yaEtIRkdUbF9OZGc#gid=0 and make it work :-) [13:00] but i don't know if it is a priority [13:00] ralsina, ack..... really clear list of tasks [13:05] ralsina, mmmmm..... some weird stuff has been reported there..... i'll need to track each person and ask [13:05] gatox: I think I fixed one of those at least [13:05] gatox: but not on stable-4 :-( [13:06] ralsina, also.... maybe something is not clear..... because it seems they expect the things being transfered from the server to the client to appear there too [13:07] gatox: then you need to clarify the leftmost column :-) === m_conley_away is now known as m_conley [14:13] gatox, got a small needs fixing [14:13] mandel, ack [14:17] mandel, fixed [14:17] mandel, it was there for when i was testing stuff [14:23] gatox: this bug is ours, right? https://bugs.launchpad.net/ubuntu/+source/indicator-sync/+bug/1063950 [14:23] Ubuntu bug 1063950 in indicator-sync (Ubuntu) "Menu item "Get more space" opens "one.u.c/services/#storage_panel" instead of "one.u.c/services/add-storage/"" [Undecided,New] [14:23] * gatox checking... [14:23] gatox: as in, that's really a bug in u1-client not in the sync menu [14:23] ralsina, i'm fixing that right now [14:23] gatox: awesome, change the affects in the bug [14:23] ralsina, ack [14:31] gatox, do you think that adding a test where you add a handler to one signal and an other handler to an other signal and assert the correct handler is called is a good idea? [14:32] mandel, will do in a few miins [14:32] great [14:34] in the meanwhile..... mandel ralsina if you can: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/add-storage/+merge/128729 [14:37] gatox, sure [14:44] mandel, new test added: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/ubuntuone-client-publishapi/+merge/128312 [14:44] ralsina, gatox, alecu the ISP eng is here to install fiber in my place, I'll me off line a little until he is done [14:46] mandel: ack [14:46] mandel, ack [14:50] alecu and I are on a call we'll miss standup, you guys go ahead without us [14:50] mandel: fiber == awesome! [14:52] mandel: you getting whole bran? [14:52] mandel: if that doesn't work, try activia! [14:52] good morning mmcc! [14:52] hi ralsina [14:53] mmcc, hi [14:53] mmcc, i have a present for you: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/ubuntuone-client-publishapi/+merge/128312 [14:53] :P [14:54] it's a fine day for reviews [14:54] mmcc: it's QA day! [14:55] ralsina: tell me more [14:55] mmcc: today QA tests our mac build to see if we do the beta release on the 11th [14:55] mmcc: so... let's do a build and give it to QA :-) [14:57] elopio: it says "windows testing" on the schedule, but it should be mac testing [14:58] ralsina: you said you wanted a windows testing session this week. [14:58] mmcc, first review my branches that fix some of the problems we were having :P [14:58] don't you want it anymore? :) [14:58] mmcc, so you can include them in the build \o/ [14:59] elopio: yes, that too :-) [14:59] elopio: I am confusing sometiimes, I know, sorry [14:59] ralsina: aha, ok. are they expecting anything to have changed since the last build? gatox' change shouldn't affect the final behavior… [14:59] because the last qa build had a workaround [14:59] mmcc: I don't know what the last build was, really [14:59] ralsina: so, are we having a windows installer too? [14:59] mmcc: I would like to get a fix I did yesterday [14:59] elopio: there are daily windows builds in jenkins again now [15:00] ralsina: aha, ok. great [15:00] me [15:00] elopio: but I am not sure it even does something on mac [15:00] ralsina, elopio: jenkins is down again... [15:00] briancurtin: oops [15:00] will take a look shortly [15:00] briancurtin: jenkins is down or builds are failing? [15:01] ralsina: the windows jenkins machine isn't connected. everything is N/A [15:01] that one about "I am not sure..." was for mmcc sorry [15:01] briancurtin: ok, it's tuesday, which is "break jenkins" day [15:01] me [15:01] me [15:03] need to go downstairs for a sec..... so here is my standup: [15:03] DONE: [15:03] Propose a branch to start fixing some u1-client sync menu related stuff, improve some other branches already proposed. [15:03] TODO: [15:03] Keep fixing u1-client sync menu and u1-cp bugs. [15:03] BLOCKED: [15:03] No [15:03] me [15:04] DONE: processed email/colonial imperialist day TODO: playlists BLOCKED: no NEXT: gatox [15:04] DONE: u1c and u1cp tests running fine (well, some failures on client, but they run to completion) [15:04] TODO: CredentialsError still coming from SSO, working out how its finding the binaries. i found one place that is wrong but then it seemed to try again with the right commdand line and failed. [15:04] TODO (on the side): i noticed we're sending post2008server in a URL post while logging in to SSO (can't remember the exact string, will dig it up). this is from the stdlib platform module not identifying Windows 8. i dont know what we actually use that for on the server side, probably stats, but it's an easy change in the stdlib. i did the change for win7/2008server [15:04] NEXT: mmcc [15:04] DONE: long weekend, recent transfers menu [15:04] TODO: reviews, set up .DMG, qa build [15:04] BLOCK: no [15:04] NEXT: [15:04] ? [15:06] who are we missing, ralsina mandel alecu? or are more people off today [15:06] * mmcc needs to check the logs [15:06] mmcc: both ralsina and I are on a hangout [15:07] alecu: ack [15:08] * ralsina declares this "day columbus noticed this was not japan" day. [15:09] Rerouting... [15:23] brb, need to let the heating/furnace dude in. it's cold in here. [15:24] gatox, quick needsfixing on that client branch [15:25] gatox: I'm also still not 100% clear why we need that change. when were multiple functions being connected to the same signal? [15:26] mmcc, i already explain that and show you some code from u1-client on friday..... and you answered that you understand [15:26] mmcc, also.... i explain it on the comment i added [15:27] gatox: sorry if I'm just forgetting - I'll look in the log on friday. [15:27] mmcc, or read the comment [15:29] gatox: ok, I had to read the comment again, sorry - can you point me to the place in syncdaemon core that was also connecting to the signal? not urgent though, I believe you - this is just for my understanding [15:30] mmcc, could you check your logs for that?? because i don't remember and i'll need to debug to find it..... i already send that piece of code to you on friday..... if you don't have it let me know..... and i'll check [15:30] ok, I'll look gatox. don't worry about it [15:32] mmcc, thx........ because it was kind of tricky to find it [15:33] gatox: yeah, I definitely know how that is [15:34] that's a good argument for keeping notes on this stuff - whenever I figure out one of these long IPC chains, I write down what calls what - it's already helped a couple times when I had to go back and check something [15:34] hello Hello! [15:35] mmcc, docstring updated [15:40] great gatox, thanks - now to try testing it on macos [15:55] goodbye Hello [15:55] oh I mean HELLO [16:01] * gatox lunch === gatox is now known as gatox_lunch [16:03] ? [16:07] apparently mandel's ISP guy is still trying to connect the optic fiber using duct tape or something, will take a while longer to show up [16:08] ha. fiber to the house, duct tape the rest of the way [16:08] briancurtin: the last mile is a sticky problem [16:08] lol [16:28] hrmm. so this branch: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/ubuntuone-client-publishapi/+merge/128312 with control-panel *trunk* fixes the share links tab for me IRL on osx [16:29] * mmcc waves [16:31] mmcc, but you will need this cp branch too: https://code.launchpad.net/~diegosarmentero/ubuntuone-control-panel/u1-cp-publishapi/+merge/128316 or you will only be able to share a file once [16:32] gatox_lunch: aha! thanks1 [16:32] ! [16:32] * mmcc was getting very confused [16:45] ralsina: how do I check that your change from yesterday works on os x? [16:46] mmcc: hmmm... [16:46] mmcc: run ubuntuone-control-panel-qt --switch-to=devices? [16:47] ralsina: with or without credentials? [16:47] mmcc: with [16:48] ok, works for me [16:48] mmcc: cool [16:49] mmcc: OTOH, if you have a "share a file" link in the menu icon's popup, does it open in the right tab? [16:49] mmcc: if it doesn't that's a separate bug [16:50] ralsina: share a file link in the menu icon? I don't know what that is [16:50] mmcc: in the icon in the mac global menu, do you have a "share a file" item? [16:51] ralsina: no, and looking, that's a # TODO in systray.py [16:52] mmcc: ack [16:53] gatox_lunch: row 8 in the sync indicator testing sheet needs to clarify that's not implemented and a bug link [16:53] ralsina, ack [16:54] ralsina, mmcc that was because de sync menu was implemented before de shares tab [16:54] s/de/the [16:55] ralsina, gatox_lunch: did you see this bug from last friday: https://bugs.launchpad.net/ubuntuone-client/+bug/1062729 [16:55] Ubuntu bug 1062729 in Ubuntu One Client "sync menu uses wrong upload size for percentages" [Undecided,Confirmed] [16:56] wondering if that's on the testing sheet - you probably won't notice unless you have lots of highly-compressible large files :) [16:56] mmcc: saw it, did nothing about it [16:56] mmcc: shhhh ;-) [16:56] I noticed it when uploading a ~200MB text .sql file [16:56] mmcc: yeah, those files are uncommon [16:56] plus, it will just show "slow slow slow done!" [16:56] :-) [16:57] heh :) ok, like I said, it's mostly a minor issue but it made debugging other stuff very confusing :) [16:57] yeah, it was showing "9%" forever and I started looking for string truncation or something :) [16:59] gatox_lunch: so, after that one is marked as unimplemented we are down to ~2 issues with sync menu. Not horrible. === gatox_lunch is now known as gatox [17:00] ralsina, yap [17:05] gatox: +1 on https://code.launchpad.net/~diegosarmentero/ubuntuone-client/ubuntuone-client-publishapi/+merge/128312 -- you still need to get mandel to approve that though [17:07] mmcc, yes..... he said he will review it when he is back [17:07] gatox cool. also +1 on the other branch [17:07] mmcc, cool.... thx [17:07] thanks for being patient with me while I figured out what was going on there [17:08] mmcc, no problem :D [17:09] gatox: it would've been easier if your comment here during lunch about "you will only be able to share a file once" was in the MP description under "HOW TO TEST" or something :) I can look at the code and understand what it's doing, but sometimes it's hard to understand what it's fixing at the top level [17:09] ok, done bugging gatox for the day :) [17:10] mmcc, jeje ok.... will add that next time.... promise :P [17:23] ralsina, so....... i've marked the share a file option in the menu as not implemented, and added the bug link..... now we have 2 reds..... and i already propose a branch for one..... and the remaining one, is the crazy thing that is happening with the shares that i mentioned [17:24] ralsina, so i'm going to start working on that one if you agree..... probably i'll need to contact people from the server team, because i'm not sure if they are sending that data now [17:25] ralsina, here is the branch for the other bug..... mandel already approve it..... is one line change: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/add-storage/+merge/128729 === eu is now known as Guest69817 [17:38] gatox: agreed [17:39] gatox: looking at the branch [17:39] gatox: global +1 [17:39] ralsina, thx [17:42] gatox: well, there's also the "sync menu doesn't show the cloud icon" but that's not us [17:43] ralsina, mmmmm i think that maybe i can prioritize this one: "Status in indicator is not updated when the status is changed from U1 control panel" [17:43] and then the new shares one [17:43] gatox: happy either way [17:43] ralsina, ok [17:44] I'm back!!! [17:45] ralsina, sorry, the guy from movistar/telefonica had some serious problem getting the connection [17:45] mandel: ack [17:45] mandel: np, it happens [17:46] ralsina, at least now I have fiber :) [17:46] which makes me very very happy [17:51] mandel: yes, slow transit is bad [17:51] ;) [17:53] ralsina, in nearly everything :) [17:54] gatox, what is the bug # for that is that indicator not updated bug? [17:55] mmcc, sorry, what? [17:55] mandel, please: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/ubuntuone-client-publishapi/+merge/128312 [17:55] gatox - this one from ten minutes ago: "ralsina, mmmmm i think that maybe i can prioritize this one: "Status in indicator is not updated when the status is changed from U1 control panel" " [17:56] gatox, sorted [17:56] mmcc, that one is only for the sync menu in ubuntu [17:56] mandel, ack, thx [17:56] gatox: ok, cool [18:10] yay, nice looking .dmg with background image and everything, all set up [18:11] mmcc, cool [18:11] ralsina: for sending to QA, is that just ASAP or when? [18:12] mmcc: asap please [18:12] ralsina: okboss [18:18] gatox: is https://code.launchpad.net/~diegosarmentero/ubuntuone-client/ubuntuone-client-publishapi/+merge/128312 waiting on anything? mandel approved but didn't flip the switch [18:18] mmcc, I did not 'cause I though we agreed to the owner to do it.. or am I outdated? [18:18] mmcc, i'll do it....... also..... this branch includes all the changes..... and there was the other one with the workaround that didn't landed..... and i think i can delete it [18:19] mandel, you're not outdated - I was poking the owner :) [18:19] :) [18:19] gatox: you're talking about the control-panel branch where you used the return value instead of the signal right? [18:20] was there a workaround branch for u1-client too? I remember one that changed a comma that did land, but it was just a comment change once it landed [18:21] mmcc, ahhh no....... i was confuse..... just for u1-cp [18:23] ok cool [18:27] mmcc: will the current and recent transfers menu item work in the public beta? [18:27] chaselivingston: no [18:28] in fact I had a note to myself, maybe we should just not show it. ralsina? [18:28] mmcc: ok, good to know, that might not be a bad idea [18:28] if it's broken, yes, disable it [18:28] we can re-enable it a week or two later [18:28] ralsina: ack. will do [18:29] ralsina, do you know why is broken? it has anything to do with the changes that you talk about in u1-client files progress? [18:29] ralsina: what's the best way to do that, just have a side branch that comments it out and build with that? [18:29] gatox: I know why it's broken, I was working on it on friday [18:29] mmcc, can you tell me why? [18:29] I don't know why it doesn't work on linux, but on osx, the timer is not called to populate the menu [18:29] mmmmm [18:30] mmcc: yes [18:30] I didn't know it was broken :-/ [18:31] gatox, in the TransfersMenu __init__ it only starts the timer if sys.platform isn't 'win32' or 'darwin', and I'm not sure why… but I have some code to fix that and try to get live-updating working on osx too, so it doesn't matter too much [18:32] mmcc, ack [18:36] sorry guys, but I have to call it a day, is close to dinner time and have to walk the dog, I'll work a couple of extra hours tom [18:38] mandel, bye === slank` is now known as slank [19:28] Does ubuntu one lock files? I misread and thought it supported webdav. [19:28] Ipsilon: no, to both [19:29] Ok, thanks. That's too bad, would have been perfect for zotero. [19:32] Ipsilon: if the app does locking through a lock file, the lock file itself will be synced but that's not reliable at all [19:32] Ipsilon: I don't know of any sync solution that reliably locks files [19:33] Well, webdav works. I believe through locking files...although maybe it's through a different method. [19:34] Works for zotero, that is. [19:46] ok, finally have current build uploading for QA. ralsina who do I ping when it's there? [19:51] OK, new build for QA here: http://ubuntuone.com/1Ny7EqmY50jLxRaufZLnrX [19:52] I guess ping elopio --^ [19:55] mmcc: elopio [19:55] mmcc: and jfunk [19:55] mmcc: and joshuahoover [19:56] chaselivingston: ^^ [19:56] saw all this, thanks, working on installing it now [19:57] chaselivingston: sorry I forgot about you :-) [19:57] ralsina: no problem ;) [20:02] ok..... i need to fix some tests and i'll have the sync menu and syncdaemon always in sync regarding the connection status [20:02] i'll propose a branch for this tomorrow morning [20:02] eod here! [20:02] bye people! [20:06] mmcc: neat choice on the black menubar icon lol [20:07] chaselivingston: is that a good lol or a bad lol? [20:07] mmcc: i think it's a good, we'll see after i use it for a while [20:08] mmcc: just can't wait to be able to get rid of the icon in the dock ;) [20:09] chaselivingston: understood. for now, how about just getting rid of the dock? I hide that sucker, makes me happier :) [20:09] mmcc: haha, i've done that in the past, but like having it shown most of the time [20:09] it takes all kinds I guess ;_ [20:09] er, ;P [20:10] lol [20:10] chaselivingston: we do have a way to do that, but we needed to fix the more critical stuff first [20:10] ralsina: sure, i understand. i know it'll happen eventually [20:24] trivial review for setup-mac anyone? [20:24] https://code.launchpad.net/~mikemc/ubuntuone-windows-installer/make-cp-fg/+merge/128811 [20:28] mmcc: i can't test, but my eyes say approve [20:30] thanks briancurtin. I assert that it works :) [20:31] nice work on the mac app guys, i think we're in a very good place for release on thursday [20:31] thanks chaselivingston , that's good to hear [20:31] mmcc: can't wait to get some user feedback on it [20:33] chaselivingston: yay! [20:33] a couple other really small packaging branches for review: https://code.launchpad.net/~mikemc/ubuntuone-windows-installer/bump-version/+merge/128813 and https://code.launchpad.net/~mikemc/ubuntuone-windows-installer/embiggen-log/+merge/128814 [20:34] ralsina: can you in particular check the version # one? [20:34] mmcc: sure [20:35] mmcc: +1 cromulent in embiggen [20:35] :D [20:36] mmcc: also +1 on the bump [20:36] we really need to automate that, we have like 5 version number-like things in -installer [20:36] ralsina: now for the trifecta, shall I globally approve the https://code.launchpad.net/~mikemc/ubuntuone-windows-installer/make-cp-fg/+merge/128811 [20:36] mmcc: looking... [20:37] mmcc: I have no idea what the flag does :-) [20:37] ralsina: yes definitely - also currently syncdaemon's version number is always 0.1 on mac/win [20:37] ralsina: hmm, let me go look and see what it does… "_ [20:38] it's is_background, and True means you have no menu bar or dock icon. False means you do [20:38] mmcc: we change that in the logs manually on windows in the last release [20:38] mmcc: we should do that for the real release on mac too [20:38] mmcc: for which process is this setting? [20:39] the last few builds have been built with is_background=False, giving us an icon. this is for controlpanel [20:39] this setting change is old, I just haven't gotten around to pushing the change [20:39] mmcc, ralsina: what is the process going to be for public testers as far as updates to the app go? [20:40] mmcc: ok, yes global +1 [20:40] chaselivingston: what do you mean exactly? [20:40] ralsina: great, thanks [20:40] mmcc: how will updates be pushed to users? [20:41] chaselivingston: we don't have an autoupdater for mac yet [20:41] chaselivingston: yeah, TODO [20:41] chaselivingston: we need to work on one. at least something that pops a "get the new version" message [20:41] ralsina: right. so will users have to come back to the download page to download updates? [20:41] chaselivingston: yeah, let's keep it simple [20:42] ralsina: ok, so for now they're going to have to either watch that page, or perhaps social media to know when to grab updates? [20:42] mmcc: maybe we should timebomb the beta :-) [20:42] chaselivingston: yes, social media && eyeballs [20:42] chaselivingston: the next one should have a simple updater [20:43] ralsina: ok, we probably need to make sure that's clear in the beginning, i'll need to make sure roberta knows about that [20:43] chaselivingston: yes, I'll tell her to add something in the communication [20:43] ralsina: ok cool [20:44] ralsina: i'll leave a note in the gdoc for her [20:44] chaselivingston: awesome [20:45] hrm. timebomb is appealing. what would be nice is if we had a web api call to ask if there's any info to show the user… like GET /message?clientversion=foo , which can optionally give us some HTML to show to the user, a place for "HEY YOU ARE USING AN OLD VERSION" or whatever [20:45] I've seen that pattern in other apps, looks handy [20:45] mmcc: we have that for windows [20:45] oh hey, great! [20:46] mmcc: but there it's done using a huge binary installbuildre creates and a xml file [20:46] oh hey, great? [20:46] mmcc: I think we can do a much better thing using 10 lines of code and a file somewhere that says 1234 (or correct number) [20:49] * briancurtin reboot [20:50] ralsina: probably, yeah. it can be useful to set it up to display a message downloaded from a known place too [20:50] good for release notes or explanations of why it's important to update === m_conley is now known as m_conley_away [21:57]