18:01:37 #startmeeting tor-browser 18:01:37 Meeting started Mon Aug 1 18:01:37 2016 UTC. The chair is GeKo. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:01:37 Useful Commands: #action #agreed #help #info #idea #link #topic. 18:01:43 hi everybody! 18:01:48 hi 18:01:56 we are probably not many today and i am basically not here as well :) 18:02:06 so we'll see how it goes 18:02:19 last week i've been on vacation 18:02:37 and started struggling with the release on the weekend today 18:03:15 i feel this struggle is continuing at least until tomorrow and then i hope i'll have another couple of days off 18:03:30 i plan to be back as usual monday next week 18:03:40 that's it for the status update from me 18:03:51 hi everyone 18:03:57 o/ 18:06:11 sorry, I'm still writing my update... 18:06:24 * boklm can go next 18:06:29 This past week I worked on #19410 and #19737, and helped build the new releases 18:06:37 This week I'm planning to publish the new releases, work on #16551 and on adding tests for #8725 18:06:50 That's it for me 18:09:37 GeKo: ugh, Mozilla tagging releases afterwords is bad form.. plus we stopped getting release emails.. I'm sorry I dropped that. I didn't expect them to do that - I assumed the release was actually delayed :/ 18:10:15 * arthuredelstein is ready to go now 18:10:16 yeah, and it wasn't :( the candidate build is from July 26 18:10:50 maybe they can add us to some email listbot about candidate builds and (hopefully) their commits? 18:11:10 might be a good idea. i'll ask them 18:12:13 This week I was mostly on vacation, but for the past couple days I worked on three patches: uplifting for the font whitelist patch (bugzil.la/1121643) and our system colors patch, and #19741. 18:12:38 The last ticket turned out to be really mysterious and I still haven't tracked it down. 18:13:24 but it is reproducible, right? 18:13:26 Also, while looking at the system colors patch I noticed that we're still leaking the user's system display settings on Windows. For example, the background color of a normally white web page 18:13:32 (i have not tested it yet) 18:13:42 GeKo: Yes, it's reproducible. Just doesn't make sense to me yet. 18:13:51 ok 18:14:35 I'll be posting a ticket about system colors on Windows. 18:14:50 I plan to continue working on those this week. 18:14:53 That's all for me. 18:15:31 okay, thanks. 18:15:54 who else is here? i know mcs/brade are away. 18:17:03 I can go next :) 18:17:13 hey! cool, sure :) 18:17:49 we have ported the Tor Browser updater patches to Tor Messenger and I am happy to report that the updater is working 18:18:06 wow, that's good news! congrats 18:18:07 cool! 18:18:21 the MAR generation is kinda manual but that's not a big concern for now 18:18:47 there is just one small issue and that is that the updater in Browser can't seem to find the dynamic libraries on OS X and Linux (Windows is fine) 18:19:07 I thought this would be covered by #4234, #13379 and #18900 (nsUpdateDriver.cpp) but I am clearly missing something here 18:19:39 for example, the error on Linux is: 18:19:41 /tmp/tor-messenger/Browser/updates/0/updater: error while loading shared libraries: libmozsqlite3.so: cannot open shared object file: No such file or directory 18:19:52 (almost similar for OS X) 18:20:18 both of these go away by setting LD_LIBRARY_PATH and DYLD_LIBRARY_PATH to . so I was wondering what is a proper way of fixing this? 18:21:13 (we can end the meeting for now and discuss this later) 18:21:55 i guess we'd need mcs/brade here. i don't know what's wrong without looking closer 18:22:23 ok thanks, I will email them 18:22:51 what is left before testing is signing the MAR files but I don't see any major blockers there 18:22:57 that's it from me and the TM team :0 18:22:58 er, :) 18:23:17 do you plan to bake your own signing keys into the tor messenger?= 18:23:39 GeKo: for signing the MAR files, yes 18:23:46 okay. 18:23:52 I thought it would be good to have redundancy. do you think that's a good idea? 18:24:31 i was mainly asking what to expect with respect to our signing machine 18:24:43 in future, ideally, we would also like to sign the binaries on Windows and OS X but I haven't thought much about that yet 18:24:50 whether you want to use that setup as well 18:25:10 well, as i said boklm and i can sign the windows binaries for you already 18:25:44 ok that's great. OS X as well? 18:25:47 and os x should be doable too 18:25:50 yes 18:25:53 nice, thanks 18:26:06 i am about to hook boklm here up as well 18:26:19 we should be ready shortly (i hope) 18:26:48 great. we plan to release in two months or so with the secure updater, so I think that should fit with our timeline as well 18:27:11 sounds good. let us know if there is anything you need 18:27:26 thanks 18:28:31 not sure if mikeperry wants to give an update 18:28:44 meanwhile i have one item to discuss 18:29:10 i hope the signing goes well tomorrow (there is still hope) 18:29:26 boklm: i wonder what to do about the alpha and the osx diff? 18:29:48 should we reapply libfaketime stuff and rebuild? 18:30:18 or should we say: we found a machine that got different results for osx due to timestamping things and we are invetsigating? 18:30:37 I am not sure it is due to timestamping, or something else 18:30:38 and ship the stuff you got? (and which i matched on the lxc machine)? 18:31:01 so I am not sure reapplying libfaketime stuff would fix it 18:31:34 i see. so, given the tight timeline it might be okay for the alpha that we point the issue out but go with the stuff we have 18:32:24 yes, I think we can open a ticket to investigate that, and point to that ticket 18:32:50 okay. 18:37:37 do we have anything else? 18:39:02 thanks everbody then *baf* 18:39:05 #endmeeting