19:00:55 <mikeperry> #startmeeting tbb-dev
19:00:55 <MeetBot> Meeting started Mon Mar 16 19:00:55 2015 UTC.  The chair is mikeperry. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:00:55 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
19:01:17 <mikeperry> ok, let's get started. I will kick things off
19:02:11 <mikeperry> Last week, I did mostly non-TBB things. I did some work to move Tor Labs forward, send some roadmap-related emails, and did some Android hacking. I did a little bit of ticket triage, and poking at #14688.
19:02:19 <mikeperry> This week, I will take a look at #15029, and either revert #10280 or try another patch. Since the original author (I think?) says the #10280 patch may be bad (without any explaination), we may just revert it. I will also continue working on #14688, will try to make a simple fix for #13650, and may try out a build with #14977.
19:02:39 <mikeperry> At the end of today's meeting, we should also discuss the tickets in https://trac.torproject.org/projects/tor/query?keywords=~tbb-4.5-alpha&status=!closed, and make sure everyone is OK with the assignments. There are some still assigned to tbb-team that we should decide how to handle.
19:02:54 <mikeperry> that's it for me
19:03:40 <mikeperry> who wants to go next?
19:06:38 * mcs can go
19:06:44 <mcs> This past week, Kathy and I completed a fix for #13548 (GeKo has indicated that he will review it).
19:06:51 <mcs> Other than miscellaneous bug triage, we spent most of the week on #12827.
19:06:57 <mcs> We thought we would be finished by now, but it has been more difficult than we expected to allow SVG in chrome documents but not content.
19:07:05 <mcs> (our current challenge is how to determine whether to allow an SVG image that is loaded due to a CSS background-image rule to be loaded).
19:07:12 <mcs> This week we will continue to work on #12827, help with code reviews, and help with other bugs that are targeted for 4.5.
19:07:19 <mcs> That's all for us.
19:07:47 <GeKo> here is what I did:
19:08:27 <GeKo> I looked at a couple of arhur's patches and merged a subset of them
19:08:46 <GeKo> then i found another Noscript bug (#15225) which Giorgio already fixed for us, thanks to him
19:09:12 <GeKo> I mainly worked on #9387 and have the explanations basically ready
19:09:28 <GeKo> some polishing is needed and I want to implement the idea mcs and brade had, too
19:09:54 <GeKo> I looked a bit at #15268 which confused me
19:10:00 <GeKo> today
19:10:33 <GeKo> I implemented the security slider portion of the MathMl patch and it is working
19:10:41 <GeKo> thanks mcs and brade
19:11:15 <GeKo> this week I plan to finish #9387, review #13548
19:11:21 <GeKo> and worked on #3861
19:11:24 <GeKo> *work
19:11:43 <GeKo> + help with code reviews and stuff for the next release
19:11:49 <GeKo> that's it from me
19:12:45 * boklm can go next
19:13:02 <boklm> This past week I added a patch for #14959 and added some noscript tests (#13053)
19:13:06 <boklm> This week I'm planning to add a test for #15225 and work on #14972
19:13:24 <boklm> that's it for me
19:13:59 * arthuredelstein can go next
19:14:15 <arthuredelstein> Last week I worked on #13670 and made some progress, but it's still not quite right. I also finished up #14324, #15086, #15207, #13891 with GeKo's help. I wrote a first attempt at a patch for #14389.
19:14:37 <arthuredelstein> This week I'll try and get #13670 finally working. I'll also try to fix the reported problem with the #5926 patch (#15268). If there's time, I'll work on submitting more patches to Mozilla.
19:14:56 <arthuredelstein> That's it for me
19:16:00 <aagbsn> is there any plan to build TBB for arm systems?
19:16:22 <GeKo> #12631
19:16:30 <aagbsn> ah awesome
19:17:33 <mikeperry> arthuredelstein: I think #13670 and #15268 are probably more important than #14389.
19:17:50 <mikeperry> esp since I'd like #13766 in 4.5-stable
19:18:08 <GeKo> indeed
19:18:29 <arthuredelstein> Yes, makes sense
19:19:29 <mikeperry> are you able to also take #7255 as part of your resizing work?
19:20:29 <GeKo> what is with #14937?
19:20:47 <mikeperry> and I'm wondering if #13650  will be too tightly coupled with #14429
19:20:54 <arthuredelstein> mikeperry: Yes. I'm not totally sure what kind of prompt we want there.
19:21:07 <arthuredelstein> (Re: #7255)
19:22:31 <arthuredelstein> GeKo: I have a workaround for #14397. But if dcf adds meek fingerprints to TBB, then we won't need to use the ugly workaround.
19:22:40 <GeKo> ok
19:23:51 <arthuredelstein> I'll post the workaround patch and then we can decide whether to land it or wait for the alternative.
19:24:01 <GeKo> good idea
19:25:56 <GeKo> re #7255: I am still against a modal dialog if we giev the user an after-the-fact notification (which we want). So, I'm in favor of trying popup notifications.
19:25:58 <mikeperry> for #7255, I am confused by Gko's comments about the type of notification as well.. Geko, are you saying you don't want to use confirmEx?
19:26:06 <GeKo> yes
19:26:42 <GeKo> I think a modal dialog is annoying for a lot of users
19:26:54 <arthuredelstein> So I guess, maybe the idea is to notify the user that TBB is constraining the window shape and explain why?
19:27:22 <arthuredelstein> So user doesn't freak out.
19:28:02 <mikeperry> I think we only need to explain ourselves when we undo a maximize
19:28:16 <mikeperry> ie just the sizemodechange event (I hope)
19:28:26 <GeKo> and fullscreen
19:28:51 <arthuredelstein> The dragging behavior is also a little weird -- do you think it's not scary enough to warrant a notification?
19:29:04 <GeKo> because I can imagine this is quite confusing if one tries to watch a video fullscreen and something strange happens with the window
19:29:09 <mikeperry> I think in the interest of minimizing popups and notifications, we say nothing there
19:29:42 <mikeperry> the ticky bit is if we can still somehow allow maximize/fullscreen without disabling everything.. I suppose even if we can't, it is fine to disable all autoresizing if the user dwants to maximize
19:30:20 <arthuredelstein> In my #14429 patch, fullscreen is still allowed
19:30:33 <GeKo> hm...
19:30:53 <arthuredelstein> It just has a black border around the edge to keep the dimensions quantized
19:31:03 <arthuredelstein> (dimensions of the content element)
19:32:15 <arthuredelstein> In contrast you click "maximize", then the window ends up being a little smaller than maximized because of quantization.
19:32:31 <arthuredelstein> *if
19:32:51 <msvb-lab> Hi folks, sorry to be late.
19:33:29 <GeKo> rereading my comment 18: the popup notification I jave in mind is only for the case we are landing #14429 and are notfying users after the fact
19:34:16 <GeKo> if we can't do that then I am all for having a modal dialog before they are shooting themselves in the foot (i.e. the old meaning of #7255)
19:34:35 <GeKo> just in case this makes things clearer
19:36:07 <GeKo> btw: I'd argue as well for having the #14429 + #7255 in the next alpha but not in the stable one given that we might need to have some release cycles before figuring out what kind
19:36:19 <GeKo> of notification really works well in which circumstances
19:38:48 <mikeperry> hrmm, so then perhaps the OCSP and locale issues are higher priority than worrying about #7255 for now
19:38:50 <GeKo> (apart from the fact that we might find some weird corner cases in #14429 itself when it gets more testing/review that we want to fix)
19:39:04 <GeKo> yes
19:39:07 <GeKo> I think so
19:39:19 <arthuredelstein> I agree that OCSP and locale are higher priority at this point
19:39:51 <GeKo> although having a fix for the original #7255 (i.e. a modal dialog or something before maximizing) would be really nice
19:42:24 <mikeperry> ok. well perhaps we can do something for that if the OCSp and locale stuff is completed. might be easy enough to do last minute as an independent thing from resizing
19:42:38 <mikeperry> and if not, postpone it
19:43:02 <GeKo> yes. I might be able to spend some cycles this week on it
19:43:52 <arthuredelstein> I do feel that #14429 patch is in pretty good shape, though.
19:44:19 <mikeperry> gamambel: are you around, by any chance? I am wondering about #14688. if we can have the NSIS create just a launch shortcut, I think that is easy enough to remove?
19:44:30 <arthuredelstein> It mainly needs more testing on other Linux desktops.
19:45:00 <GeKo> I look at it tomorrow again. we'll see how it goes :)
19:45:05 <arthuredelstein> Thanks :)
19:45:46 <gamambel> mikeperry: NSIS can generate an uninstall.exe and place it somewhere. yes it is trivial, but then you really want to add tor to the "add/remove software" section of the registry etc
19:46:02 <GeKo> ugh
19:47:11 <gamambel> what i don't like is that if you add an optional "install" and "portable install", the UI looks ugly. if you want anything more custom, you have custom strings and then ...transifex?
19:47:43 <gamambel> you can also simply throw a shortcut on the users desktop or wherever, and not deal with uninstall
19:47:48 <gamambel> doesn't feel clean either.
19:47:48 <mcs> Can we just leave it to the user to delete the shortcut or is that too sloppy?  In other words, do not provide an uninstaller.
19:48:17 <gamambel> from my experience, for "installed software" on windows, people expect to be able to remove it from within the software panel.
19:48:30 <gamambel> most people don't know how to delete an icon. and if they delete an icon, why it doesn't delete the software.
19:49:01 <gamambel> yeah we could make it so the shortcut is an optional item that people can select
19:49:02 <mcs> You are probably correct about people
19:49:14 <mcs> about people's knowledge
19:49:14 <gamambel> and just don't deal with uninstall, like now
19:49:28 <gamambel> mcs: i've worked 2 years in an IT hotline :)
19:49:39 <gamambel> and had some usability classes at uni
19:50:02 <mikeperry> I am becoming more convinced that hat the portable app model shouldn't even be the default then :/
19:50:16 <mikeperry> it seems like something only advanced users could manage to handle anyway :/
19:50:51 <gamambel> i agree. i think the default should be "installed", which means shortcuts, uninstaller, integration in "add/remove software" panel
19:51:32 <gamambel> there's much to dislike about that, i know
19:51:42 <GeKo> but that does not work without admin privs which needs some more NSIS hackery wich UAC dialogs and such IIRC
19:51:46 <mikeperry> well, in the interest of making it slightly easier, maybe we just do the shortcut for now by default, but have it be an option that can be unchecked?
19:52:04 <mikeperry> and not worry about the uninstaller + registry yet
19:52:08 <gamambel> sounds ok
19:52:11 <GeKo> s/wich/with
19:52:23 <GeKo> +1
19:53:18 <arthuredelstein> Some windows programs have an "uninstall" shortcut, which at least makes it more explicit to naive users that deleting might not be enough.
19:53:21 <mikeperry> gamambel: can you add that in a commit a commit in your nsis repo?
19:53:53 <gamambel> i'm not sure i have merged all patches someone else did
19:54:16 <special> maybe off topic: I have had success with inno setup (instead of nsis) to do this.
19:54:19 <gamambel> ( mine is https://github.com/moba/tbb-windows-installer )
19:54:36 <gamambel> i think someone made some changes. where is the copy tbb uses these days?
19:54:40 <mcs> Currently we build from https://github.com/MarkCSmith/tbb-windows-installer/ but it would be great if changes there were merged into the other)
19:56:08 <gamambel> mikeperry: can do. i can only test on XP though.
19:56:10 <gamambel> ok
19:56:26 <gamambel> i'll find out if there's anything to merge,and then add the shortcut as an option. when do you need it?
19:56:48 <GeKo> gamambel: if you upload bundles I can test on Win7, and 8
19:56:52 <mikeperry> by next tuesday (March 24th) if possible
19:57:05 <mikeperry> but earlier may allow more testing
19:57:12 <gamambel> yes i will do it within the next days
19:57:29 <mcs> I can generate a pull request if that would be helpful.
19:57:47 <mcs> (for the older changes)
19:58:09 <gamambel> ye why not
19:59:00 <mcs> https://github.com/moba/tbb-windows-installer/pull/1
20:00:18 <gamambel> i will find some time tomorrow or on wed to play with it
20:00:28 <gamambel> adding a new section is trivial, but i want to find out how to make it visually more appealing
20:02:47 <mikeperry> ok, sounds good
20:03:34 <mikeperry> I think we covered all of the tbb-team tickets then
20:04:10 <mikeperry> is everyone Ok with their assignments for https://trac.torproject.org/projects/tor/query?keywords=~tbb-4.5-alpha&status=!closed
20:04:23 <mcs> A quick question: is there anything remaining to do for #13685 ?
20:05:01 <mikeperry> I think we need to add another warning to the 4.0.5 blog post
20:05:08 * mcs is OK with assignments in trac
20:05:11 <GeKo> yup
20:05:11 <mikeperry> and maybe ask someone to send a tor-announce mail
20:05:24 <mcs> OK.
20:07:14 * arthuredelstein is also OK with assignments
20:07:24 <mikeperry> probably also tweet the https://blog.torproject.org/blog/end-life-plan-tor-browser-32-bit-macs blog post again
20:09:51 <mikeperry> ok, I think that wraps up the meeting then?
20:09:52 <mikeperry> anything else?
20:10:36 * GeKo has nothing more
20:12:28 <mikeperry> ok, thanks everyone!
20:12:33 <mikeperry> #endmeeting *baf*