18:59:35 <GeKo> #startmeeting tor-browser
18:59:35 <MeetBot> Meeting started Mon Feb  1 18:59:35 2016 UTC.  The chair is GeKo. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:59:35 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
18:59:43 <arthuredelstein> hi everyone
19:00:05 <GeKo> welcome to the first tor browser meeting after the 5.0 series went EOL
19:00:24 <GeKo> i might start this time with the status updates
19:00:41 <GeKo> last week i helped with the releases and fought with our signing setup
19:01:14 <GeKo> then after all releases were out i started investigating and triaging things that broke
19:01:54 <GeKo> i triaged tickets for feb 2016 and tagged them with TorBrowserTeam201602
19:02:45 <GeKo> this week we'll probably want to release a fixup stable version (we can talk later about it) and i finally try to get to the design document update
19:03:19 <GeKo> oh, while filing #18198 i found a fix for it which I'll post later
19:03:28 <GeKo> that's it for me
19:03:52 * mcs will go next
19:04:01 <mcs> Last week, Kathy and I created a patch for #18144 (needs revision).
19:04:05 <mcs> We did some testing of 6.0a1 and filed #18154.
19:04:11 <mcs> We made some progress on #13252.
19:04:17 <mcs> And then we dropped everything to debug #18168 and create a patch for it.
19:04:24 <mcs> This week we will create a revised patch for #18144 and get back to work on #13252.
19:04:45 <mcs> And help with any critical 5.5 bugs.
19:04:50 <mcs> That's all for us.
19:05:24 * arthuredelstein can go
19:05:36 <arthuredelstein> This week I made patches for #18169, #18172, #18030
19:05:44 <arthuredelstein> I wrote revised patches for #17965.
19:05:51 * huseby next
19:05:56 <arthuredelstein> I worked on revisions for https://bugzilla.mozilla.org/1121643
19:06:03 <arthuredelstein> And I did some rebasing work for #15197.
19:06:15 <arthuredelstein> This week I plan to work on review #18168 and work more on rebasing and upstreaming the font patch.
19:06:23 <arthuredelstein> That's all for me.
19:06:55 <huseby> arthuredelstein: re 1121643
19:07:07 <huseby> can we add the ability to have firefox report a list of fonts programatically?
19:07:23 <huseby> are you baking the whitelist into the code?
19:07:42 <arthuredelstein> No, it's a pref
19:08:13 <arthuredelstein> What reporting do you have in mind?
19:08:58 <arthuredelstein> The pref is called "font.system.whitelist" and the value string can look like "Arial, Courier, Monaco, Times", for example.
19:09:58 <huseby> arthuredelstein: I want to eventually have firefox report all of the fingerprinting types of variables to some concensus mechanism that then says back to all firefoxes, hey, the concensus is that we all have Arial and Times so just report that until the next concensus round
19:10:10 <huseby> so i want to be able to dynamically set what firefox reports as the installed fonts
19:10:33 <huseby> i want to do that for fonts, graphics capabilities, named css colors, etc
19:11:06 <huseby> so that all users that have the master "concensus" pref turned on, will report the exact same stuff to fingerprinting scripts
19:11:11 <huseby> making a huge cohort
19:11:15 <arthuredelstein> The next revision of the pref that I'm working on will hopefully let you dynamically set the whitelist (from chrome code).
19:11:26 <huseby> everybody will just be a regular tree in a vast forest
19:11:36 <huseby> nice
19:11:53 <huseby> arthuredelstein: my plan was to try to expose these things via an API so that add-ons could control it
19:12:04 <huseby> that way we can have different add-ons that do this sort of thing
19:12:18 <arthuredelstein> I'm not sure I see the advantage of this approach.
19:12:19 <huseby> I can see there being some disagreement on the best concensus mechanism
19:12:36 <huseby> e.g. sending it all to a central server vs. a distributed concensus
19:12:43 <huseby> anyway, it's all researchy
19:12:54 <GeKo> arthuredelstein: fwiw it seems the important piece of 1237831 did not get  backported to mozilla-beta.
19:12:57 <huseby> the first step is to just make it settable at runtime
19:13:10 * huseby will go now
19:13:19 <huseby> the origin attributes work keeps chugging along
19:13:26 <huseby> I've got more patches out for review
19:13:27 <GeKo> arthuredelstein: am i missing something? if not could you try to get that into it?
19:13:30 <arthuredelstein> GeKo: Oh no.
19:13:36 <huseby> we're writing test to validate isolation as we go.
19:13:43 <arthuredelstein> GeKo: Yeah, I'll look into it.
19:13:44 <huseby> we're touching lots of stuff and it has me nervous
19:13:57 <arthuredelstein> huseby: Are you using our isolation tests?
19:14:22 <huseby> the tracking spreadsheet is here: https://docs.google.com/spreadsheets/d/1MzKb8Bodhp3JhqpLArpnVqT4hkV-e1owXyM_yBZKlOs/edit?ts=568337c9&pref=2&pli=1#gid=1965805550
19:14:32 <huseby> arthuredelstein: not yet
19:14:37 <huseby> are those upstreamed?
19:14:46 <huseby> are they in central yet?
19:14:55 <huseby> I should upstream those as a conditional compile if they aren't
19:14:57 <arthuredelstein> The isolation tests? No.
19:15:05 <huseby> that way I can TDD the isolation
19:15:20 <arthuredelstein> That would be cool.
19:15:22 <huseby> arthuredelstein: can you point me to where the isolation tests are
19:15:24 <huseby> ?
19:15:27 <huseby> I'll cherry pick them ASAP
19:15:28 <arthuredelstein> Yes, I'll send you a list.
19:15:33 <huseby> danke
19:15:36 <huseby> that's it for me
19:15:49 * boklm can go next
19:16:03 <boklm> This past week I helped on the release of the alpha and hardened versions
19:16:10 <boklm> I looked at #18127 and rebased our gitian patches on upstream gitian in order to try the Debian LXC patch
19:16:19 <boklm> After trying a KVM build with this gitian branch I got a non-matching 6.0a1 build (but didn't investigate why yet)
19:16:32 <boklm> I rebased my tor-browser split branch repo on 38.6.0esr in order to push it to Mozilla Try
19:16:35 <boklm> This week I will push the split branch repo to Mozilla Try and compare results with the previous release to check for any regression in the tests
19:16:44 <boklm> Investigate the non-matching build with the rebased gitian and try an LXC build
19:17:02 <boklm> And help with the fixup stable release if we do one
19:17:11 <boklm> That's it for me
19:18:13 <GeKo> okay, who else is here for a status report?
19:19:34 <GeKo> let's start the dicussion then. i've just one topic which is a follow-up 5.5.1
19:19:48 <GeKo> which basically contains the fix for #18168, #18169 and #18172
19:20:12 <GeKo> does taht make sense?
19:20:15 <GeKo> *that
19:20:46 <GeKo> i think we can skip the alphas as these bugs were there for weeks/months and nobody noticed
19:21:00 <GeKo> it seems our userbase could be more widespread there :/
19:21:26 <arthuredelstein> Do we have an estimate on the size of the alpha userbase?
19:21:53 <mcs> We definitely need more people to run the alpha. I wonder if it would help if we started calling it a beta at some point during our cycle?
19:22:04 <arthuredelstein> GeKo: I think your plan makes sense.
19:22:24 <mcs> Kathy and I have not had a chance to look at #18170 yet which might be another candidate for the 5.5.1 release.
19:22:27 <GeKo> arthuredelstein: not really. we could try to look at the update check pings
19:22:40 <boklm> I think it makes sense to have a 5.5.1
19:22:53 <GeKo> mcs: yeah, but i'd say this can wait a release
19:23:11 <mcs> OK; no problem from our perspective.
19:23:30 <GeKo> not sure if renaming it to "beta" would make sense. we could think about it
19:23:53 <arthuredelstein> Another possibility is to make it easier to reach on the website.
19:24:25 <GeKo> maybe but i wonder if people that would try an alpha would really have a problem finding it on our website
19:24:51 <arthuredelstein> Sometimes I have trouble finding on the website ;)
19:24:56 <arthuredelstein> *it
19:24:59 <mcs> Probably we need to more actively recruit people to run the alpha.
19:25:00 <GeKo> which does not mean we should tr burying it
19:25:13 <GeKo> arthuredelstein: ha, okay, fair point then :)
19:25:30 <GeKo> okay. this is something to mull over at least
19:25:47 <GeKo> for the release i thought about tagging it tomorrow anf getting a build started
19:26:26 <GeKo> arthuredelstein: could you have a look at #18168 today then? i looked at the code and compiled it on all platforms and it seems to be good
19:26:35 <GeKo> i plan to look closer tomorrow and do some testing
19:26:36 <arthuredelstein> Yes, will do
19:26:53 <GeKo> and could you do the small fixup for #18172?
19:27:26 <GeKo> apart from those we are fine i think
19:27:29 <arthuredelstein> Yes, I'll do that today as well
19:27:35 <GeKo> cool, thanks
19:28:29 <GeKo> what else do we have to talk about?
19:31:10 <GeKo> thanks, then and this meeting is over *baf*
19:31:13 <GeKo> #endmeeting