Hi all!

We had another Tor Browser team meeting yesterday 1800 UTC in #tor-meeting at irc.oftc.net. The meeting log can be found at:

http://meetbot.debian.net/tor-meeting/2018/tor-meeting.2018-08-13-17.59.log....

The entries from our meeting pad are below:

Discussion: - The TBA strings file seem to contain Fennec strings. Compare:

https://gitweb.torproject.org/translation.git/tree/en/android_strings.dtd?h=...

https://hg.mozilla.org/mozilla-central/file/812e732d0928/mobile/android/base...

mcs and brade: Past two weeks: - Tried a few more things for #26514 (intermittent updater failures on Win64). - Fixed #26985 (Tor Launcher help button icons missing). - Fixed #25509 (Tor Launcher misleading proxy prompt). - Created an updated patch for #26960 (implement new about:tor). - Worked on desktop onboarding: - Created a patch for #26961 (new user onboarding). - Created a patch for #27082 (enable a limited UITour). - Researched #26962 (new feature onboarding). This week: - We will be unavailable on Thursday and Friday of this week due to family commitments. - Try jacek's suggestion for #26514 (intermittent updater failures on Win64). - Incorporate final artwork for #26961 (new user onboarding). - Work on #26962 (new feature onboarding). - Review our notes from #22074 (undocumented bugs since FF52esr) and file additional tickets if necessary. - Revisit #26381 (about:tor page does not load on first start in localized Windows bundle). Does anyone else have time to work on this ticket? -pospeselr: I can take a look! - After the Tor Browser nightly builds include the network team's fix for #26876, do some testing for #26251 (Adapt macOS snowflake compilation to new toolchain).

boklm: Past two weeks: - made a patch adding an option for incremental firefox builds (#27045) - found some way to make a program requiring GLIBCXX_3.4.22 for #25485 (Fix libstdc++.so.6 compatibility for Tor Browser) and reviewed patch from sukhe - reviewed #27020 (RBM build fails with runc version 1.01) - made patch for #27101 (Error building tor in tor browser nightly builds) - the machine doing nightly builds has some network issue, but I won't be able to fix it until next week, so I set up a new replacement VM - fixed some testsuite issues: #26981, #27106 This week: - continue working on #27105 (fixing testsuite for esr60) and subtickets - look at #12968 (HEASLR) - help with #25485 if needed (libstdc++ issue with updater) - help with building new release

pospeselr: Last weeks: - got cross-compiled incremental mingw windows builds working, initial setup and build still takes awhile (but less than a full rbm build), but subsequent builds after changes around ~2 minutes, mostly linking (in a VM) - workflow for printf debugging windows tor-browser is now way less painful - winsetup and winbuild scripts available in a branch here: https://gitweb.torproject.org/user/richard/tor-browser.git/commit/?h=winbuil... - shouldn't we apply ifdef'd versions of the build-time windows specific patches ( 1464084.patch, 1467041.patch and strcmpi -> stricmp) to the firefox source directly rather than applying them in rbm? [GeKo: it's meant to indicate their temporal status essentially being patches not meant to be carried around in tor-browser, although this could possibly be debated in the strcmpi->stricmp case] - fix for #26874 ( Investigate UNC path restrictions failure in Tor Browser 8.0a9 ) - turns out this is actually broken in ESR60 when FF is *not* installed to an SMB share

This week: - uplift fix for #26874 - update wiki with cross-compile mingw build instructions - review #27045 ( Add option for firefox incremental builds ) - review gk's patches - #26381 (about:tor page does not load on first start in localized Windows bundle) - #21785 (Keep an eye on the Storage API ) - Portland for Rustconf Thursday-Friday - my initial reason for going to this was to get some hands on experience with rust, but do we as a team have any goals we wish to accomplish with a presence here? - considerations for rust for tbb standalone launcher, rust-qt bindings?

igt0: Last Weeks: - Make sure the #26884 works with noscript webextension (it does!) - Worked on TBA onboarding #25696 - The final assets are missing, I am working together with Antonela on it. - The design is slightly different than the one in the bug. In landscape mode the UI doesn't have enough space, so the whole view scrolls. - Started #27111 (about:tor for mobile) This Week: - Finish #27111 - Update #26574 addressing Geko's comments - Review mobile related patches

sysrqb: Last weeks: Audited third party libraries in Fennec (#26982, #25851 (plus child tickets)) Reviewed permissions requested by Fennec and created patches (#24796, #26825, #26826) Investigated Fennec's Firefox Account code and nearly finished a patch for it (#26858) Researched how Mozilla localize Fennec and build localized releases (#26843) This week: Finish Firefox Account Manager patch (#26858) TBA onboarding ticket Make sure we're ready for releasing the first alpha version of TBA RustConf on Friday

tjr mingw - Have a build in -central that builds.... just doesn't run =(

pospeselr: you using msvcr whatever spec file? nevermind!

tjr: this is using mingw-clang w/ ucrt so we shouldn't need it

fuzzyfox needs some feedback from david; but it probably getting closer

arthuredelstein: Last two weeks: * Was afk for a few days * Audited HTTP2 implementation and patch for #14952 * Wrote patch for #24056 (UI locale is detectable by button width) * Wrote patch for #17252 (Confirm TLS session resumption/ID respect FPI) * Confirmed #21787 (Make sure exposing calendar info does not leak the locale) * Reviewed #26655 (onion button wrong size and color) * Re-verified #17512 (Tor Browser installer isn't properly localized) * Opened #27123 (ping/settings timing side-channels, #27098 (monthly giving banner), #23925 (2018 donation banner) #27097 (Join email list) * Updated Tor Browser locale monitoring: torpat.ch/locales * Wrote patch for #26609 (investigate whether the -moz-windows-accent-color-in-titlebar media query adds a fingerprinting vector) -- patch soon Coming week: * Revise #14952 as needed * More tbb-fingerprinting patches * torbutton banner

GeKo: Last Week: -being on vacation (the week before mainly reviewing patches, working on #26475 and a bit on the network code review) This Week: -trying to get releases out -I need help with reviews, in particular a second review for #26456, #24056 and tickets I provided patches for (#26628, #26833, #26655) -work on my backlog -work on #26475 -work on network code review

sukhe: Last Weeks: - #25485 (libstdc++): merged but will need revision - #26476 (Windows crash): still investigating, doing log comparison now - no longer blocks on tjr - #12968 (HEASLR): no progress so now asked boklm to look into it - (offline two days last week) This Week: - #26476 (Windows crash) - Revise #25485? - Hooman and I are going to resume #25483 (Windows reproducible build of snowflake) - anything else I can pick since the above two involve building and won't take up all my time? [GeKo: I'll look at some tickets after the meeting, e.g.:#27061 and #24465]

sisbell: Last Week: Confirmed that generated Android APKs have different checksums after 2 builds (classes.dex different) Investigating diffs and android grade builds This Week: Continue investigating why classes.dex isn’t reproducible Setup Mac environment for builds Attending RustConf Thurs/Fri

Georg