- Fixed #23266 (pub use privacy error) and a little refactoring of resolve
- DST coercions RFC (https://github.com/rust-lang/rfcs/pull/982)
- Fixed #18601 (coercions are valid casts, trivial cast lint, #23630)
- adjusted above to all allow (RFC issue and implementation)
- one day sick leave
- work on unqualified assoc items
Plus some minor bits and pieces for rustc performance tracking, rustfmt, coordination of refactoring project
FIx Self:: in impls, item scoping. Set q2 goals.
- Off Thurs/Fri
- Various bits of AWSY cleanup and testing, e10s followup
- landed all parts: Bug 1131416 - Desktop syncing module for reading list service
- filed, landed: Bug 1148060 - test_Sync.js fails due to defining
deleteon an object pre-39
- landed: Bug 1148201 - Fix desktop reading list client's use of Last-Modified, If-Unmodified-Since, If-Modified-Since headers
- closed: Bug 1143822 - Rename reading list SQLite database for final release
- closed: Bug 1147713 - Various ReadingList tests are going to permafail when Gecko 39 merges to Aurora
- assigned: Bug 1147554 - 13% win7 non main startup file IO bytes on fx-team (v.39) on March 22 from push b5bfda73ba54
- assigned: Bug 1149105 - Various sync failures including "Failed to add a new item from server record: Error: The item must have a url"
- assigned: Bug 1149023 - unable to remove reading list item
- filed, assigned: Bug 1149302 - Error when adding an item to reading list
- filed: Bug 1149336 - Fix reading list Sync.jsm TODOs
- filed: Bug 1149337 - Beef up reading list sync tests
- reviewed: Bug 1147113 - Filter the article properties in ReaderParent.jsm instead of ReadingList.jsm
- reviewed: Bug 1143797 - Allow clicking on suggested explanation text to see overlay explaining the suggested tile
- reviewed: Bug 1146249 - Tiles on the newtab page don't wrap properly
- reviewed: Bug 1143745 - Update the way Firefox reads directoryLinks.json sent from the server for the tiles v3 endpoint
- reviewed: Bug 1140496 - Only show a suggested tile url for some number of times or until clicked
- reviewed: Bug 1145303 - Move "What is this page?" to the bottom middle
- reviewed: Bug 1136208 - Rename various references of related in Firefox
- reviewed: Bug 1148217 - Readinglist sync engine should catch and log errors accessing the local store to prevent entire sync failing
- reviewed: Bug 1148252 - Reading-list sync engine should gracefully handle local item already existing on the server
- reviewed: Bug 1141617 - Land suggested tiles string changes as code for uplift to Aurora 38
- reviewed: Bug 1136203 - Remove history thumbnail/title replacing functionality
- reviewed: Bug 1148273 - Readinglist scheduler doesn't handle FxA error states correctly.
- reviewed: Bug 1146146 - Maximize the number of rows of tiles by reducing the suggested explanation maximum line count to 2 instead of 3
- reviewed: Bug 1149021 - Suggested tile with just an image shows a thumbnail instead
- reviewed: Bug 1148938 - Temporarily test suggested tiles for Nightly and Aurora
- needinfoed: Bug 1146358 - The attempt of adding a page to Reading List from Reader View fails and causes an error
- needinfoed: Bug 1147554 - 13% win7 non main startup file IO bytes on fx-team (v.39) on March 22 from push b5bfda73ba54
- bug 1146565: [EME] EME-free Firefox fails to uninstall any pre-existing CDMs - Fixed and uplifted to aurora (FF28).
- bug 1146201: [EME] Delay navigator.requestMediaKeySystemAccess while we wait for CDM to download - Reviewed patch.
- Downloaded all data for update hotfix analysis, set up system to generate report on update hotfixes, initiated deduplication job (has been running for 5+ days now)
- Generate report on update hotfixes, to be used in report on update orphaning.
- Installer/updater code reviews
- bug 1147008: Uninstalls of GMPs should have additional tests
Bug 1147953, bug 1144317 (refresh driver silk windows 10) - feedback, IRC discussions with mchang on windows 10 issues, causes, how to land, and stuff. he was testing silk timing on in a windows VM, now he uses real HW.
Bug 1147487 (21% tsvgr_opacity PGO linux regression) - too many hours on IRC and at the bug with Gijs. Still not resolved, but it seems on track now.
Investigated responsiveness metric During talos runs (FX_REFRESH_DRIVER_FRAME_DELAY_MS) - e10s does better: shorter tail, 1/4 of e10s samples beyond 17ms - https://etherpad.mozilla.org/tJ9ehJIKB0 and (roberto) http://nbviewer.ipython.org/gist/vitillo/cb6f1304316c1c1a2cbc
started examining yoric's cpu measurements patch. didn't get too far, discussed some metrics (frames as as unit of time is probably not good).
The big things from the past few weeks:
- Bug 1148032 - BroadcastChannel API bypasses private browsing mode
- Bug 1147990 - Pass W3C websocket test suite
- Bug 1113086 - Implement AudioChannel API into BrowserElement - default mute/unmuted by pref
- Bug 1148527 - bug 1145631 screwed up the indentation of way too many files.
- Bug 803537 - XHR + blob: URI crashes in workers
- Investigated telemetry alerts
- Histograms are automatically merged now when using the Spark API
- e10s analysis http://nbviewer.ipython.org/gist/vitillo/cb6f1304316c1c1a2cbc
- Bug 1139513 - Warn and gather data if ServiceWorker hits max workers per domain limit - reviewed
- Adblock plus hang reduction analysis - http://nbviewer.ipython.org/gist/vitillo/3ff6d0a85825ced3f0df
- Pushed PR 44 on data-pipeline
- Bug 1148473 - Lower BHR threshold - ready to commit
- Deal with expiring histograms
- More of the same
- Reviewed bug 1147243, bug 1146869, bug 973933, bug 1146096, bug 1148062, bug 1147685
- Landed bug 1142908 to make arm simulator builds use tooltool gcc for host, and relanded bug 1142352 and bug 1142420 for dbaron.
- bug 1147183: UnicodeEncodeError: 'ascii' codec can't encode characters in position 39-44: ordinal not in range(128) when running mach build python/check. Landed.
- bug 1147217: Improve l10n repack error message when locale doesn't contain necessary files. Landed.
- bug 1147283: Replace mozpack.path with mozpath. Landed.
- bug 1147207: Rework the l10n repack code to handle giving it separate directories for different parts of localization. Landed. Partially backed out, fixup waiting for review.
- bug 1147723: Avoid non TEST_PASS/TEST_UNEXPECTED_FAIL output from test_files.py. Landed.
- bug 1147760: Using mozpack.files.FileCopier in-place can fail with XPTFiles. Landed.
- Finished and landed bug 1138293 and bug 1142434 removing many moz_* functions from mozalloc.
- Finally landed bug 1138824 (Avoid installing import libraries in $(DIST)/lib) after NSS was updated on trunk.
- bug 1149416: nsTArray.h:188:44: error: attempt to free a non-heap object 'tautoconst' [-Werror=free-nonheap-object]. Landed.
- More WIP towards builds with no compilation.
- bug 1134920 and bug 1134923: As removing nsMemory and NS_* allocation functions has an impact on the various xpcom glues, studied that impact. Will try to find a minimal way forward, but thinking about dropping the xpcom standalone glue on the medium term.
- More of the same.
- Looooots of time hunting down elusive segfault.
- Testing profiling on a slow Windows machine.
- Bug 674779 - Add per-compartment CPU accounting - applying feedback, adding tests, fixing ux
- Bug 1147664 - PerformanceStatsService should offer a detailed mode that doesn't collapse the platform in a single group - filed
- Bug 1147348 - nsIPerformanceSnapshot should expose the timestamp of the snapshot - filed
- Bug 1146948 - about:performance should let me pause/unload a tab - filed
- Bug 1146947 - about:performance should let me switch to a tab - filed
- Bug 1146945 - about:performance should display the
- Brainstorming with JSAPI about memory
- Brainstorming wit dbaron about other operations
- Bug 992896: [OS.File] Add a function watching for a file being changed under Linux - reviewing
- Bug 1083361 - Expose an API to track DOM Promise which are left uncaugh - updated
- Bug 1087110 - Intermittent browser_experiments.js | A promise chain failed to handle a rejection: - at resource://app/modules/experiments/Experiments.jsm:876 - Error: uninit() alrady called - finished (landed)
- Bug 1140558 - Telemetry subsessions should be submitted with their own environment, not the new environment - reviewed (landed)
- Bug 1076775 - Implement History.removeHistoryByFilter - unbitrotting, applying feedback
- Generally, unbitrotting
- Bug 1147373 - JS backquotes break syntax highlighting - filed
- Eyrolles event
- More discussion about memory
- Bug 1083361 - Expose an API to track DOM Promise which are left uncaught - finish, land
- Finish unbitrotting, land
- Tuesday and Wednesday at Adobe office, Thursday and Friday in SFO.
Written, awaiting review: * bug 1133351
- Sort out Q2 goals
- Security bugs
- Land bug 1133351
- Finally fixed the SM(cgc) build naming stuff
- Turned out to be that my github pull request from a month ago was never merged
- Got something sorta kinda numerically stable from gc-ubench - bug 1147995
- Tracked down the destructor weirdness with the rooting analysis - bug 1056410
- Fixed all the false positives blocking the landing of array handling in the analysis, a longstanding hole - bug 963738
- Another swing at jorendb.js. once again, enough to get it working for my purposes, still a little shy of something I'd be ok landing - bug 944164
- Finally landed the sixgill fixes on the tooltool server, fixing memory usage, some unimplemented constructs, some other stuff - bug 1091298, bug 1126544
Landed the heap size graphing in gc-ubench - bug 1125412
Landed timeouts for SM(...) builds - bug 1135399
- Failed attempt adding OSX arm-sim builds. Will have to circle back to this. bug 1146520
- work on improving hazard build runnability
- fix osx arm-sim builds
- respond to needinfos
- mdns crap for b2g
- weirdo jsautokwgen build crap
- fixed up FetchEvent.isReload
- addressed html5ever review
- investigated synthesizing redirections some more
- Servo reviews
- wrote patches to get e10s redirections for SW intercepts working
- worked on viewport resizing + responsive images
- filed issues blocking the jQuery and React test suites, as well as Dromaeo
- slapped together the start of WebDriver integration
- Tackle some more responsive images intermittent failures and unlanded fixes
- Address the e10s SW redirection review
- Address e10s SW + tabchild review
- Write a patch for e10s + SW script fetching
Opened zero-copy html5ever PR
Started working on another iteration of that idea
Rustup for: urpc, string-cache, rust-phf
Revived Servo PR 4215 for replacing stb-image
Q2 goals planning. Thinking more about concrete deliverables for security.
Keep working on zero-copy string representation. Do some serious fuzz testing of my new library's rather
Land stb-image replacement.
- Looked into profiler backend implementation. Tried about 3 or 4 ways of slicing into the codebase to refactor it out. Not possible to refactor without major changes, and would probably need Benoit's input as well. Decided it's better to focus on implementing short-term needs, and consider writing a new profiler backend from scratch once Benoit gets back.
- Measured performance overheads of profiling on Windows, OSX, and Linux. Measured distinct cost of both invoking the sampler (i.e. the cost of interrupt), as well as cost of sampling itself (i.e. the stack walk and associated operations). Results posted on bug 1143598. Decided high-resolution profiling will require lazy symbolication of JS frames.
Help review/land lazy symbolication (bug 1137569) of JS frames in profiler.
Add pref for choosing profiler frequency.
Move OSX profiler implementation to use pthread_kill instead of thread suspend/resume.
- Final review / code merge PR #1420
- Review / merge PR #1426
- Review / merge PR #1429
- Review / merge PR #1432
- Review / merge PR #1433
- More work on Heka buffered routing branch
- Reproduced / started debugging reported LogstreamerInput log rotation bug
- More work on Heka buffered routing branch
- Resolve LogstreamerInput log rotation bug
- Release Heka 0.9.2
- Began redesign of Sync panel https://www.dropbox.com/s/n2g9jpqoy0d8qdp/Firefox-In-Content-Prefs-FxA-Sync-Signed-Out.png?dl=0
- Password mgr, saw first incremental password change in Nightly launch
- Engagement flow https://www.lucidchart.com/documents/view/b9161199-e94f-4a1b-8af0-2f8756038605
- Review feedback on Firefox Account identity control https://www.dropbox.com/s/jlkvr5xlqzzbvvy/Identity%20Control.pdf?dl=0
- Review upcoming work in FxA roadmap https://docs.google.com/document/d/1A4DaBA1sWYmJ1p1ILQTWzEeTTfl61TuhapBctDcwJ0Q/edit?usp=sharing
- Continue work with Holly and Michael on iframed Sync FxA
- Review upcoming work in FxA roadmap https://docs.google.com/document/d/1A4DaBA1sWYmJ1p1ILQTWzEeTTfl61TuhapBctDcwJ0Q/edit?usp=sharing
- Scaled all of the icons for the Windows main window to 2x
- Search works and meetings
- Control Center visual design
- netvc BoF at IETF 92
- worked on refactoring ogg reader code in daala examples
- simplified daala header handlings
- worked on gathering statistics for block size decisions
- Servo interviews
- Servo interviews
- more Q2 planning
- Q1 debriefing
- try to finish up last week's daala experiments
- Talk with people to get up to date on EVERYTHING
- More reviews
- Assemble issues around perceived perf and quality
- Continued mockup work for a new search+awesome-bar feature in FF.
- Participated in a ideation session with Yahoo around incorporating new search products.
- Working on a reduced feature set alpha version of search that's in development now.
- Providing feedback for the iOS version of FF.
- Look at full screen + multiple other permissions at the same time (games)
- Possibly run Firefox download test with usertesting.com to see if Google ads are making a difference in preventing people from ending up on 3rd party sites.
- More Q2 planning
- cartoon nearing completion
- big update to the way FxA loads to reduce jank/improve feel on mobile (on today's train)
- got up and running with universal search builds
- wrap up chronicle classic obligations
- actually do coppa stuff
- define Q2 goals
- minimally style search iframe
- PTO Friday/Monday
- Task Continuity US: Finishing final report.
- Task Continuity US: Writing blog entry on study.
- Task Continuity Asia: Paperwork for recruiters and travel.
- Task Continuity Asia: Reviewing field guide
- Password Manager: Finalizing contractor hiring
- Search Prototype: Prepping small usertesting test
- Collaborated with Stomlinson and Rfeeley on a variety of Accounts related things
- Updating copy on /Sync product page per Rfeeley guidance
- Account Verified snippet [ github issue ]
- Copy for "send to device..." module, /sync product page, /iOS product page, FTE landing page
- FTE "reader view friendly" landing page designs are final (collab w/ Lee Tom, Mmaslaney, Agibson) - [ PDF ]
- Launching 2 variations of Firefox family navigation this week, which is a unifying nav across primary Firefox product pages
- Usertesting.com and GA testing this and next week
- Wrap up copy for all 38.1 campaign projects
- Continue to collaborate with Rfeeley and Stomlinson on Account Verified snippet for 38.1
- Persona workshop with Lindsay Kenzig
- Task Continuity: Created a near-final draft of the Task Continuity research report. Sharing with Bill today
- Task Continuity: Worked with Markus to establish a framework model for task continuity
- Task Continuity: Drafted blog post on U.S. research
- First Run: On-Boarding Sync-up & Fennec first run planning with Anthony
- MoFo Research: Met with Laura to help her plan Chicago mofo research
- Task Continuity: Finalize presentation
- Task Continuity: Plan Asia research
- Task Continuity: Finalize/post blog post
- First Run: Draft testing plan
- MoFo Research: Connect Laura to local community leaders
• Updates to tiles UI on Fx 39 submitted
• Updates to on-boarding for tiles in progress - UI changes made, just waiting on final copy from team
• Work with Aaron L. (new hire) to develop mobile strategy for tiles
- One more pass on keyboard nav.
- Set up MozReview, and use it for a review request.
- Reading List:
- Continue looking into Reading List polish bugs.
- Mapped out initial First Run tests w/ Gemma
- Bug 1140169: Polish Android Menu (bugs filed)
- Co-ordination around Tracking Protection, Search, Reading List, etc
- Bugs backlog
- Presented at EFL Dev Day
- Meetings with various Samsung Research groups
- Kick off Q2 planning
- Expense reporting, HR stuff, etc.
- Fiddle with rust-in-gecko
- Working on RTL layout, submitted PR #5349.
- Layout PR reviews.
- Dependency rustup fixes (rust-lang/num#69, PistonDevelopers/image#344)
- Rust doc fixes (e.g. rust-lang/rust#23644).
- Playing with servo-shell.
- Still working on overconstrained layout bugs with writing-mode.
- File issue and create a test case for block formatting context impacted by a sibling float.
- Implement :focus pseudo-class, for use in servo-shell chrome.
- Hello Web Sharing review with Telefonica
- Text Chat in Firefox Hello: https://docs.google.com/a/mozilla.com/document/d/1HwYbM4Eru3rHVT3MStO8eeTunJ7pN5PYBingtmwe-bA/edit?usp=sharing
- Conversation Window redesign
- Other Hello Bugs: Bug 1131584, Bug 1145053, Bug 1104746, Bug 1137197, Bug 1146919, Bug 1115340, Bug 1146919, Bug 1129172, Bug 1133532
Continue with redesign, finalize chat, more web-sharing.
- Reading List / Reader View Polish & Bugs
- Task Continuity Model (with Gemma)
- dialog for blocked unsigned addons
- Task Continuity planing
- Reviewed patches, specifically jesup's
LazySingletonThreadpatch in bug 1145354. Got some helpful feedback on the review from bent.
- TSan runs to check goodness of some recent patch landings and to diagnose possible intermittents. Filed a few more TSan bugs.
- Wrote up a bit for Q2 goals.
- Looked at some power consumption stuff. Intel's VTune purports to come with power-measuring things.
- 1:1 with Nick.
- Fought neck strain + headaches all week.
- More of the same.
- Getting goals figured out, both Q1 and Q2.
- Take one day off in observance of my birthday.
- First steps in making the lua_sandbox a standalone package https://github.com/mozilla-services/lua_sandbox/pull/84. Including updates to the circular_buffer, bloom_filter, cjson, and hyperloglog modules.
- FxChannelSwitching https://bugzilla.mozilla.org/show_bug.cgi?id=1132079
- Cleanly stopping a terminated sandbox: https://github.com/mozilla-services/heka/pull/1432
- CPack modifications https://github.com/mozilla-services/lua_sandbox/issues/85
- PPA setup https://github.com/mozilla-services/heka/issues/1417
- Assist mreid with running Heka reports out of S3
- Make sub benchmarks visible.
- Better noise detection
- Improved GUI (e.g. single regression view)
- Fix sec bug 1146410
- Landed filtering at andor: bug 1138881, was backed out due to failures
- WIP: Knockout: faster string compares. Good initial numbers
AWFY: - Created tmp solution to jemalloc breaks OS X
- Completed bug 1143796: Increasing TelemetryScheduler ticking interval when user is not active (Pending Review)
- Completed bug 1139754: Remove idle-daily pings (r+, but waiting for Telemetry Phase 2 to land before checking that one in)
- Completed bug 1120379: Telemetry needs to send ping deletion messages to the server when FHR is deactivated (Pending Review)
- Completed bug 1137353: Port the client id loading from DRS to TelemetryPing (pending 2nd round of review, don't expect big changes)
- Completed bug 1147522: Move SessionRecorder out of datareporting to a more general shared location (pending review)
- Updated bug 1137252 and pending review on it.
19 reviews including bug 1144366, 2 feedbacks, 6 needinfos
Patched and landed, or debugged and landed:
- bug 1143810 (remove some JSSetterOps from XPConnect)
- bug 1139683 (SetExistingProperty rewrite)
- bug 1142794 (mass SetProperty type signature change)
- bug 1138489 (remove JSPROP_INDEX)
- bug 1144819 (stop using a class JSSetterOp in a particularly weird case)
- bug 1142761 (a trivial refactoring)
- bug 1142775 (simplification in our native [[Set]] code)
- bug 1142784 (preliminary cleanups to ES6 DefineProperty)
- bug 1146165 (fix a poor choice I made earlier re: Proxy::set JIT interaction)
That is a lot of refactoring, but all of that came out of necessity. My only goal was to get NativeDefineProperty rewritten without having impossible-to-diagnose test failures.
Posted patches for review in these bugs (all these patches have green try server runs, except the two trivial ones at the end):
- bug 1138499 (Assert descriptor sanity on entry to DefineProperty and exit from GetOwnPropertyDescriptor --- this is an awesome patch and I am dying to land it before we break it)
- bug 1142828 (refactor to avoid having [[Set]]-specific paths in NativeDefineProperty code)
- bug 1147660 (refactor to separate ES6 checks from implementation details in NativeDefineProperty, since the ES6 checks need to be almost completely rewritten)
- bug 1148568 (standard-compliant DefineProperty calls in JSON.parse with reviver)
- bug 1148652 (NativeDefineProperty ES6 compliance for arrays, typed arrays, arguments)
- a test for bug 1123875 which got fixed at some point
- update jsapi-tests/README in bug 1145636
- add a couple tests and post patches for review in bug 1148750
- land the fix for the security-sensitive bug
- rebase 30 patches across the T* changeset
- clear my request queue
- mentoring: bug 1063946
- inline clamp in ion (bug 1084319)
- inline int32x4.bool in ion (bug 1146363). Note that with this and the previous patch, all SIMD operations for int32x4 and float32x4 are inlined in Ion.
- make more SIMD node movable (bug 1146311)
- cleanups in the interpreter (bug 1146295)
- some pull requests to the polyfill repository (added checks for shuffle/swizzle/load/store + wrap the polyfill in a IIFE).
- AWFY: added support for local runs: running
python driver/dostuff -s printwill run AWFY locally, without the need for setting up a distant PHP server.
- reviews (bug 1092544, bug 1146223, bug 1147144)
- Fixed Math.ceil so that it returns Number by default, not Double (bug 1147405)
- looked at the assertion triggered in bug 1145644
- fixed non-ion builds (bug 1144692)
- opened bug 1146893 (tabs can't be moved anymore), resolved as duplicate
- fuzz bug 1147662
- prepared slides and give a quick presentation of ES6 features at a open-source/libre meetup at Lyon, France
PTO all week
- 2015-03-23: wrote servo PR #5323 Fix warnings in script.
- 2015-03-23: reviewed r/4335 Add TODO comments for a recently opened issue
- 2015-03-23: reviewed r/4334 dom::urlsearchparams cleanup and documentation
- 2015-03-23: wrote servo PR #5324 Remove allow(missing_copy_implementations) attributes.
- 2015-03-23: wrote web-platform-tests PR #1684 Remove grouping-blockquote test.
- 2015-03-23: wrote web-platform-tests PR #1685 Rewrite pointerevent_constructor to avoid the explicit_done.
- 2015-03-23: submitted servo PR #5327 Fix double-panic when the script task panics.
- 2015-03-23: reviewed r/4341 Fix double-panic when the script task panics.
- 2015-03-23: reviewed r/4342 Fixed some deprecation errors in components/net.
- 2015-03-24: filed servo issue #5340 Panic during panic causes huge memory spike
- 2015-03-24: wrote servo PR #5341 Update web-platform-tests.
- 2015-03-24: reviewed r/4346 Cleanup and modernize script::dom::document
- 2015-03-25: filed servo issue #5353 Code duplication in components/servo/lib.rs and ports/gonk/src/lib.rs.
- 2015-03-25: reviewed bug 1147248
- 2015-03-25: wrote servo PR #5354 Cleanup canvas_paint_task.rs.
- 2015-03-25: filed rust issue #23696 TcpStream::duplicate is not implemented.
- 2015-03-25: filed servo issue #5355 Move devtools to std::net.
- 2015-03-25: wrote presto-testo PR #44 Start organizing tests per-spec.
- 2015-03-25: wrote rfcs PR #1013 Replace duplicate() by try_clone() in std::net.
- 2015-03-25: wrote servo PR #5356 Move devtools to std::net (fixes #5355).
- 2015-03-25: wrote servo PR #5358 Fix warnings related to std::fs::PathExt on Android.
- 2015-03-25: wrote servo PR #5362 Remove old_path usage from layout.
- 2015-03-26: wrote servo PR #5381 Stop using old_path.
- 2015-03-26: wrote servo PR #5382 Remove unused imports in layout.
- 2015-03-26: wrote rust-stb-image PR #66 Modernize.
- 2015-03-26: wrote mozjs PR #30 Move to a name without hyphens.
- 2015-03-26: wrote mozjs PR #31 Update build.rs.
- 2015-03-26: wrote rust-mozjs PR #138 Update build.rs.
- 2015-03-26: reviewed most of r/4250 Support setting prefs in wptrunner via ini files
- 2015-03-26: reviewed r/4378 Suppress unused import warnings
- 2015-03-26: reviewed part of r/4348 Implement Element.outerHTML setter
- 2015-03-26: reviewed r/4376 Add tests for HTMLMarqueeElement
- 2015-03-26: wrote rust-mozjs PR #139 Update mozjs_sys to its new name.
- 2015-03-26: wrote servo PR #5383 Update js.
- 2015-03-26: reviewed r/4385 make no_jsmanaged_fields not require imports
- 2015-03-26: wrote servo PR #5388 Stop using old_io in net tests.
- 2015-03-26: wrote rust-mozjs PR #140 Correct a hyphen I missed.
- 2015-03-27: reviewed bug 1142497
- 2015-03-27: reviewed rust-core-graphics PR #34 Use libc from crates.io
- 2015-03-27: reviewed r/4398 Add space in mach command suggestion.
- 2015-03-27: reviewed r/4392 Cleanup and modernize script::dom::element
- 2015-03-27: wrote web-platform-tests PR #1688 Add a test for XMLHttpRequest.send with a URLSearchParams argument.
- 2015-03-27: wrote servo PR #5413 Simplify URLSearchParams::serialize's percent-encoding.
- 2015-03-27: landed bug 1145553
- 2015-03-27: reviewed bug 1124091
- 2015-03-27: reviewed r/4404 Changes for better servo integration
- 2015-03-27: wrote servo PR #5416 Remove some int/uints.
- 2015-03-27: wrote servo PR #5418 Cleanup contenttest.
- 2015-03-27: reviewed some of r/4348 Implement Element.outerHTML setter
- 2015-03-27: reviewed r/4404 Changes for better servo integration
- 2015-03-28: reviewed csswg-test PR #750 Add missing quotes in script src attribute
- 2015-03-28: reviewed r/4398 Add space in mach command suggestion.
- 2015-03-28: reviewed r/4404 Changes for better servo integration
- 2015-03-28: commented on r/4411 Add Accept header to HTTP loader
- 2015-03-28: reviewed r/4412 script: Optimize JS rooting to not move the entire
Rootstruct from the stack to the return out-pointer.
- 2015-03-28: reviewed r/4414 script: Fix O(n^2) dirty bit marking when setting style on all immediate children of a node.
- 2015-03-28: reviewed r/4250 Support setting prefs in wptrunner via ini files
- 2015-03-28: reviewed r/4410 add readonly pathname function in location
- 2015-03-28: reviewed r/4419 Use box syntax instead of Box::new()
- 2015-03-28: rebased servo PR #5416 Remove some int/uints.
- 2015-03-28: wrote servo PR #5431 Remove some unused imports.
- 2015-03-28: wrote servo PR #5432 Use the new io's stderr handle to write out the hard-fail warning.
- 2015-03-28: reviewed the failures on servo PR #5274 Added check for the existence of for and element attribute when preparing a script
- 2015-03-28: wrote servo PR #5434 Move to new io in util.
- 2015-03-28: reviewed the DOM part of r/4402 Canvas: added arcTo() support.
- 2015-03-28: rebased servo PR #4620 Implement a replacement for Vec<Root<T>> as servo PR #5436 Implement a replacement for Vec<Root<T>>
- 2015-03-28: addressed a comment on servo PR #5436 Implement a replacement for Vec<Root<T>>
- 2015-03-28: wrote servo PR #5437 Remove some int/uints.
- 2015-03-29: reviewed r/4419 Use box syntax instead of Box::new()
- 2015-03-29: wrote libfreetype2 PR #8 Update build.rs and travisify.
- 2015-03-29: rebased servo PR #5404 Use usize rather than u64 in MemoryReport. as servo PR #5441 Use usize rather than u64 in MemoryReport.
- 2015-03-29: reviewed r/4352 Fix more gonk warnings
- 2015-03-29: rebased servo PR #5342 Fix more gonk warnings as servo PR #5442 Fix more gonk warnings
- 2015-03-29: wrote servo PR #5443 Use more usize.
- 2015-03-29: addressed a comment on libfreetype2 PR #8 Update build.rs and travisify.
- 2015-03-29: wrote servo PR #5444 Fix gonk warnings.
- 2015-03-29: reviewed r/4437 Fix useless assert in script::dom::element
email@example.com 2 days ago
- Added servo shell project.
- Support keyboard events and focus for iframes and servo-shell.
- Fix root window size being passed to pipeline when clicking links.
- Support iframe navigation updating the parent iframe subpage id.
- Add mozbrowser events for location + title change.
- Glutin fix - Set WM_CLASS based on window title.
- Glutin fix - Add Expose event handler which is needed on some window managers.
- Review css transitions and some other PRs.
- Opened tracking bugs for react.js support.
- More testing and functionality for servo shell.
- Try some ideas for fixing reftest race conditions.
firstname.lastname@example.org 4 days ago
infra work for ME. Doing upsampling of ME ref image right before sub-pel refine. Currently matched to master with < 0.3% bit rate increase and the reason seems the abscence of padding for 1x padding, while 2x ref image has it. This work will make plugging in SATD simpler. Codes checked in yushin/no_sad_exp01 (Note : sub-pel refine disabled in od_mv_est() ).
Continue moving upsampling of ref image for ME, then plug in SATD metrics.
email@example.com 4 days ago
- Health Report: ready for merge
- Dashboard Builder: front-end functionality beta-ish
- Dashboard Builder -> beauty pass
- Dashboard Builder -> server API tweaks
firstname.lastname@example.org 5 days ago
- Bug 1129786 - Unlazify initialization of some PLDHashTables. Simplifies some code, thanks to PLDHashTable now being automatically lazy.
- Bug 1146216 - Remove MakeSlotArray(), because it's dead.
- Print memory reports in tree form. Nicer and easier to read that way.
- Move profiler code from
utilinto a new crate
- Rename lots of profiling-related things.
- Reuse the stack array in Frame.reset(). A small memory churn reduction.
- Avoid creating intermediate strings when generating CSS color values. Another small memory churn reduction.
A month of vacation.
email@example.com last week
- bug 1144481 - Prototype state-mirroring machinery to reduce dependency on the MediaDecoder monitor
- bug 1144486 - Prototype state-watching machinery to reduce notification bugs in media code
- landed: bug 1144487 - Create a class to abstract out the semantic differences between nsIThread and MediaTaskQueue
- landed+backedout: bug 1142336 - Unify state machine and decode thread pools and enable parallel state machine execution
- bug 1144491 - Disable skip-to-keyframe logic during regular automation and test it separately
- bug 1144519 - Rename and streamline media threading assertions
- bug 1144991
- bug 1145203 - Stop synchronously draining the decode task queue when spinning down the MDSM
- landed: bug 1136873 - Use MediaPromises for metadata decoding and fix up threading model
- bug 1145686 - Enforce that MDSM state changes only happen on the state machine task queue
- bug 1146482 - Remove the MediaDecoder monitor
- bug 1146485 - Use MediaPromises for resource / CDM waiting instead of NotifyWaitingForResourcesStatusChanged
firstname.lastname@example.org last week
- Fix any fallout from incremental compacting
- Learn about ES6 modules
email@example.com last week
- Worked on diagram of v2 pipeline with whd: https://mana.mozilla.org/wiki/display/CLOUDSERVICES/V2+Data+Pipeline
- Started writing up onboarding document (for services): https://mana.mozilla.org/wiki/display/CLOUDSERVICES/Cloud+Services+Metrics+Onboarding
- Finished accounts dashboard
- Assorted other last minute tasks related to exec dashboards (mailing list setup, marketplace dashboard changes, review sessions, etc.)
- Caught up with loop team at reporting meeting
- Finish onboarding draft doc & get revew
- Meet with Edwin & IT about metrics/dashboarding roadmap
- short week, pto
firstname.lastname@example.org last week
• Finish up on Version 1 of Reading List / Reader View
• Start working on Version 2 of Task Continuity and Reading List /Reader View v1.2
• Hello Design needs
email@example.com last week
- Bug 1077104 (atomics for effect - optimization) - landed
- New features, notably atomic operations on float32 and float64 (nearly done)
- Spec work
firstname.lastname@example.org last week
Updated iOS UI with:
email@example.com last week
firstname.lastname@example.org last week
- reviewed, landed, and fixed fallout from landing Spark Scheduled Jobs for Telemetry
- reviewed and landed a bunch of PRs for data-pipeline
- Tested a different fork of the
goamzAWS Library for go to enable partial S3 downloads in Go
- Implemented a prototype of reading specific records from S3 files.
email@example.com last week
- bug 1143540 (cubeb crash) opened
- bug 1120331 (ipc crash) crash queries
- bug 1143958 (shutdownhangs) opened
- bug 1143968 (cubeb crash) opened, investigated
- bug 1143972 (dom crash) opened, investigated
- bug 1123511 (a11y crash) crash queries
- bug 1124206 (necko crash) noise-making
- bug 1123465 (writecombine crashes) queries
- bug 1136968 (dll block) wrote patch, landed
- bug 1138794 (xp crash reports) landed, uplifted
- bug 1131180 (js crash) investigated
- bug 1137716 (gfx crash) adapter queries
- Part-time this week and next
- Mostly looking at major crashes and needinfo
firstname.lastname@example.org last week
- Finish landing bug 1137180
- Maybe look at removing BaseShape::metadata, if someone else doesn't do it first
email@example.com last week
- Servo: rustup rustup rustup
- Host a Paris Rust meetup
- Servo reviews
- Successfully moved into a new home in Paris
firstname.lastname@example.org 2 weeks ago
*Correct reference frame tracking in encoder and decoder *Code MV reference list *Simple MV reference index coding
email@example.com 2 weeks ago
- Chased profiler regressions (LUL failing); turned out to be bug 1132586
- Looked a bit at bug 1139386 (jemalloc crashes on EC2), failed to be useful.
- More Fx/V mochi runs.
- Started work on reinstating LUL (native unwind for SPS).
- bug 1022583 (enable LUL by default on desktop linux). Refresh. Re-consider xpcshell test failures.
- Conclude cannot fix it properly with the unwinder thread in place. Make plan to enable LUL without unwinder thread (bug 1141712). Start to implement. Initial run successful.
- V: quite some patch reviewing and coordination: ubsan cleaning, icc as compiler, stack seg 344139, sigaltstack extension 344886.
- V: bug 345125. major overhaul of register allocator and representation of registers. Reduces reg-alloc time by about 1/3 and gives up to 7% memcheck speedup in very JIT intensive situations. Like starting Fx on V. x86/amd64 only right now.
- More LUL work -- tidy up and nuke all traces of the infamous Unwinder Thread.
- V: fix up remaining backends to use new reg-alloc framework.