Summary
-
The XPToolkit team resolved 18 bugs in the last week. For details,
see our
resolved
bug list.
- Reviewed embedding requirements documents to try to determine what
is needed, but have more questions than answers. Will meet this week with
PMs and embedding customers to solidify them.
- Will meet with Jud's team this week to get braindumps and handoff of embedding responsibilities.
-
Special recognition for Michael Kaply, who identified the cause of Bug 56337,
apparently by backing out checkins one by one until he reached the culprit
from October 1.
Highlights
- Mike Pinkerton (pinkertoni)
- more work on the never-ending embedding DnD task
- Chris Saari (saari)
- had several meetings discussing image lib, necko/the new cache design,
and plugins
- Daniel Matejka (danm)
- My horrible Nemesis Bug 56337 has been tracked down and understood, though no fix that isn't itself a
regression has yet presented itself. But knowing that much is a
relief.
- Eric Vaughan (evaughan)
- worked with Pierre and Mark on brain storming ways of making the style
system smaller. Did some protyping. Meeting tomorrow to discuss ideas
and a plan of attack.
- Stuart Parmenter (pavlov)
- David Hyatt (hyatt)
- worked with pinkerton on drag & drop embedding issues
- Dan Rosen (dr)
- more xslt and xpath preliminaries.
- Brian Ryner (bryner)
- Tried to help pink on mac mousewheel code
Lowlights
Priorities
-
pinkerton:
- land a bunch of code, and work with hyatt on embedding DnD
- saari:
- investigate making plugins play nice with events
- more image lib design work, actually dig into the code/bugs
- hyatt:
- evaughan:
- Come up with a fast small solution to make style smaller. And make it
SO. :-)
- pavlov:
- danm:
- dr:
- nail down a plan whereby i can do my work for the next few months.
organize with heikki based on xml/dom team's priorities.
- triage current bugs according to embedding priority, find something
that i can hack on for embedding.
- bryner
- Bugs, especially longstanding mousewheel ones. Hope to
knock out several of them at once with a partial rewrite.
Issues
Escalation
People
Highlights
- Identified printing as large memory consumer since it recreates all
of the layout objects (frames) when the document is printing.
- Verified that the backbuffer is generally allocated using VideoCard
memory on WIN32 so reusing that memory for layout will not work.
Lowlights
Dogfood
Risk
Features
Escalations
Issues
People
- Kevin will be on Jury Duty starting 12/4/2000 (Will be out for the
rest of the week if selected)
Priorities for next week
- Don Cone
- SMIL teleconference
- Work on Frame/crashing problem for printing. Bugs 47478,
56146,58900, 59006
- Rod Spears
- bug triage
- Investigation of form control bloat.
- Waqar Malik
- bug triage
- Crasher bugs related to Linux printing.
- Eric Pollmann
- Bug triage
- Prioritizing bugs
- Kevin McCluskey
- Bug triage
- Prioritize and schedule tasks.
Accomplishments
- Don Cone
- Looked into the printing crasher bugs.
- Worked on general and 32 and 16 bit image support for printing
- SMIL meeting and HP/Gateway meeting.
- Rod Spears
- Investigated overhead related static id's vs. NS_GET_IID. Wrote app
to remove static id's and replace them with calls to NS_GET_IID.
- Removed unused code in form control frames. Ready to checkin, waiting
for reviews.
- Waqar Malik
- 4639 Printer dialog needs to be transient
- 17141 printing doesn't respect "quiet" param
- 14715 Assertion failure: me == CurrentThreadId(), at jslock.c:699
- 45251 test9.html does not display properly
- Eric Pollmann
- Kevin McCluskey
- Vacation - Wednesday.
- SVG weekly teleconference
- Discussed approaches to shrink style structures within layout.
- Worked on document which describes how to add SVG to mozilla.
- Investigated memory usage of images on WIN32. Determined that weshould
always attempt to keep images as DDB's (Device Dependent Bitmaps)on WIN32
since DDB's are often allocated using Video card memory as opposedto (DIB's)
Device Independent Bitmaps.
- Discussed approaches to shrink memory usage when printing with Steve
Clark and Don Cone.
Summary
Highlights
- pnunn
- Bugs reviewed: #60789
- Bugs worked on: #55997
- Scanned new bugs for duplicates, WFM and accepts.
Ordered bugs into major food groups to create meta bugs. There will be about 18 meta bugs.
- Attended Cache meeting on Tuesday and Wednesday.
- Holiday on Thursday and Friday
- darin
- Triaging bugs.
- Cache design meetings.
- Landed fix for bug 45747.
- Received windows machine!
- neeti
- Participated in the cache design meeting for the new cache.
- Worked on publishing diagrams showing the interaction of the various HTTP objects.
- gordon
- Extensive Cache Design Meetings (Tuesday, Wednesday)
- gagan
- Cache meeting to hash out the cache cleanups.
- status/reports up and running! Currently being used by 17 people and growing! :)
- wrote an FAQ for status/reports tool.
- Working on expanding the project plan for Necko and Imglib to include time swags.
Lowlights
Issues
Plans for the next week
- pnunn
- Create meta bugs. Prioritize meta bugs.
- Have identified many bug duplicates and need assign as such.
- neeti
- Work on publishing the diagrams for the current state of HTTP.
- Bug triage of HTTP bugs into various categories.
- gordon
- Make first pass at new Cache interfaces.
- Hold next iteration of Cache Design meeting.
- gagan
- Changes to status/reports
Summary
- We are focused on stability, bloat and memory leak reduction and performance
work.
- The XML/DOM team resolved ~15
bugs last week (bugs not assigned to us do not show up).
- Triaged Clayton's bug list.
People
- Nisheeth will be on vacation from November 20th to December 17th.
- Johnny will be on vacation from December 18th to January 3rd.
Accomplishments of last week
- Peter Van Der Beken
- Harish Dhurvasula (harishd)
- Reviewed Vidur's and Rickg's code.
- Tom Pixley (joki)
- Heikki Toivonen (heikki)
- Worked on 59674, compiler warnings about uninitialized variables.
- Worked on 49572, implement responseText property (XMLHttpRequest object).
- Worked on our team's task list.
- Triaged Clayton's bugs.
- Johnny Stenback (jst)
- Worked on reducing the size of the in-memory and on-disk size of the
content model.
- Went through part of my buglist to prioritize the bugs.
- Nisheeth Ranjan (nisheeth)
Plans for this week
- Harish Dhurvasula (harishd)
- Prioritize my bugs ( Crash,MLKs,Correctness )
- Triage my bug list
- Heikki Toivonen (heikki)
- Triage and set milestones for bugs on Nisheeth's list.
- Work on our team's task list.
- Check in the fixes I have in my tree.
- Johnny Stenback (jst)
- I'll continue with what I did this week and also start looking into
converting the DOM code from using IDLC to using XPConnect.
Lowlights
- Peter Van Der Beken does not have a computer yet.
Escalations
Mailnews Accomplishments
- Resolved as fixed
13 bugs in Bugzilla.
- Jean-Francois Ducarroz did some performance work on replying to a message
andposted the results to n.p.m.mail-news.
- Gayatri Bhimaraju made it so there's feedback when there are no search
results.
- Navin Gupta made it so that folder-only-folders don't display an error
dialogwhen selecting them.
- Seth Spitzer removed a lot of bloat from the subscribe dialog.
- Blake Ross removed mailnews' dependency on the browser for themessage
pane context menu and made File | New Message work in the composewindow.
Profile Manager
- Bhuvan Racham is going to hand over ownership of the Profile Manager
code to Conrad Carlen. They'll be working together to make sure there's
a good transition and Bhuvan will be around for consultation.
We continue to concentrate on bringing the OS/2 version of Mozilla up to
par with the other platforms.
Main areas of focus lately have been printing and font rendering.
We feel like we are making pretty good progress.
In addition, we have begun the major task of removing XP_PC defines from
the Mozilla build.
We have made a Win32 Bidi Mozilla
available.
We have the Bidi code merged to the tip and are still waiting for approvals
to check it in.
Bugzilla: implemented a script which kills runaway threads in order to
prevent the database from getting out of sync or horking. Seems to be
working nicely so far.
Despot: checked in a patch from zach@zachlipton.com turning off
deletion of users, as this could leave dangling pointers. Disable user can
be used as a workaround.
Previous Updates
|