maintained by Tim Rowley <firstname.lastname@example.org>
Last Updated Wednesday January 3, 2001
This status update page is updated every
week. To get updates and news throughout the
week you are invited to check out mozillaZine
a site devoted to Mozilla advocacy.
Friends of the Tree
- Brian Ryner and John Myers for excellect tree cleaning work. They both killed a ton of unused code. - Dan Mosedale
- David Baron for profiling and code cruft removal. - Stephen Donner
Missing Status Reports
The following engineering groups failed to
submit a status report by the Wednesday
6:00pm EST deadline.
GTK based plugins should handle focus properly now. This means that if
you want to want to write a plugin you can do it in a toolkit.
Here are minutes from the 12/20 footprint meeting (better late than
never). Please look through this to remind yourself what you were
supposed to do and reply to this message with a quick status update.
Then everyone can be up-to-speed and we can just go through questions
and problems. (Tip o' the hat to buster for the pointy-haired meeting
- Ongoing memory tracking (curt)
- Builds were stable (as of 12/20) and we were expecting to see some
regular reports posted. Looks like things may have degenerated
- Still problems with combo boxes crashing embed build. Couldn't find
a bug filed on this, although I'm pretty sure there is one. (rods?)
- winEmbed crashing with `embed' builds?
- Code coverage
- evaughan did
- discovered amount of unused code in layout DLL == sum of unused code
in all other DLLs. jst is doing some DOM refactoring; should re-
measure after his work lands.
- Style system
- attinasi's style context sharing has been checked in.
- attinasi/pierre/evaughan discussing doing sharing at finer
granularity, sharing CSS declarations, selectors. pierre has
been collecting statistics that point towards this. (pierre,
maybe you could share your info with the group
- Trace malloc porting
- mjudge still working on it
- Allocation analysis
- brendan/waterson discover race condition bug that was causing bad
data to show up. Still can attribute only 20-25% of growth to
- jar/thesteve to investigate fragmentation.
- kandrot to instrument sbrk() and mmap().
- buster to look at frame manager structures
- cls to determine if compiling mozilla with gcc-2.95.2 requires
libc upgrade to use resulting binary.
- cls determined that -O2 and -O3 do not make significant difference
with respect to overall DLL size.
- pavlov to investigate new gcc code-gen rumors
- chofmann to bug waldemar about commercial compilers we could try
- harishd decides parser nodes cannot be removed. Will change to
arena allocation of nodes and fix leaks.
- thesteve to begin using `--enable-boehm' build to track down
- Project Fu
- waterson to set up space on gila/website for coordination.
Our two main areas of focus of late have been internationalization
(particularly DBCS) and printing. We have made some big strides with both.
We are looking at putting together a Mozilla 0.6 very soon that will
Bidi is currently in the process of being reviewed. See