Title: Necko, Imagelib and XPCOM's status
submitted on 2001-10-08
Necko, Imagelib and XPCOM's status
submitted on 2001-10-08
| Summary |
- gagan
- http://status/bugs/ is back online. Please file bugs for it under any component you like :-P but make sure they are assigned to [EMAIL PROTECTED]
- darin and pavlov fixing away bugs.
- dougt has been helping out with MAPI issues. Thanks Doug!
- triaging continues for the group (lots of bugs targeted for 0.9.6)
- untargeted bugs are still high-- we may need a group bug-triage session to weed them out.
| | Highlights |
- darin
- fixed bug 94004 "Offline: redirects not cached"
- fixed bug 102332 "9K leaks on tinderbox: cycle relating to HTTP handler"
- landed initial framework for http/1.1 pipelining (bug 93504)
- checked in "iDNS support" patch, bug 42898
- triaged a bunch of bugs
- worked on the mach-o OSX port
- neeti
- Worked on bug 96461, ComponentManager could PLHash or PLDHash. Have a patch ready, waiting for reviews.
- Triaged bugs.
- gordon
- Bugs closed (WFM, INVALID, DUP) 91120, 91198, 95851, 96032, 103328
- Triaged 0.9.5 and untargeted bugs.
- dougt
- Completed first swing at freezing the nsIServiceManager. Waiting of some inital reviews and then I will post a patch. The changes include redefining the interface in IDL, factoring out the helper classes and functionality into a seperate file, removing the nsIShutdown listener, and preserving the old functionality as nsIServiceManagerObsolete. Still some things to think about such as removing (Un)RegisterService*. Details will be posted to the newsgroup.
- There wasn't a test case to enumberate the registered ClassID's, so I wrote a protocol handler that enumerates all of the components printing out information about each of them (about:components).
- Spent most of Friday looking at component unloading. I have proof of concept that modifies the nsISupport and the factory macros to Lock and Unlock the objects' factories. The Modules can query on the status of the factory as a necessary condition of unloading. The one thing that I do not like about nsIFactory's is that in order for them be part of the unloading story, the need to stay resident in memory. Maybe it is a small price to pay?
- pavlov
- Bug 92248 - Flickering on mouseovers. Checked in fix for this on the trunk.
- Bug 103466 - Crash when trying to load images with no source.
- gagan
- http://status/bugs/ is back online with v1.3. Fixed a bunch of bugs with it.
- triaging continues for the team. We have still too many untargeted bugs across the whole group.
| | Lowlights |
- gordon
- So many bugs, so little time.
- dougt
- Spent two+ days on reviewing MAPI support and fixing a known security hole in MAPI for the 0.9.4 branch.
| | Plans for next week |
- darin
- land some (hopefully all) of the changes for mach-o on the trunk.
- get back to working on freezing xpcom/io interfaces.
- triage more bugs.
- neeti
- Continue work ComponentManager cleanup/bloat.
- pavlov
- Play bug triage.
- Fix bugs!
| | Issues |
- gagan
- access to a machine outside the wall is still needed for publishing, hosting necko tests scripts and perf results.
|
|
|
