Firefox2/StatusMeetings/2006-04-04
From MozillaWiki
<< previous week | index | next week>>
Meeting Details
- 11:00am PDT (18:00 GMT)
- Mozilla HQ, 1st floor conference table
- +1 866 216 2181
- join irc.mozilla.org #bonecho for attendance taking
In Attendance [edit]
schrep, linus, ben, annie, pam, rstrong, brettw, jhughes, sherman, darin, beltzner, mconnor, dietrich, markp, gavin, ealden, corneilus
Agenda [edit]
- Development Roundtable
- Any other business
Developer Roundtable [edit]
Add notes about work being actively done at the moment, areas where you'll need help/review, or things which the rest of the team should be made aware of. Leaving the section blank or linking to current discussion threads in dev.apps.firefox is perfectly acceptable.
MoCo Team Update [edit]
The status graph below shows relative workloads, in terms of estimates in days for a2-targeted bugs. Right now, Rob Strong is the long pole due to the dual load of extension manager bugs and the new installer, but we should be able to shift a few bugs elsewhere by the end of this week. From our perspective, we should be ready for a2 by the end of April at the latest.
note: Robert's estimated time is now 15 days (e.g. 5 days less) after moving several of the non-feature bug target milestones to b1.
Visual Refresh [edit]
- not a lot of response from the community
- beltzner has someone in mind to help out
- might also be able to get someone on contract
Bookmarks and History [edit]
- latest swags are for approximately 22 days for A2
- goal is to be feature complete for A2 with the exception of tagging
- beltzner: any chance of tagging make it in for B1?
- jhughes: we're interested in seeing it in there, so we're optimistic there
- shaver: I'm interested in getting some extension hooks in; where should I go for adding them?
- beltzner: there's a requirement to have that for A2
- ben: that's getting built as we're building the infrastructure; post to dev-apps-firefox to discuss specific APIs that might be required
- shaver: if tags is out for A2, shouldn't that be updated on the wiki
- action beltzner to update planning docs, include changelog so it's a living document
- schrep: if we include tagging, then we should be creating A3 so that alphas are unfinished features, and betas are relatively stable and finished products for wider testing
- mconnor: the idea was that we were going to hold the line on features to preserve our ship date, though; I don't think we should be pushing an alpha just to get this?
- schrep: the goal is to not insert features between alpha and beta, and get realistic about the impact of inserting a new feature like tagging after A2 ships
- ben: I just want to preserve the alpha baseline
- schrep: understood, but if you're doing that because we can't fit the feature, then we can't fit the feature without moving our ship date
- tabled for any other business
User Experience: Tabbed Browsing [edit]
In MV last week, patches should start surfacing in the next few days.
User Experience: Session Restore [edit]
- Fixing review issues and bugs
- New patch should be done today
User Experience: Spell Check [edit]
- dictionaries will either have to be bought or extensions
Feeds: Parser [edit]
Feeds: UI [edit]
Extension Manager: UI [edit]
- Extension Manager UI wiki page available
- planning to have a patch submitted by this weekend
Extension Manager: Blocklisting [edit]
Extension Manager: Locale Packs [edit]
Search: UI [edit]
- should be moving into UI space in the next little while
Search: Plugin Format [edit]
Distribution: Installer [edit]
Distribution: EULA [edit]
Distribution: Channel Support [edit]
Infrastructure: Performance [edit]
- might look at a compiler change (this might just be gossip!)
- dveditz: that might end up resulting in library incompatibilty problems
- agreed to take that part of the discussion offline
Infrastructure: Platform Uplift [edit]
On the radar/under investigation:
- JS engine mass uplift for JS 1.7(?) proposal -- need details from brendan and co.
- Canvas performance improvements (vlad)
- Proposal for better hooks for additional canvas contexts, which would need to use non-frozen API vigorously (vlad)
- Some timer fixes, as a subset of Darin's THREADS_BRANCH trunk work; idle timer and queue-service time bounds could be nice and small (vlad investigating)
- Mac event loop improvements (mento)
- this stuff should be in for B1 since it's large enough that we'll need the testing
- need to figure out how to get API hooks in various places for extension authors
- schrep: any word on basic implementation of offline storage?
- shaver: haven't seen anything about that, but if we were going to do it, we'd need to do it soon, since it would have significant effects into the codepath. Seems risky.
- mconnor: is there anyone going through the platform (blocking1.8.1?) nominations?
- shaver: not that I know of
- action: schrep to organize a triage meeting for these
Infrastructure: Instrumentation [edit]
- progressing with client development work
- targetting alpha 2 for first-use
Security: Phishing [edit]
- looking at refactoring the code (for A2) to make it easier for other parts of the platform to get at it
Security: SSL [edit]
User Documentation [edit]
- working with Basil to draft requirements and plan for integrated help systems
- currently in a lot of flux
- when a plan is ready it will get pushed to dev-apps-firefox (this week)
Microsummaries [edit]
Added architectural details to the microsummaries writeup and posted pieces of the writeup to mda.firefox in five separate posts on Microsummary Definitions, Specifying Microsummaries, Back-end Implementation, Datastore, and Front-end Implementation. I'm looking for feedback on all aspects of the architecture. This week I'm going to:
- meet with Places folks to hash out the technical details of microsummaries integration with Places;
- migrate good prototype code into Firefox;
- get cracking on the implementation.
Other Business [edit]
- schrep: want to get some input on where we think a good line is in terms of creating a competetive product
- ben: couple things; first, the targetted places work for A2 should make us competetive, second, the additional value add (full text search, tagging) would make us extremely compelling
- linus: from what I hear, priority seems to be: full text search, the idea of "starring" a page, tagging (I think this is suspicious), scrapbook-feature (ability to grab text/images of a page and store it for offline viewing)
- annie: I've also heard the desire to take notes on/about a page
- shaver: something that's useful for extensions and adds rich metadata is the ability to track the inbound and outbound sources and destinations for links; so I can know I got a link from someone/someapp and sent it to someone/someapp
- sherman: advanced support for detecting and handling microformats
- beltzner: detection and quickpaths of other types of content like media and documents
- shaver: lots of metadata things happening such as RDF/A, micro-summaries/formats; making it easier for extension authors to get at this stuff would make it easier for our browser to be the core of experimentation around that
- linus: what's the one killer feature though?
- beltzner: with the caution that adding features for features' sake isn't the best way to do things?
- ben: agreed; and I think that better search results, including the ability to annotate pages in a lightweight fashion is the most globally useful feature
- linus: so what you're saying is that out of all this, better bookmark search?
- beltzner: I think that in today's web, features that help our users find the bits of the web that they find/found interesting is globally useful
- schrep: are we happy with the A2 targets as our feature set?
- ben: yes, I think so
- brettw: I think it's a good upgrade vector, but not neccessarily a switch vector
- schrep: so what we're saying is a bunch of stuff is better, and some little things are new
- myk: what's the relative priority of making our current users happy vs. leapfrogging IE functionality to draw new users?
- ben: our bookmarking and history story will leapfrog them
- ???: IE's smooth bitmapping seems really nice
- ???: the story when IE7 comes out will be IE7 vs. Fx2, and we need to make sure we have a bunch of things we can point to
- beltzner: and I want to make sure that "we do basic browsing simpler than the other guy" is eligible for that; driving the feature set from a headline-seeking attitude seems wrong to me
- shaver: we are, however, in a competetive environment and need to convince people to switch to us; one thing I'm hoping to do is build a bunch of subcommunities around extensions that are compelling to those communities
- linus: early data indicates users primarily use their browsers for research, so tools for collecting information seem to be paramount. the number one reason for leaving Firefox appears to be apathy - nothing compells them to use the browser, so they are entropically drawn to the blue "e"
- mconnor: I think that there isn't a lot we can do for those users; they don't seem to care/want to make a choice
- shaver: we need to explain to those types of users why they *do* want to make a choice about their browser
- sherman: these types of users are swayed by feature differences
- linus: back to the core question, though, do we think this list of features is compelling enough?
- schrep: actually, what I was trying to get to was: is adding additional features to this A2 list worth taking the impact to schedule? given infinite time and energy, what would we want as killer feature?
- shaver: my wave-a-wand feature would be the support for search across full-text and metadata about the subset of the web that I experienced and how I experienced it
- myk: extract search results for specific data and display them more usefully in the product chrome
- brettw: an A3 would allow us to fit in tagging and Annie's idea about creating notes (which could be searched)
- ben: I think that's even ambitious for an A3
- shaver: there's lots of ways we could annotate entries based on the search queries that led there, as well as words that were found in those pages
- schrep: hitting A2 and the set of features for A2 still seems reasonable; we can explore these other things on a slow-burn now, discussing them in dev-apps-firefox, and then make a call on A3 as we near the A2 deadline
