Calendar:Status Meetings:2007-10-10
From MozillaWiki
<< previous meeting | index | next meeting >>
Meeting Details
- Wednesday, October 10th, 16:00 UTC
- Phone meeting
- Join #calendar-mtg on irc.mozilla.org for attendance
Telephone Info
- Toll free numbers
- US/Canada: 866-692-3163
- Netherlands: 0800-020-1392
- Germany: 0800-000-3441
- Participant Passcode
- 3182189
- Conference controls
- Press *1 private help menu
- Press *6 mute or un-mute individual line
- Each time you say something, please say your name first so people know who's speaking
Agenda Items
- Chat with David A about Lightning/Thunderbird/MailCo
- 0.7 RC Status
- When is code freeze for RC2 scheduled?
- Need to move forward the *_RELEASE tag on all changed source and l10n files.
- 0.7 Status
- status of blocker bugs
- l10n status
- Should we exclude mn locale from release?
- What to do about localizations with issues (eu, pa-IN, tr)?
Meeting Log
- Attendees: mickey, ulf, stephan schafer, joey minta, martin S., clint, david A, simon, Philipp, dmose
Calendar Chat With David A
- What is the current plan for 1.0 - what is the feature set
- Interoperability
- Timezones
- Meeting support
- Free-Busy more coherent support for ICS, CalDav
- Improve CalDav support
- See Calendar:Roadmap for details
- What is the timeline
- (simon): plan is to hit 1.0 in 2008. We are about to release 0.7 in the next few weeks. But we are going to put in one more release - 0.8 to tie up the loose ends
- Then a 0.9 as a polish and fit and finish release
- Expects 0.8 at start of 2008, April/May: 0.9, and June/July/aug 1.0
- What can DavidA do to help? In the long term, obviously he can't do much until he gets up and running himself. We have the most ambitious extension on top of thunderbird, and he wants to understand where the architecture of Thunderbird isn't quite right for extensions and what can be improved with that?
- We will think of ways that "MailCo" can help us
- One thing that is hurting us is that the entire extension architecture is based around browsing. People try to install Lightning and they try to install it in Firefox.
- Why not use a different mime-type for Tbird extensions vs. Ffx extensions.
- It's something that is done in the FFx code and there has not been enough will to overcome the inertia to change it.
- Poor addressbook integration comes up in terms of features that people complain about. And it becomes more an issue with calendar and mail together.
- There are several extensions in that space that add needed/additional fields
- LDAP write support for Thunderbird
- DAV support adjustments might be needed to platform library -- ask Bruno about this
- All documentation for extensions is focused on doing the browser extensions, nothing geared for thunderbird or calendar or any other product.
- Makes it difficult to develop ext's for Tbird
- What has been lacking in the past was communication from the tbird team. Look at how much MOCO employees blog about what they are doing, and then compare that to what you hear from the Thundrbird devs
- For example, when we started blogging on the calendar blog, we really started to generate a lot of interest around the program.
- Communication is going to be incredibly important, and it's going to be even more important for tbird to get people outside of "mailco" to care about the project and contribute to it.
- David A. needs referrals to people who might want to be mailco engineers, and is looking for leads
- Simon knows of a few people working on the mail/news side of seamonkey and it might be good to start there.
- Another idea, he should look at people doing low-level extensions and try to recruit people there.
- David A wants to figure out the right way to do user experience and UI in an open source project
- "Lightning has gotten much better over the last three releases" (direct quote from David! Yay us!), and he wants to know where to give feedback
- He wants to know how we've been doing this, and how we can better make changes with the Thunderbird UI changes, and what changes in Thunderbird could be made to better accommodate the user experience we are striving toward.
- Simon - We can put you in touch with Christian who does our UI reviews and design evaluations
- Christian returns from vacation next week, and we will have him contact you.
- We have no thoughts right now on how to change the Thunderbird UI, since we've been moving in the direction of complete independence from the Thunderbird UI.
- One thing that would make every extension developer's life easier would be to open up the "Folder pane" to easier integration. Several of us have been defeated when trying to push integration into that area of the Thunderbird UI.
- David A: What is the architecture of lightning/sunbird?
- It is based on a C iCalendar library that called libical that is used by us, Evolution, and (we think) KDE. Very little is done to maintain it, but it might be reasonable that if changes need to be made, we could leverage support from those other two organizations to help with those changes.
- Above libical, we have a series of C++ XPCOM interfaces that serve as a wrapper to libical to implement things like Timezones, Calendar Components, Recurrence rules, DateTime objects etc.
- Above that, specific objects are implemented in JS, these are: Calendar Provider (Server) Interfaces, Event objects, Task Objects, Recurrence Handler objects, Views, etc etc.
- The rather thick JS layer allows people to easily plug in and make substantial contributions to the project without having to really understand the intricacies of XPCOM and C++ development. (which is generally not the case in Thunderbird).
- MVL couldn't be here today, but sent me a question to ask: (ctalbert reads mvl's email:) "I don't have many questions. The only thing I would like to find out is how much interest there is in "MailCo" to support us. And with support I mean things like paying to send people around the world for calconnect, calendar face-to-face etc, keeping tinderboxen running, and stuff like that. I think that's more important then hiring a developer. But maybe it's a bit too early to talk about that."
- Not too early to talk about it
- It's important to have a vibrant calendar community with a Thunderbird component.
- It's important to do the right thing for calconnect and organizations like that and until DavidA has people to send to that, he will send us
- Generally, he'd like the two projects to get closer together and not farther apart. His tinderboxes (which he doesn't have) are (will be) ours as well, as far as he is concerned
- The tinderbox support is really important - because right now the tboxen etc are moving farther away from the core, and shoved into the community projects, but there isn't good support or know-how in the community to administer those systems.
- A lot of this issue has also been IT misunderstandings as well.
- Is having a build person who is not a software developer a good person? You want someone to engineer a better system overall, not just fix the current problem.
- The initial tinderbox stuff was done by a mishmash of build and software engineers, and it worked, but the code was unmaintainable.
- The build team here has been doing build auotmation for a long time.
- David's getting more to the point that he meant that we need to hire people that will fix the issue forever versus just fix it for now.
- It would be great to have somebody for Ause to talk to in order to help get issues resolved.
- SSA: What are your future plans for "MailCo" and your long term goals? When do we expect the next thunderbird release?
- Next release will be the point release for the security stuff on the 1.5 product line (1.8.0 branch).
- There is no roadmap in place yet, DavidA is still in process of finding out what needs to happen
- DavidA needs to understand what is currently going on, how thunderbird is being produced w.r.t. build/qa/addons etc and to establish contact with all those people and figure out what the right transition process and timeframe is.
- Also, he needs to build the core team in parallel, specifically, he's looking for:
- an architect who understands what needs to change and how to take an incremental process to change things
- front end - xul / js/ ui talent - tbird needs love there.
- QA Coordinater to build a vibrant community surrounding thunderbird QA.
- IT/Build - not sure what the boundaries are there
- And...he still needs to set up the organization - legal and accounting issues to figure out. Like...What is its *real* name?
- Also, He doesn't want to allow thunderbird 3 to be released without a team to support it.
- What will the revenue model for this "MailCo" be:
- No answer yet.
- It's only his second week on the job. With the granted funding from the Mozilla Foundation, he has time to figure it out.
- He doesn't want to come up with a premature plan that pigeonholes Thunderbird into an inconvenient niche that might not be the right answer for the users. He would rather evolve Thunderbird into a more interesting product, build a community around it, and then worry about how to make money.
- LibMAPI
- Two guys out of France who have written a MAPI implementation that can talk to an exchange server and communicate email, contacts and calendar information.
- He's been in contact with them. They are fascinating group.
- They have a library that talks mapi that you can talk to an exchange server so that the exchange server thinks that you are outlook.
- DavidA's question: How flexible is the calendaring infrastructure? Would it be easy to include something like this into Lightning/Sunbird?
- Our answer: It's very easy, all in javascript, and we specifically designed the calendar system to be very pluggable in this regard so that people could extend the product as needed. Just look at what Philipp was able to do with the Provider for Google Calendar and the Sun guys with the Sun WCAP Provider.
- Why isn't the Google Calendar stuff built-in?
- Google calendar is on the list of things to integrate once it is ready
- Historically we just wanted to limit the product to only have built-in providers for standards-based calendar servers, and that is why it was originally slated to be an extension
- Philipp: Really wants to have recurrence support before integrating.
- Simon: We've thought about evolving that historical perspective because there has been great interest in the Gdata extension.
- Feel free to email him whatever issues that we have. Especially regarding issues that he can move forward.
davidascher at gmail
Status Call
- RC2
- No major bugs, quiet test day
- Still issues with some localizations
- 3 l10ns that are having issues
- mn seems to be dead - we will exclude it from RC2 as well.
- 2 l10ns: basque and turkish simply don't have default preferences yet. No default categories is the biggest issue there
- punjabi language about half of their strings are in english. - Should we actually release a l10N like that?
- Filed a bug about this, sent a mail to them
- Posted again in L10N newsgroup
- Maybe we can leave it this way but just have them leave the strings in english.
- Maybe we can release it for 0.7, or 0.8. But definitely not for 1.0
- Simon will add this to the release notes and we'll release it.
- German localization also has very odd/creative wording
- We haven't been able to contact Thornsten
- We don't want to go in and fix it ourselves because it is the Localizer's call and the Localizer's job to do so.
- We will create a patch with the changes to these strings and ask for Thornsten's review.
- Hungarian issues were fixed today
- Macedonian bug was filed by Simon, and he thinks we should wait a few days for them.
- It's a very small fix. Localizing one line for default categories
- Code Freeze for RC2
- Announce a code freeze for it
- Simon will blog about this
- Friday midnight UTC should be the codefreeze, and that gives Ause two days for the tinderbox run. On Monday he will create the localized versions.
- We will smoketest them on Tuesday.
- Updates
- Ause has been in touch with Preed.
- There is a lot to do and that work might be prohibitive for 0.7
- ctalbert: Follow up with preed and see where we are at the moment with regard to updates
- We might release 0.7 using the old AUS system.
- 0.5 will update to 0.7 using the old AUS
- 0.7 will update to 0.8 using the new AUS
- Roundtable
- Is it necessary to move the _RELEASE tag? Ause says no, not for this release because he is releasing it from a 0.7 branch.
- RC2: we will wait for the localizers and then we will be all set for the release. Everything else is in place.
- Simon will blog about the code freeze.
- Sun Marketing guys are preparing something for the heise.de guys.
- Ctalbert will do an english press release.