This week’s Random Notes

• Running MozMill tests for Thunderbird.
• Thunderbird:Start Hacking
• window.dump

Status Updates

Zach Church

I've got some Mozmill tests working, and my patch is ready; I just submitted it.

Unfortunately, I didn't read Blake's link about pdiff soon enough. I had been making commits to the default branch in my repository instead of the pbranch, so I had some troubles exporting a diff. In case anybody else runs into the same problem, here's what I did:

http://zgchurch.com/2010/01/28/mercurial-pdiff.html

I'll post next week when my patch gets rejected, so we can all see common reasons for rejection ;)

Tim Miller

So I woke up on Saturday and headed in to school, set on having a productive, useful day and make some progress on my bug. (Adding support for new taskbar icon functionality in Windows 7 to the Mozilla base library.)

However, I could not get Thunderbird to compile. The error was related to a data type defined in the Windows 7 SDK, and was in a file i had not modified. So I began my quest to resolve it.

One of the first things I did was to check the SDK version windows was set to use. It was set to v6.0A. I have no idea why it was set to that, but it was wrong so I fixed it. Still didn't compile though.

Fast forward six hours. Still not compiling and I'm pulling what's left of my hair out in frustration. I was talking to sid0 on irc though and he mentioned a useful bit of information. The console checked the SDK version on startup, so it had to be restarted if the version was changed. Which I did six hours ago. Sure enough, restarted the console and it worked.

So pretty much my whole Saturday was wasted on a something that took a total of 30 seconds to fix. But at least it is compiling now. Now I can write some tests and submit a patch so I can get back to working on my original bug. (Making an the Windows 7 taskbar icon for Thunderbird show unread messages.)

Marcel Guzman

For some reason, three classes of mine decided to make an essay due during this week (yes, seriously :(), so I didn't have time to hop on IRC and talk to the people I needed to talk to about testing the activity manager changes I've made. I did manage to steal a few minutes or an hour here and there however, and I spent that working on getting folder compacting to log in the activity manager.

So far, I've figured out how to add a new "Activity Module" and I've tried hooking that up to the compacting events that get fired. This currently doesn't work so I'm working on my gdb skills to try and figure out what's wrong.

Kefu Zhao

This week I am still working on bug 461669.

First of all, I spent time studying some header definitions in the source of an email because I need to decide which “From” address to use when user clicks “reply to list” base on different headers shown in the email. As a result, making clear of where and what headers are added to the email is very important. These headers include “Delivered-To”, “Received… by… for”, “List-Post”, etc.

Then, I tried out TB to reproduce the bug and found a problem that I am not sure which “From” address is supposed to use (from the perspective of user requirement). Currently I have to clues:

1.used the account which the email is associated to regardless of headers in the email.

2. try to read the headers and see whether can I decide which “From” address to use.

Next week I am going to talk to Jens about this bug.

Lindauson Hazell

This Week 01/24/10 - 01/30/10:

[Bug 495815]

I'm a bit puzzled... I've gone though the mailWindowOverlay.xul and added in the missing command elements associated with the incorrectly enabled menu-items.

Sadly the changes in the script don't seem to have any effect on the incorrectly enabled menu-items.

There is a comment in the code that mentions a utilityOverlay.xul. I'm gathering that it is being used as a temporary bug fix. I think the utilityOverlay.xul file is overriding the mailWindowOverlay.xul.

Next week 01/31/10 - 02/06/10:

Continue woking on [Bug 495815]. Specifically looking into the utilityOverlay.xul file.

Evan Stratford

This week, I continued work on vCard imports:

• Added file dialog filters, completing the frontend work.
• Wrote nsVCardAddress::ReadRecord(), nsVCardAddress::ProcessRecord(); these do the work of reading vCard objects from file and parsing them into nsAbCard objects respectively.
• Implemented adding the newly created nsAbCards to the database - first end-to-end working version!

Wei Xian Woo

Still working on bug 408338. I think I made pretty good progress this week.

For my first patch, I intended only to support periodically saving session state so that we can restore a user's session in the event TB crashed and deferred the other requested features to subsequent patches. However, I ended up finishing off supporting saving of session state for multiple 3-pane windows as well. Details:

• For context, the 3-pane window is the "what you think of when you think of Thunderbird" window.
• Prior to this, we only supported persisting the state of the last 3-pane window that was still active at the time Thunderbird shuts down CLEANLY. If TB crashed, we feel bad and start up cleanly next time for the user.
• Firefox restores state for all windows only if Firefox crashed. On a clean shut down, state is persisted only for the last window.
• In an effort to keep the same behaviour across Mozilla apps, I implemented TB's new Session Restore Manager to imitate the session saving/restoring behaviour in Firefox.

So, what's next?

• Tests for patch 1!!!
• Richer state!
• Support other window types!

What happened to that Python script I said I was going to post on my blog?

• I stopped blogging when I came to Waterloo, so I don't actually own a blog anymore. For the past week, I have been working on a WordPress theme for my new blog. I expect the script to be posted by the end of next week. Sorry for the delays!

Who’s writing the blog post this week?

Tim Miller