Internal Affairs

Not dead yet

I just read an article about Mozilla over at eWeek entitled A Fighting Dragon. I would like to thank Jim Rapoza for writing that article. I must say that I am glad that there are people who realize that Mozilla is not dead. Exciting times are on the horizon...

Rules

I landed a pretty cool fix for Hixie the other night. It only took two and a half years for it to land, but hey it's in.

In other news, after a few days of random crashes which were quickly fixed, I haven't seen any more regressions from my principal landing last week. Looks like it's settling down finally.

Blocker Lettuce Tomato

I've gotten my feet wet now, I suppose. I fixed a blocker bug only to find out it caused another blocker the next day. Nice.

Landing

So it's in the tree as of last night, with ~20k reduction in footprint on linux boxes and ~10k on Windows boxes. Go download a nightly. You know you want one.

Yanked

I was reading this week's edition of the SitePoint TechTimes and noticed a little blurb about the end of Netscape. The article goes on to say that Netscape developers will be sorely missed and alludes to the end of Mozilla development. I'd like to remind Kevin Yank about the many ex-Netscape developers who still plan to work on Mozilla in their spare time. Many of the people who worked for Netscape were very passionate about the product they produced, and though this is a set back, there are a good many developers who will not go quietly. Some of the Netscape employees were involved with Mozilla in one way or another before they got hired and I expect them to continue to do so.

Looking at the list of checkins which occured since that fateful day, I count 17 developers who were affected in one way or another and have checked in code to SeaMonkeyAll. Still more have been active in Bugzilla and/or IRC by doing either code reviews or providing thoughtful comments to bugs.

Yes, we will miss Netscape as a company, and we grieve for and will definitely miss those who have lost their jobs. Please do not assume that Mozilla development will come to a halt, though. Don't forget that companies other than Netscape have vested interests in Mozilla's success. And Netscape the company may be dead, but Netscape will live on through what it's created, and through other contributors who are taking up the helms.

Zip-A-Dee-Doo-Dah

So in the past few days, I've started working on Bugzilla again for the first time in nearly 2 years. It's amazing how much the source has changed, but how easy it is to get back into it. It almost looks like it hasn't changed. And that makes the transition back a little easier. I'd still rather be working on Gecko, but this doesn't look like it will be a bad re-assignment at all.

Bounty Hunter

So about two or three weeks ago, someone was on #mozilla seeking help with some Mozilla thing for his company for a small bounty. I don't really know why I answered the call, but I did thinking it was just a desparate cry for help, and it turns out that the bounty hunter was true to his word, which I did not expect. Paypal payment came through shortly thereafter, and today the funds were officially deposited into my checking account. I may have to try this some more....

The last patch

So, the last patch I am working on under the Netscape banner is getting closer to landing. Mitch said that I can land it without his review, provided I get r/sr from other people. He will look at it in depth when he returns from his vacation. In other news, I have r=jst.

Why 2k

Oddly enough, it was exactly 2,000 days between January 22, 1998 and July 15, 2003.

And so it goes

Well, Mozilla and Gecko shall live on with or without Netscape/AOL backing. I will continue working on it in my free time, as will others. I will be pushing in my existing patches as quickly as possible, and working on new ones as well, although at a somewhat lessened pace. Based on my talks with a few reviewers, I expect to land my principal re-write by this upcoming Monday.

Earthquake

Sigh.

Netscape is dead. Long live Mozilla.

Deepest sympathies and best of luck to everyone.

Marked man

David Bradley profiled my nsIPrincipal patch earlier and said that it showed marked improvements in DHTML perf.

Fear Factor

The only thing we have to fear is fear itself. Former President Franklin D. Roosevelt.

Wow. I just saw Bowling for Columbine. Go see it if you haven't. It was a very well-made and thought-provoking documentary. This sheds some insight as to why foreigners don't want to live in America.

What's in a name?

As part of the DOM Level 3 kick I've been on, I'm now working on an implementation of Document.renameNode(). Seems pretty simple right? Think again.

Some of the issues I've run into are:

• retargetting nodes' nsIEventListenerManager and its nsIJSEventListeners
• reparenting element nodes' attributes NamedNodeMap and childNodes NodeList
• transferring ElementCSSInlineStyle
• computed style stores a reference to the content's frame which will need to be updated
• other as of yet unknown things may hold onto similar things which point to the content
• XBL, XSLT, etc. will have to be rigorously tested to make sure that nothing bad happens

With all that said, though, I've almost got a working version in my tree. I hope to finish it up tomorrow, and start writing testcases. Some things may crash for holding on to things too long, but I'll tackle those as they come. I'm still debating whether I want to allow renaming of XBL or XSLT nodes just yet. I may just have to put them to the fire and see how they fare. Boris summed it up pretty pretty well:

<bz> caillon: who are the perverted individuals who thought of this idea?

Looks like tomorrow will be an interesting day.

Aggregation

Thanks dbaron for fixing my mess.

The principle of the matter

I uploaded my nsIPrincipal patch just now to its bug. I don't really like the idea of trying to push it in without being able to accurately and thoroughly test it, but the good thing about Mozilla is that so many people download it daily and can test it for me. This will ultimately be the best test harness anyway, since there are so many different combinations of things that this patch could affect. I'd rather get this patch landed early in beta so as to give more time for people to catch regressions. I don't foresee many, but you never know....

One, Two, Three

I've been on a DOM Level 3 kick lately. I've recently added some extra things to Node, and now I'm working on Document.

Testing

So I've gotten my rather large patch for combining three of our principal objects into one polished enough to where I think it is ready for testing. I asked around and got a link to the internal CAPS regression tests. The only problem is, most of the tests fail. On trunk nightlies. So I have no real tests to compare my patch to. On the bright side though, a build with my patch shares the same bugs as the trunk does... Maybe I'll just attach it to the bug, get reviews, land it, and hack any regressions.