February 2010 Archives

With all the excitement around those two great articles on JagerMonkey yesterday, I forgot to blog about another amazing development, the DirectWrite & Direct2D landing in the nightly builds (though pref'd off.)

DirectWrite and Direct2D are Windows Vista and Windows 7 APIs for text and 2-D graphics that can be hardware accelerated.

This is brand new code and there are sure to be bugs. If you'd like to help us test these changes and you're on a supported platform, please download the latest nightly build and make these changes to enable the features:

1. Enter 'about:config'
2. Click through the warning, if necessary
3. Enter gfx.font in the 'Filter' box
4. Double-click on 'gfx.font_rendering.directwrite.enabled' to set it to true
5. Below this, right click and select New > Integer to add a pref setting
6. Enter 'mozilla.widget.render-mode' for the preference name, 6 for the value
7. Restart

(To disable, set gfx.font_rendering.directwrite.enabled to false, delete mozilla.widget.render-mode, then restart.)

If you find bugs, please report them here or to Bugzilla.

note: Some extensions like stylish and adblock+ may break the new features. If you're not seeing the new DirectWrite and Direct2D changes, try starting in Safe Mode.

update: If you want to help others find and test this new feature, please digg it. Thanks.

update2: The fonts look really great with DirectWrite enabled. Really great.

Here are two really good blog posts on what's coming for Mozilla's JavaScript engine:

JaegerMonkey – Fast JavaScript, Always! -- David Anderson gives a nice overview of how we currently work and what's coming up.

Starting JägerMonkey -- David Mandelin also offers up some good high level information in addition to going into some pretty specific details.

We're probably not far from a world where all of the browser vendors are sharing these performance techniques and when we get there, as I noted in an earlier post, I think we'll see this particular performance "war" wrapping up with everyone winning.

Madhava Enros, from the Mozilla Mobile team, has just posted a nice new photo of Mobile Firefox (code named Fennec) running on a Nexus One.

Firefox Mobile, already shipping for Maemo devices like the N900 (available in the Ovi Store) has become my primary mobile browser but I can't wait for the Android version to reach a stable enough point that I can start using it regularly.

Go team!

Congratulations to Opera on the release of their new JavaScript engine.

Looks like we've got one more browser in the JS speed race and off to a pretty amazing start as well.

So how close will JS get to native code speeds? If you could take the fastest parts of each vendor's implementation and combine them to form a single engine -- that is, if you look at the fastest score for each of the sub-tests in these suites, regardless of which browser it came from, how close would that put JS to native code in execution performance?

Are we 2x slower? 10x slower? 50x slower? I'm assuming in a year or two, all of us will have copied the best techniques from each other and we'll start to reach an equilibrium that's about as good as it's likely to get for JS performance. When we reach that peak, what will we have?

And, maybe equally as important, where will we all go next in the performance competition, in a couple of years when we've all reached the limits of JavaScript performance? DOM speed? Rendering? Is there anything more that can be done to squeeze performance out of the networking stacks? The various parsers?

It's been fun watching JS get about a million times faster over the last three years and the competition's been really good for the Web, but once we've done what we can for JS, where do we go next in benchmarking performance?

Mozilla’s non-profit mission is focused on self-determination and individual empowerment. We believe that people should have the opportunity to participate in the Web on their terms and that they should have the information they need and the technology choices available to them to ensure that their experience of the Web is under their control.

This week, as part of their settlement with the European Commission, Microsoft will begin rolling out a "browser choice screen" to all Windows users in the European Economic Area. This choice screen will be delivered by Windows Update and will offer people in Europe the opportunity to learn about the different Web browsers available to them and provide a convenient mechanism to download any of a dozen different Web browsers.

But let's not forget that the browser choice screen is just one part of Microsoft's agreement with the EC and beyond that its roll-out represents just one small step to restore self-determination and individual empowerment for users in just one part of the world. This is clearly a positive development, but we've got a long ways to go in helping people the world over understand that they do have the opportunity to participate in the Web on their terms.

On that front, Mozilla has just launched a new program called Open To Choice where we'll be talking with the world about why choice matters. I hope you'll join us in commenting, blogging, or tweeting about this important issue and that together we can make a difference in the lives of the billion or so people online who haven't given any real thought to what the Web can be and how they can help make it so.

Microsoft has posted a draft of the ballot that will appear via Windows Update for IE users in the EU.

The first screen of the ballot features 5 browsers (four rendering engines) randomly ordered and the scrolled off screen section features 7 browsers (three rendering engines) also, but separately, randomly ordered.

Each browser has an icon, a short description, an "Install" link (actually a download link) and a "Tell me more" link. The icon, description and two links were provided by the browser vendors and may be updated periodically over the next 5 years.

A couple of interesting things to note: One, there are a lot of Trident-based (IE) browsers on the ballot. Fully half of the browsers there are based on Microsoft's rendering technologies. Two, not all of the browser vendors were able to provide translations of their browser descriptions. Many European browser users are comfortable with English, but I suspect being the language odd men out on the ballot will negatively impact those browsers uptake. Three, only a couple of vendors provided a "Tell me more" page with information targeted specifically at users encountering this ballot. And finally, vendors made very different decisions in how prominently their brand (as opposed to the browser brand) feature in the ballot. Google's image, for example, features their company brand much more prominently than their browser brand, while Apple's name is nowhere to be found in the Safari image.

What do you all see? Any questions?

update: A couple of people have mentioned that they don't recognize some of the tier 2 browsers. That's likely going to be a very common case for most European users. Most of the second tier browsers are generally unknown or are only known in one or two countries (outside of Europe). I was amused when I went to look up the SlimBrowser on Wikipedia that the article had been deleted "Reason given: Non-notable software near-ad." Not all of them are that bad, for example, Maxthon and Avant are pretty well known in China and Sleipnir is somewhat well known in Japan. Flock has a small but global following, and the others are genuinely rare.

But, in Europe, I don't think any of the tier 2 browsers, maybe with the exception of Flock, really have any significant penetration. For example, the tier 2 browser with the most users, Maxthon, has less than one quarter the usage in Europe as Netscape 6 (Netscape 6!!)

"There was a lot of confusion when it came out on Tuesday, and people thought that somehow we were publishing their email addresses and private information, which was not true."

It's my fault? I was confused? My addressbook isn't private? Really? Really!?

Help us build an identity for the Mozilla Developer Network.

Congratulations to Opera Software for shipping a Windows beta version with support for Open Video.

Thank you to what ever lawyers there finally let the engineers show off their Theora+Vorbis in Ogg video work that goes back more than 2 years.

Both ears are now quite clearly up.

Sodapop turns 9 weeks old tomorrow. The most notable change from a week ago is that one of her ears has started to stand up. Eventually (probably in a week or so) both will be standing tall. The next change, besides weight gain and better manners, will be when more of her silver hair comes in. You can see a bit of it around her eyes, but she's still mostly black and tan.

Sodapop at 8 weeks

Sodapop at 8 weeks 6 days

For those people paying attention (and anyone who might want to update my Wikipedia entry,) Sodapop the yorkie is the third child in our family behind Ptolemy the cat and Munch the bunny.

The Mozilla Store needs your help! If you have a minute, please take this quick survey