Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror

Comment Re:Everything old is new again (Score 1) 158

NaCl is open source but tied to totally undocumented Chrome internals via Pepper, which makes it pretty hard to adopt without adopting Chrome wholesale.

Worse yet, NaCl is tied to particular hardware, which means that if it gets traction on the web the bar for a new hardware platform would become very high (think "ARM would not have been viable if this had happened 15 years ago" high). PNaCl, if/when it starts working would help with that problem, but not the Pepper dependency.

Comment Re:No (Score 1) 213

> with an active community that cares about
> standards,

Sometimes. And sometimes not so much. Compare Gecko and WebKit's CSS 2.1 support (based on the official test suite) at http://www.w3.org/Style/CSS/Test/CSS2.1/20110323/reports/results.html for example.

Or note the behavior of the editors of the Transitions and Animations spec drafts (who did nothing for a few years, until editors who were not associated with Apple took over the job).

The WebKit community has many members who deeply care about standards. How much the community overall and the project as a whole care is very variable.

> has an explicit policy of trying to behave like other
> browsers where possible

For various values of "possible". e.g. https://bugs.webkit.org/show_bug.cgi?id=36084 is a long-unfixed spec-compat and other-browser-compat issue that's "impossible" to fix because there are Dashboard widgets and such depending on the current WebKit behavior.

> listens to feedback, and fixes bugs.

Unless it's inconvenient to one of Apple or Google to
do so, of course.

But yes, generally speaking it's a reasonably run development project, which means it listens to feedback and bugs generally get fixed eventually.

> They are the opposite of IE in those critical
> respects.

You must be thinking of IE6 circa 2004 or so, when there was no IE team.

The IE team did quite a bit of listening to feedback and fixing of bugs in the late 2000s, and is doing it again now that it exists again. Of course they're not fixing them in IE6 no more than WebKit is fixing bugs in whatever fork it is Android 2.2 is shipping. But you might want to take a look at IE9 and IE10 sometime if you haven't already.

Comment Re:TLDR (Score 2, Informative) 208

Actually, the address space is in fact key. The goal is to load a known DLL at a known address in the process address space, not a known address in physical RAM, because process address space is what you can see in the code that you will then try to run that will try to call into the DLL.

So in a 64-bit process, this technique is pretty hard to pull off, since it does in fact rely on address space exhaustion.

Comment Re:How does firefox handle searches? (Score 1) 101

SSL by default for Google since Firefox 14, back in July 2012. See https://bugzilla.mozilla.org/show_bug.cgi?id=633773

For other search engines it depends. For example, Wikipedia has asked that the search through their search plugin keep happening over HTTP for now (see https://bugzilla.mozilla.org/show_bug.cgi?id=758857 ).

Comment Re: Content management (Score 2) 244

> People ought to know that the prefixed attributes
> are in beta and may change.

That would be true if WebKit didn't explicitly promise to never remove or change them. Which they do. So people assume they can use them with no fear.

> Fortunately none of the vendor-specific extensions
> are anything but minor enhancements,

That's just not true for transforms, where not supporting them makes a page done entirely using positioning via transforms totally unreadable.

Or for animations where an element is display:none or off-screen and then animated in: no CSS animations means you never see the element at all.

Seriously, I suggest using a non-WebKit mobile UA for a bit and seeing just how broken some sites are.

Comment Re:I find Trident faster than WebKit. (Score 3, Insightful) 244

Actually, in a very real sense the engine _does_ belong to the competition. To actually get your code landed in WebKit you have to convince the current project maintainers (mostly Google and Apple) to accept it.

Which means that if you want to do something that Google and Apple don't (both, often!) approve of, you have to maintain it as a separate branch and deal with the merge pain. No different from other projects where you have to collaborate with others, but a lot different from having control over the code as Microsoft does with Trident right now.

Comment Re:Arguments of convenience (Score 4, Interesting) 244

Everyone and their mother designing "mobile" sites. For some big names, Google, Disney, Comcast, DirecTV, Flickr will all sniff whether you're on "mobile" and either serve you WebKit-only sites or detect that you're not using WebKit and serve you totally different, mostly unusable, sites than they do to WebKit-based browsers.

You should really try using a non-WebKit browser on Android. It's worse than trying to use a non-IE browser in 2000-2001 or so.

Slashdot Top Deals

"We shall reach greater and greater platitudes of achievement." -- Richard J. Daley

Working...