View Full Version : General Browser Trends

06-20-2006, 06:44 PM
Things I have noticed lately (relevant to web developers at least):

The Mozilla camp is focusing on their scripting support. Various extensions to ECMAScript have been incorporated lately (generators, list comprehensions, etc), and there has been renewed work to put Python in the same position as Javascript with respect to DOM scripting, XPCOM writing, etc. This work would also extend nicely to Ruby and whatnot. Also, the switch to the Cairo engine, which is allowing crazy increases to performance and amazing svg foreignObject support.

The Opera camp has focused on their DOM. I don't see much changes in CSS between Opera 8 and 9 (other than opacity: support), but the differences in DOM support is staggering. It is my opinion that Gecko has finally been surpassed in DOM support. Also, they have been focusing on SVG to a large extent (and finally made a somewhat crippled XSLT implementation available).

The WebKit people... the nightlies have been incorporating KSVG piece by piece, gradually getting better and better support, though last I checked had no support for the SVG DOM. I honestly don't know what WebKit is focusing on.... I might say SVG, but it certainly doesn't seem with as much zeal as Opera with SVG or Mozilla with scripting.

Microsoft of course: security. No real new features in IE7 that make our lives significantly better as web developers.

Summarizing as simply as that is a little misleading, but it seems to be the predominant focuses. Any thoughts?

06-21-2006, 06:02 AM
Personally I think that firefox should hold up with the scripts, and start seeing some serious CSS3 features roll out. I really want border-image. It would be nice to see the memory footprint start to get smaller, too.

I've not used Opera since v5, nor ever used webkit so I can't comment. :(

MSIE is dying. There is absolutely no doubt about it. It needs SVG, decent javascript support, and still is lacking in elementary CSS3 specs.

06-21-2006, 06:24 AM
MSIE is dying. There is absolutely no doubt about it. It needs SVG, decent javascript support, and still is lacking in elementary CSS3 specs.

Let alone CSS2. As for Firefox, it already has some serious CSS3 support (-moz-appearance, many CSS3 selectors, even the CSS3 Column Layout support), though more never hurts.

As for scripting, I think its crucial that *someone* starts moving forward on it. Despite the advent of all the AJAX-stuff, the web is stagnating. Everything you see now is the tail end of the wave of new stuff from a couple years ago. And no new amount of DOM or CSS support is going to revolutionize the experience 5 years from now. SVG is a step in the right direction, but I honestly think Mozilla has it right by concentrating on improving the very language (and expanding the available languages) that drive web applications. Javascript is absolutely one of my favorite languages, but it doesn't lend itself well towards large codebases, which is where web applications seem to be going. JS2.0 seems to remedy this, but Python has also shown itself to work well in large environments, and making the environment language-agnostic would open up web-developer to even more coders experienced in perhaps other fields. (I would *love* to do all of my web-scripting in Scheme or Haskell!)

06-21-2006, 01:29 PM
the web is stagnating. Everything you see now is the tail end of the wave of new stuff from a couple years ago.

That is a very, very good point indeed.

It would be nice to be able to be comfortable with using a language knowing that support for it is standard across browsers. Opera and Firefox tend to meld quite well, and they are both in active development, which is exactly what a web browser needs. They wern't designed to be stagnent.

I'm rather worried that we're getting towards and alphabet soup of langauges however. We've got things like Perl, PHP, ASP, Ruby and JSP for server side stuff, but apart from PHP and Ruby, they are growing worthless. ECMA/JS2.0 for client side stuff. XHTML/XML/Your Format Here for page content. SVG/PNG/Jpeg/Flash/Mp3/Vorbis/Others for multimedia. If things are going to progress into the next generation, we need to drop the crap, and concentrate on moving only the solid technology's.

From my point of view, the technology's that will/should survive are these:

PHP - It just keeps evolving, but unless PHP 6 brings some speed enhancements and drops some depreciated stuff, we'll be stuck with PHP 4, which will become stagnent very quickly.

Ruby - I've never used it, but I'm hearing good things. I've no doubt it will progress with all this new Ajax rubbish that people seem to want.

JS2/ECMA - Hard to tell where it is going at the moment. We're seeing some nice things with Ajax, but at the end of the day, it wont move the web forward very much, apart from making pages pretty and faster to use.

PNG/GIF/SVG - I think these are good standards. GIF is really seeing some age now, but the overhead on PNG's favor the GIF's file sizes for small images. SVG could well replace Flash if it's scripting is improved. I will rejoice on the day that flash is finally dead.

06-21-2006, 02:55 PM
I don't think the next few years of innovation in the web space will have anything to do with cutting-edge technologies per se - it's the application and combination of those technologies that makes them useful.

With the advent of better CSS support in IE7, we now have a pretty decent toolkit with which to build the next generation of apps - we don't really need much more in the way of DOM support. With the exception of solid SVG, which I agree is a necessary thing - we can already do what we need with what we have.

What I'd like to see over the next few years is a crystallisation of existing technologies and techniques, particularly AJAX - put the brakes on slightly when it comes to pushing its boundaries, and concentrating on making the set pieces work properly - without driving a ditch through usability and accessibility, leaving whole groups of people behind.

06-21-2006, 03:36 PM
I'm slightly worried by the current state of W3C projects; CSS2.1, which has been regarded by most as the latest standard still hasn't left Working Draft stage yet; I wish they've moved it to Candidate Recommendation and Recommendation by now--unless they're still waiting for IE to catch up.

Also, I wonder what the future for XHTML looks like now that the WHATWG people are drafting up what they call Web Applications 1.0; I can see what they're trying to achieve, but I don't think a second standards body would really improve things. And although Web Applications 1.0 seems to be meant to become (part of a) W3C recommendation (which one?) at some time in the future, seeing W3C's slow pace as of late I wonder if that's ever going to happen.