News for hackers. Stuff that matters.
Re:Comet is dead, long live websockets by endtwist on Wednesday December 16, 2009 @12:22PM (#999211)
I find articles like this annoyingly premature. Sure, we're seeing the birth of a whole new way of communicating with the backend, which will only server to improve web apps in a positive manner.

At the same time though, support for WebSockets in major browsers is woefully little, with only Chrome currently supporting it in an active "release," and (I believe) support available in the Firefox trunk for a later version. This still doesn't account for Safari or Internet Explorer, the latter being the biggest impediment to adoption.

Additionally, I certainly don't think this is the beginning of the end for "AJAX" — not in the least. AJAX still has very strong grounds in pulling down data from the server that doesn't require some sort of long-running communication.

Yes, let's get excited, but a little bit of a sanity check should be done if you're ready to call AJAX "dead" (or "dying").

Re:Comet is dead, long live websockets by pkulak on Wednesday December 16, 2009 @01:15PM (#999338)
It will probably be in all non-IE browsers soon enough, and then it becomes like border-radius. You get it if you have a real browser, otherwise some Flash fallback library will kick in.
Re:Comet is dead, long live websockets by 10ren on Wednesday December 16, 2009 @08:08PM (#1000159)
It might get adopted quickly on a niche platform that relies heavily on webapps, and needs the efficiency because it is a relatively low powered device.

Like a phone.

Re:Comet is dead, long live websockets by anonjon on Wednesday December 16, 2009 @12:39PM (#999251)
I agree with you, but why can't this be used as an excuse to kill internet explorer?

It seems that IE has been a consistent impediment to web based technologies for upwards of 10 years now. There have been so many other things that have been horribly crippled because 'internet explorer doesn't support it'.

People are becoming more and more savvy from the standpoint of computer use, and putting up a redirect to install a new web browser... like Chrome or Firefox... is something the majority of people should be able to handle/understand.

I'm tired of waiting 10 years to actually use the technology that we have /today/.

Re:Comet is dead, long live websockets by qeorge on Wednesday December 16, 2009 @04:01PM (#999709)
It seems that IE has been a consistent impediment to web based technologies for upwards of 10 years now.

This is patently untrue. In 1999, IE was the best available browser (Netscape would have been your other option).

It wasn't until Firefox in 2005 that a real alternative emerged (IMHO).

Sorry to be nitpicky, but it matters.

Re:Comet is dead, long live websockets by jasonkester on Wednesday December 16, 2009 @04:14PM (#999736)
Indeed. IE pushed many of the technologies that make today's stuff possible. Divs, usable Document Object Models, vector graphics, and XmlHttpRequest, to name a few.

I'd go so far as to say that it was 2008 before something genuinely better came along in the form of Chrome. Firefox did a good job of getting Microsoft to release IE7, but it was never compellingly better in the way that Chrome is.

Re:Comet is dead, long live websockets by daleharvey on Wednesday December 16, 2009 @12:45PM (#999265)
I think this is a point that a lot of people miss, everyone is quick to blame those big corporate it teams that wont upgrade browsers that its their fault.

people dont upgrade unless they have a reason to, not because there is a better rendering engine or more plugins, by pulling your hair out and slowing yourself down to work on support for ie, you are also giving users a reason to not upgrade.

but (IT wont let you install firefox because x only works on ie6) works both ways, if the rest of the web shoots forward with these new technologies, companies will catch up.

Re:Comet is dead, long live websockets by tedunangst on Wednesday December 16, 2009 @03:07PM (#999599)
Indeed. I can't believe the IE team hasn't implemented a standard whose draft specification is dated today. Why are they so fucking slow?
Re:Comet is dead, long live websockets by axod on Wednesday December 16, 2009 @05:00PM (#999849)
Websocket has been coming for about 12 months. Most of the spec was sorted out months and months ago.
Re:Comet is dead, long live websockets by anonjon on Wednesday December 16, 2009 @04:11PM (#999727)
Whether or not you are serious, indeed they should have something in the works for a standard whose draft specification is dated today.

I think the bigger issue is that Microsoft doesn't bother adhering to standards;

Or they make up their own standards when there are already perfectly good standards that provide the same functionality;

Or they implement part of a standard, but certain stuff is completely wrong, and when it is pointed out, they don't do anything about it.

And then they force you to conform their arbitrary non-standard stuff, because they can bully you with their 70% market share.

This isn't just about being slow, its about their course history of being slow AND maliciously protectionist AND incompetent.

And they should get ditched if they continue to behave in such a manner.

Re:Comet is dead, long live websockets by jasonkester on Wednesday December 16, 2009 @06:38PM (#1000028)
Historically, Microsoft has put forth their own standards well before anybody else, only to see the official "Standard" be written differently by a 3rd party.

DOM, CSS, Box Model, Events, SVG, and a bunch of other things they did before anybody else (in a straightforward and reverse-engineerable way) were ignored by later groups who put together slightly incompatible "standards", some of which had no reference implementation. NN6, for example, was out for a full year before it implemented the W3C Events model that the Netscape team had proposed (which of course didn't exactly match the reference implementation that IE5 had in place 2 years earlier).

It's only recently that the Standards folks have managed to spec out things that the IE group hadn't done yet.

Re:Comet is dead, long live websockets by allertonm on Wednesday December 16, 2009 @02:57PM (#999581)
What makes you think this excuse to kill MSIE will work when 10 years of other excuses have failed?