I don't think I've ever seen so many warnings on a single Wikipedia article.
To be honest, the ridiculous ways in which you have to implement trivial network applications on the Web sometimes makes me think whether the Web may have severely slowed down technological progress. People are still building IRC, usenet, FTP and e-mail, but this time in the stupid way.
This is not an interesting post. The OP on Leah Culver's site is one of the most ridiculously obvious things ever written on the web; a "definition" of AJAX? Really? She offers no insightful commentary (linking to Wikipedia).
Every post on her blog is a cleverly (but not that cleverly) disguised self-promotional stunt. I've never seen somebody on the web so obsessed with herself. Who, exactly, does she think her audience is? A bunch of Diggers?
If somebody wanted to post about the "Comet" PL to YCN, they could have just posted the link to the Wikipedia article and started a discussion about that.
I remember when the controversy first flared up there. Basically, there's one Wikipedia user who has an axe to grind against a particular publication, and parlayed that into the whole slew of nasty warnings you see now.
I actually tried to take part in the discussion for a little while, but gave up when it was clear that the dude was just pursuing some personal dislike of a source rather than any real questions about the reliability of information.
I was always fascinated how a single function - XMLHttpRequest() has grown to be promoted into a TECHNOLOGY but it just outdone itself by becoming two (!) TECHNOLOGIES by taking a little longer to return to the caller.
scanf() must be screaming with envy. Poor old scanf()... future holds nothing for it.
For things besides chat, Comet sounds simpler than XMPP.
I'm planning a project management project with a whiteboard, this article is going to save me the time I was going to spend trying to make a chat server act like things it's not.
It will be killed as soon as WebSockets get implemented.
WebSockets will bring the real-time web alive. It will be the greatest advance in internet since its early days.
Google Wave will ride on WebSockets. Web alerts will be the next gold mine. iPhone like push messages for email, feeds, etc., web chat without hacks, instant collaboration and much more.
We really don't understand right now the impact of the real-time web, but it will be pervasive.
I'd adventure to say by the end of the year all major browsers will implement WebSockets: Firefox, Safari, Chrome and Opera are working on it as we speak.
I've had enough of this real-time web silliness. Threading and networking on the client side is exactly what it seems. "Real time" is something entirely different than a client-server metaphor, which is all this browser thing has ever been.
The difference? Real time applications run discrete logic on bare metal and have sub-millisecond response time. Web applications that use multiple data sources with short update periodicity is what it says on the tin.
As many HN links hint, REST = CRUD, AJAX = TCP/IP and what was old is new again.
However, while the future may look like a cheaper IBM-style mainframe-to-the-world thing, it just looks like web services via vendor lock-in to me. Like a low-rent version of SUN's old slogan.
I don't think I've ever seen so many warnings on a single Wikipedia article.
To be honest, the ridiculous ways in which you have to implement trivial network applications on the Web sometimes makes me think whether the Web may have severely slowed down technological progress. People are still building IRC, usenet, FTP and e-mail, but this time in the stupid way.