Re^2: CGI and System Administration

I am going to beg to differ on Dojo. Anything that requires you to break standards to use it (Dojo requires non-XHTML attributes) is broken out of the box. Also, I don't know about the current version but earlier versions have the worst documentation of any software I've ever tried to use; including the fact that the version number is nowhere in the code so it's hard to even know which missing docs to try to find. YUI has nice tech docs and a big community. Ext JS is a pretty fork off it. jQuery, to me, is the most Perl like hitting the sweet spot of syntax and brevity, and it has, in just two years, spawned about 600 known plugins (alla CPAN) and probably another 600 you can find on your own. Its Google groups are very active, 100 posts per day on the main one. There are others too, like Mootools and MochiKit. (Update: Prototype is probably not on the list of good choices.)

Comment on Re^2: CGI and System Administration

Replies are listed 'Best First'.
Re^3: CGI and System Administration by dragonchild (Archbishop) on Apr 04, 2008 at 20:22 UTC
I used Dojo quite extensively for a year and I never used a non-XHTML attribute. That's only one way of using it. My criteria for good software: Does it work? Can someone else come in, make a change, and be reasonably certain no bugs were introduced?	[reply]
Re^4: CGI and System Administration by Your Mother (Archbishop) on Apr 04, 2008 at 20:59 UTC
Thanks for noting it (and good for you for taking the high road on it -- it's a pet peeve that so many "use strict or die" Perl hackers have a "So what, as long as it works in IE," attitude about XHTML/CSS). I don't know if it was the earlier version of Dojo (I think it was 0.4) or what but the code I inherited was rife with it and there were no alternatives offered in any forum post, code sample, or the very, very few documents I could find.	[reply]
Re^5: CGI and System Administration by dragonchild (Archbishop) on Apr 05, 2008 at 01:37 UTC
Oh, you can so write your stuff with the non-standard attributes and, in fact, that's the recommended way to do it. That said, I find it a lot nicer to use code that I can actually control. I don't trust anyone to parse my XHTML other than the browsers and then, I only trust so far. I certainly don't trust someone walking the DOM. I don't even trust my own DOM-walking code. Hence, why everything gets created by hand. My criteria for good software: Does it work? Can someone else come in, make a change, and be reasonably certain no bugs were introduced?	[reply]