#6 small fixes to 0.2.4

[Anonymous]

*suggestion for smaller and faster(less try blocks)
XMLHttpRequest code
* suggestion for working around the ie mem leak problem
when using addHandler() ( commented out both on the top
and in addHandler() )
* Changes to addHandler()and removeHandler() to avoid
using eval, don't automaticly think that the browser is
if is w3c event model is unsupported and test for the
correct methodes in removeHandler()

changes taken from jaxx : javascript asynchronous xml &
xslt, a soon publicly small & fast ajax lirary

[Nobody/Anonymous]

xajax with some small fixes

[Eion Robb]

Logged In: YES
user_id=1414103
Originator: NO

Hi there,

Thanks for the suggestions. I'm interested in your fixes for the addHandler() methods.. mainly because I wasn't aware that there was any memory-leaks in there. :)

The XMLHTTPRequest stuff using browser version detection I'm not so fond of though. Not to say I'm fond of using try/catch blocks either, however using the code as you've shown, doesn't allow for other versions of the XMLHTTP library to be loaded, eg version 4 which is still publically available (although deprecated) for download, and is sometimes used as a last resort. Might look at using some kind of static way of detecting the XMLHTTP version though so that only the first call to getRequestObject is slow (on IE)?

So to reiterate, I'll look through your changes in addHandler and probably will apply them, but wont change the XMLHttpRequest object as you've done it.

Cheers,
--BigBrownChunx

[Eion Robb]

Logged In: YES
user_id=1414103
Originator: NO

Just had a look through our admin conversation logs, and the reason we used eval() statements there, was that things didn't work with Safari otherwise. Weird bug.