I develop a variety of high-end online interfaces, and some of them
are very complex. For example, one page has about 40k of DHTML
(generated by PHP), 20k of JavaScript (linked via a <script> tag), and
an 8k CSS file (linked via <style>). This loads fine in Safari 1.2,
Netscape 7.1, or Mozilla 1.6, in IE 5.1.7 under Mac OS 9, in any
recent Windows 2K/XP browser, or in any recent Linux browser.
But on OS X 10.2.8, using IE 5.2.3 or 5.2.1, my browser gets stuck
with a spinng ball. Cancelling the load, then looking at the page
source, shows that IE has downloaded about 4k of the DHTML, then
simply stopped. This happens whenever the page is over 35k or 40k.
If I chop out some of the DHTML so it's down to 35k, the page loads
perfectly.
Any ideas on how to prevent this lockup (other than splitting up the
content)? Is it a server configuration issue? A header of some kind?
Is there any way of enabling a debugger in IE so I can see what it's
getting stuck on? I appreciate any suggestions.
Mandy - 04 Jun 2004 21:03 GMT
I've noticed this on 5.0 and 5.23. Does it appear to happen when the
included *.js file is greater than 35k, or the sum download is greater than
35k?
This has been bugging us (no pun intended) for a few weeks, and we almost
feel a sense of relief knowing that it isn't just crappy code... our code
may be fine, just too much for these 2 browsers to handle.
I've seen a similar thing on some AOL/IEs, too... that does NOT make me feel
better.
> I develop a variety of high-end online interfaces, and some of them
> are very complex. For example, one page has about 40k of DHTML
[quoted text clipped - 9 lines]
> If I chop out some of the DHTML so it's down to 35k, the page loads
> perfectly.
Dekortage - 07 Jun 2004 14:42 GMT
I'll be more specific. The first time I noticed this problem, it was
with a PHP file that was delivering (or not delivering) a large chunk
of DHTML based on prior user input. Without that chunk, the DHTML is
roughly 33k, with about 56k of external JavaScript (cached) and an 18k
CSS file (also cached), and the page loads without problem. _With_
that optional chunk, the DHTML grows to around 57k, and the page does
not load, getting stuck with a spinning ball after about 4k of data.
I have not had the time to cut out smaller chunks of the DHTML to
determine the exact kilobyte size that IE trips on (e.g. trial and
error). I have now witnessed the problem on a couple of other
similarly large and complex pages.
It is very odd -- and very frustrating -- that these same problematic
pages load without problem in IE v5.1.7 on Mac OS 9 (at least so far).
Clearly, a bug was introduced in v5.2.*. Of course, it works fine in
Microsoft Windows versions of IE 6 (haven't tested IE 5 under
Windows).
But, like you, I am relieved to know I'm not the only one seeing the
problem. :-)
> I've noticed this on 5.0 and 5.23. Does it appear to happen when the
> included *.js file is greater than 35k, or the sum download is greater than
[quoted text clipped - 6 lines]
> I've seen a similar thing on some AOL/IEs, too... that does NOT make me feel
> better.