Hi Jack,
On 31-Dec-08, at 3:27 AM, Jack Killpatrick wrote:
Hi Bob,
I created the listnav plugin. Thanks for the nice comments about it
in your post, glad you like it.
I certainly do. Nice work!
I'm curious about the issue you're seeing when using Blueprint CSS.
In your post you said that the issue goes away if you remove the
blueprint screen.css file. You mean if you remove that, but keep
listnav in place, right?
Yes, the only change is to remove the blueprint/screen.css style
sheet. I can leave all the other ones in place, including the others
that come with blueprint.
Does the perf change much if you remove listnav and keep blueprint
screen.css?
Yes, the problem goes away as well. It would have been very tempting
to leave it at that, but the differences between browsers was a flag
that it wasn't this simple. I probably should have been a little more
explicit about that, if the jQuery document ready function is removed
-- which is equivalent removing listnav in the simplified version I
had worked down to -- the performance problem goes away.
Is there a way I can take a look at your example? I'd like to give
it a look and see if anything jump out at me.
Yes, actually. I put some demonstration stuff on one of my websites
yesterday afternoon, but annoying internet connection problems (it is
winter here) prevented me letting anyone know.
Here's a link:
<http://www.raconteur.info/stuff/strange-css-js-problem/fast-for-webkit.html
>
There is a link on that page to the slow version, so you can flip back
and forth.
I'm seeing a flash on safari like I reported on FF from that URL. Even
more interesting, occasionally (rarely, but annoyingly often the
*first* time, sigh) Safari doesn't show the
problem.
If you download all the files from this tarball:
<http://www.raconteur.info/stuff/strange-css-js-problem/strange-css-js-problem.tgz
>
and play with it locally (e.g. a file: URL) you'll not see the flash
in Safari after the cache is loaded.
Something to do with latency affecting the timing?? Maybe the css is
being delayed just enough?
Cheers,
Bob
Thx,
Jack
Bob Hutchison wrote:
Hi,
[reposted, didn't appear in archives... sorry for the noise]
I've encountered, and worked around, a rendering performance
problem in WebKit based browsers.
I'm working on a webapp and a particular uncontrived page is
rendered by FireFox in about 0.5s while Safari takes about 12s --
that's about 20 times slower. And 12s is way too long from any
user's point of view.
I really don't know how jQuery/ListNav and the Blueprint CSS
stylesheet are interacting, and I certainly don't know whose
'fault' it is, but I do know it can be avoided by applying the
style sheet after ListNav has finished doing it's thing on the page.
I've written in more detail about this on my weblog:
<http://recursive.ca/hutch/2008/12/30/how-to-destroy-page-rendering-performance-using-just-javascript-and-css/
>
Cheers,
Bob
----
Bob Hutchison
Recursive Design Inc.
http://www.recursive.ca/
weblog: http://www.recursive.ca/hutch
