Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

CPU utilization up in recent weeks

1 view
Skip to first unread message

JoeS

unread,
Jun 25, 2007, 7:51:21 PM6/25/07
to
I test the nightly Thunderbirds and have noticed a sharp drop in XBL (marquee) and DHTML performance in recent weeks.
I don't really look at performance on a daily basis, but here are my recent findings.
Current trunk CPU pegs with a scrollong marquee, same for some of my favorite DHTML scripts.
I have a "checkpoint" build ==
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9a5pre) Gecko/20070504 Thunderbird/3.0a1 ID:2007050403
Performance seems fine there, although not as good as branch.
Would you like input on these issues, maybe a tracking bug. Reflow?

JoeS

Martijn

unread,
Jun 25, 2007, 8:23:41 PM6/25/07
to JoeS, dev-tec...@lists.mozilla.org
Hi Joe,

Those things don't sound like things that have to be tested in thunderbird.
They could just as well be tested in Firefox.
What are your results when you view those testcases in Firefox trunk?
Is there a place somewhere where we can see those testcases?
If you really see DHTML performance slowdown, then yes, you probably
should file a bug on it. Testcases and a regression range would be
very useful.

Regards,
Martijn


2007/6/26, JoeS <joesa...@gmail.com>:

> _______________________________________________
> dev-tech-layout mailing list
> dev-tec...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-tech-layout
>


--
Martijn Wargers - Help Mozilla!
http://weblogs.mozillazine.org/qa/
http://wiki.mozilla.org/Mozilla_QA_Community
irc://irc.mozilla.org/qa - /nick mw22

JoeS

unread,
Jun 25, 2007, 9:20:50 PM6/25/07
to
I don't run current Firefox trunks, but I do have a SeaMonkey
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9a5pre) Gecko/2007053102 SeaMonkey/2.0a1pre
The CPU pops to 100% there also.
This is with WinXP but noted also by a Linux user.

Sometimes I think it is easier to see these kinds of regressions in Newsgroups with knowledgeable users than in
random Web observations.
The regression was noted here:snews://secnews.netscape.com:563/f5n7kt$4l...@ripley.aoltw.net
Attached an html file that shows the problem. Just a simple vertical marquee.

Joe

--
div.sign { position:relative; bottom:0em; width:90%;
background-color:#fffacc; border-style:dotted; border-width:1px;
-moz-border-radius: 1em; font-style:italic; text-align:left;
padding-bottom:2px; margin-bottom:1px; }
* JoeS ...**"Get a Gecko" TB Release Vers
<http://releases.mozilla.org/pub/mozilla.org/thunderbird/releases/1.5.0.10/>
TB Trunk w/Plugins
<http://ftp.mozilla.org/pub/mozilla.org/thunderbird/nightly/latest-trunk/>
Seamonkey <http://www.mozilla.org/projects/seamonkey/releases/> How
to post to MTMM <http://ilias.ca/mozilla.test.multimedia>
*

marquee.htm

Martijn

unread,
Jun 25, 2007, 9:39:36 PM6/25/07
to JoeS, dev-tec...@lists.mozilla.org
2007/6/26, JoeS <joesa...@gmail.com>:

> Sometimes I think it is easier to see these kinds of regressions in
> Newsgroups with knowledgeable users than in
> random Web observations.
> The regression was noted
> here:snews://secnews.netscape.com:563/f5n7kt$4l...@ripley.aoltw.net
> Attached an html file that shows the problem. Just a simple vertical
> marquee.

I can't see the attachment, it's somehow scrambled here.
I see some text of it. You seem to be using a 1px dotted border, right?
Does it make a difference if you remove that border?
In that case you might at least suffer from
https://bugzilla.mozilla.org/show_bug.cgi?id=379834

Regards,
Martijn


--

JoeS

unread,
Jun 25, 2007, 10:26:44 PM6/25/07
to
Sent the file as an attachment to your gmail
Joe
0 new messages