[NTLK] Weird Firefox behavior: Please do a test for me

Lord Groundhog lordgroundhog at gmail.com
Sat Jun 25 08:48:34 EDT 2011

~~~ On 2011/06/23 23:18, Jon Glass at jonglass at usa.net wrote ~~~

> I'm replying onlist, because what I'm seeing is odd enough that others
> may want to try this as well...
> In Firefox 4.0.1 on my MacBook Pro 2.66mhz 4 gig of RAM, OSX 10.6.7,
> highlighting that line bumped its cpu to 100%+. Also, though, clicking
> and selecting several other lines did the same, or nearly as
> high--near 100%. I got curious, and tried a couple other web pages. It
> seemd rather random, but I could repeat this with other entire-line
> selections, but not everywhere.
> Next, I got _really_ curious, and tried it in Safari...

Also going onlist with this because my results are again different and a
little odd.  

First, my test setup is Firefox 4.0.1 on a 2007 MacBook 2.4GHz Core 2 Duo
with 4 GB RAM, running OSX 10.5.8.

I selected the single line as requested and monitored the results on the
OS's own Activity Monitor. The CPU activity went up -- briefly, for about 2
seconds -- to 2.2%.  It then dropped again to 0% and stayed there.

I tried selecting and deselecting several times, tried the same thing again
with different lines.  Always the same result.

Finally I selected the whole column of lines, and then the whole page. That
was the only way I managed to get a serious increase in CPU usage, all the
way up to 55.8% on those occasions.  Even that lasted less than 10 seconds
though, and usage then dropped back to 0% and stayed there.

Having repeated this numerous times in Firefox and in Safari, I cannot
reproduce the problem on my setup.

Out of curiosity, what does a Newton do with this page?  ;-)


~~~ ~~~ ~~~

³Any sufficiently advanced magic is indistinguishable from a Newton.²
            -- ref.:  Arthur C. Clarke

(With thanks to Chod Lang)

~~~ ~~~ ~~~
Fight Spam.  Join EuroCAUCE: http://www.euro.cauce.org/
Get MUGged and love it: http://www.oxmug.org/
Join today: http://www.newtontalk.net/

More information about the NewtonTalk mailing list