Re: [Gnash-commit] [SCM] Gnash branch, release_0_8_10, created. release_0_8_9_final-1429-g4ae56c3

[Rob Savoye]

On 01/23/12 14:21, Sandro Santilli wrote:

> Yes, I'm sure reducing the heartbeat is very helpful in reducing
> CPU load. I have some ideas too about how it can be done w/out

  Gnash needs to have better performance, especially when using AGG. A
cpu load over 70% is just unacceptable, yet I get a cpu load of 99%
still for many SWF files. I did much testing with real-world examples,
which obviously produced different results than a custom designed test
case. For some of those files, I was getting a 30% drop in cpu load.
That particular test case was full of Timers, I never noticed any
problems with those after adjusting the time interval.

> Moving there in small steps would be advisible, IMHO.

  You're talking about a 3 line change, I think that would be considered
a small step. :-)

        - rob -