[Testing] Automatic power management

[James Cameron]

On Thu, Aug 26, 2010 at 04:20:54PM +0530, Asiri Gunarathna wrote:
> So James, you also told that there s a side effect of two unrelated
> changes in a complex system, while we talk about didn t stop when
> playing "Blu may" on browser. Gathering all these things, can I raise
> a defect as in sugar-labs; already playing audio file didn t stop even
> enabled the automatic power management, [...]

Like Paul, I'm a bit confused as to what you are getting at.

You need to express your opinion of what the software behaviour should
be, along with arguments justifying your opinion.

So far your opinion seems to be based only on the dubious validity of
previous test case documentation, or the different results obtained when
playing audio files on different laptop models.

It is not at all important how many seconds should elapse before
suspend, when there is variable activity and usage of the laptop.

However, there is one point I'd draw from your write-up that may be of
interest:

- there is an inconsistency in laptop idle suspend behaviour when
  playing audio, between Record, Browse, and Jukebox,

The cause of this inconsistency is that the CPU power cost of playback
varies according to the application design.  It is fully expected.

Now, if it became necessary to fix this inconsistency, we might do it by
suppressing the idle suspend function in those activities when a
playback is happening.  We know we can do that.

But for what reason?  What is more important; playing music, or saving
energy?

-- 
James Cameron
http://quozl.linux.org.au/