[sugar] another response time

Gary C Martin gary at garycmartin.com
Wed Oct 1 14:15:08 EDT 2008


On 1 Oct 2008, at 13:49, Mikus Grinbergs wrote:

>> as you probably installed previously the Wikipedia activity, my guess
>> is that the jffs2 gc thread was taking most of the CPU.
>
> If I understand correctly, this raises the possibility that other
> actions performed *prior* to the launching of an Activity can
> noticeably affect the time it takes to launch that Activity.
>
> Would someone else please launch XaoS, and see what kind of
> "response time for the launch" they get?
>
> Tried it again, after the XO had sat there overnight (having now
> hopefully done everything it needed to, for jffs2 housekeeping).
> For me, the "launch screen" for XaoS pulsed for 100 seconds before
> the "activity screen" was drawn.  [My guess is that it is
> calculating the fractal picture before showing it.]

Hi Mikus,

XaoS on my XO B4, launches instantly. No seriously, it's absolutely  
the fastest launching activity. Here's the rub, because its so quick,  
it even beats the launcher window getting going, the working activity  
is hidden behind the launcher pulse effect window, which now can't  
tell that the activity is actually already running so does its blind  
'I'll sit here and strobe for a fixed time-out because I have no idea  
what happened.'

Take a look in the frame and you'll see the default grey circle that  
you can switch to.

Worth a ticket if there's not one already, it's an interesting case in  
how quick XO-1 HW really is with the right software (must be and  
absolutely massive price being paid currently for having the Pyhton  
environment I'd guess, but I'm sure there is a whole heap of  
optimisation to be made if/when made a dev cycle focus and team  
resources made available).

Made some great strides for 8.2, fingers crossed for 9.1.

--Gary


More information about the Sugar mailing list