Re: How should at-spi be turned on?
- From: Bill Haneman <bill haneman sun com>
- To: Michael Meeks <michael ximian com>
- Cc: merchan baton phys lsu edu, gnome-accessibility-list gnome org
- Subject: Re: How should at-spi be turned on?
- Date: 01 May 2003 11:17:00 +0100
On Thu, 2003-05-01 at 06:02, Michael Meeks wrote:
> Hi Gregory,
>
> On Wed, 2003-04-30 at 16:48, Gregory Merchan wrote:
> > But my reason for playing with at-poke may be a reason to always load the
> > modules. Last time I saw platform-wide GUI scripting discussed, use of AT
> > was suggested as the way to do this. Using AT would should solve at least
> > two problems "GNOMEScript" would have. As it's in the libraries and
> > should already provide complete functional coverage, there's no need for
> > extra work to support scripting. (AppleScript had this problem, afaik.)
> > Because AT should be working regardless of where the X clients are running,
> > there's no need for another rendevous layer like ICE's.
>
> IMHO; the accessibility code being 'always on' will have a very
> significant performance impact; and is not really the right way to go
> about setting up a general scripting infrastructure - although, clearly,
> there are many useful ideas / implementations in there.
Michael:
I think Greg's point is worth exploring more. In actual usage I don't
agree about the implied severity of the performance impact. I use
pretty slow machines ;-) and I rarely if ever notice any difference in
performance when I flip the accessibility switch.
I don't doubt that the performance impact may be "technically
significant" but that would be true of any wide-ranging scripting
facility, and in this case it may be worth that cost. In either case
we'd need to do something more than gut-level analysis before making a
decision IMO.
regards,
- Bill
> Regards,
>
> Michael.
>
> --
> mmeeks gnu org <><, Pseudo Engineer, itinerant idiot
>
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]