[sldev] Valve/Steam Hardware survey.

Callum Lerwick seg at haxxed.com
Wed Jun 13 22:14:19 PDT 2007


On Wed, 2007-06-13 at 23:07 -0400, Jason Giglio wrote:
> Callum Lerwick wrote:
> > On Wed, 2007-06-13 at 16:12 -0400, Jason Giglio wrote:
> >> It's in the official minimum requirements.  But there's no requirement 
> >> that SL run fast on those chips, so if it's just going to be slower on 
> >> those chips, I say go for it.  SL is probably already pretty horrible on 
> >> such chips, making it a little worse isn't going to matter much.
> > 
> > Yeah, that's a solid philosophy. Lets give tax breaks to the rich as
> > well. (Dzonatas isn't the only one who can stretch a metaphor. :)
> 
> To abuse it even more, taking performance away from chips made in the 
> last 5 years to support these platforms that can't realistically run SL 
> anyway, is more like taxing everyone to make sure that only homeless 
> bums on the street have $50,000 a year in income, even when the middle 
> class is being taxed to well below that income.

Back to the original topic, you were rather abstract to begin with. If
you're talking about SSE, a 1.4ghz T-bird is a perfectly reasonable
thing to run SL on. I'm not talking about benefiting "platforms that
can't realistically run SL anyway".

Though really, I couldn't give a rats ass if LL decides to require SSE
or not in their builds, I have the source and my Fedora build will
continue to support T-birds.

Of course, this is all moot unless anyone can actually derive benefit
from SSE. Lets see the code. In my experiments with OpenJPEG, its pretty
much memory bandwidth bound, SSE actually slows things down if anything,
due to the stricter alignment requirements.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://lists.secondlife.com/pipermail/sldev/attachments/20070614/1e39ec80/attachment.pgp


More information about the SLDev mailing list