SL Target Audience (was RE: [sldev] branches/shadow-draft missingllrender.cpp)

[Soft]

On Wed, May 28, 2008 at 9:27 AM, Matthew Dowd
<matthew.dowd at hotmail.co.uk> wrote:
>
> I do have issue with the "if you machine is older than a few years, stop
> being a skinflint and buy a new one" implication of the e-mail though!

That's not quite accurate. The implication was that if you want new
high level rendering features, you need high level rendering hardware.


> There is a perception that LL is concentrating on fancy features which
> require top end, up to date machines at the expense of working on improving
> the stability and performance of SL on mid-range/older machines. DaveL's
> comments, and observations such as below about SL's performance on lower end
> machines (and there seems to be some controversy as to whether WL with all
> the shaders turned off runs better or worse than pre-WL on older hardware),
> only re-inforces this perception.

Perceptions aside, some objective data would be really helpful. A
great way to deal with WL issues would be to create a meta JIRA for
supported hardware configurations that perform worse under the
released version of the WL-enabled viewer, and to define some
objective tests for these issues.

The Lab doesn't have every combination of hardware out there, and so
far as I know WL performs similar to - or better than - previous
viewers on the hardware we do have. Some specific benchmarks and test
scenarios identifying specific rendering problems on hardware we don't
have are guaranteed to help more than talking a graphics heavyweight
off of heavyweight graphics work.