[Mono-dev] simd: more accelerated classes

Miguel de Icaza miguel at novell.com
Wed Jan 14 17:19:07 EST 2009


    Speaking in terms of our policy as a team:

> First, it would require to be API stable, which will be a pain in the
> neck during the initial ramp up while the design is flushed out.

This is the major problem, but there is nothing stopping people from
developing the library, trying it out, exploring its use, and once we
feel comfortable with the result, we would be happy to include it with

The support angle is not as bad as you think, and in fact for a
technology like Mono.Simd and a higher level library, it would be easy
to produce a test suite that goes with it, a test suite that runs on
every single invocation of the buildbot.

Having this *in* Mono (once stabilized, GACed, etc) means that if we
ever regress, we can catch this almost immediately, as opposed to having
to wait for a pass of QA releases and previews, and having third parties
come to us and say a month later "You broke SIMD".

So I suggest that such higher-level library first gets developed
elsewhere under a suitable license (MIT X11): prototype, evolve, tune,
fix, adjust and unit test.

Once that is done, we should include it with Mono.

> As for hosting the project in the mono svn server, there isn't much of
> an advantage for it, really. I, for one, would rather use github for
> example. But if you really fancy it, please talk to Miguel, as he is
> the one that can make such arrangement.

We can put it on Mono svn.


More information about the Mono-devel-list mailing list