[Bf-committers] Blender on Windows - some thoughts about XP and32bit

[Sergey Sharybin]

We would need to drop XP at some point, but it should not be ASAP. It
should be a clear for everyone process with defined release when we're
dropping. Ideally it should be connected with switching to a new official
compiler for windows. Meaning, when we're considering MSVC 2013 as a
default to build blender on windows then we might stop maintaining old
libraries and let XP support fade down naturally.

As for dropping support of 32bit -- i don't see good reason for this just
yet. It might be 2-3 people who supports libs on windows, but i only know
one who supports all the libs libs on linux. It's not much of hassle if we
update libs only when we need this (meaning not just to be on a leading
edge with brand-new libs).

Dropping 32bit windows builds would mean we're dropping atom support as
well? It might be still useful to be able to run blender on a netbooks i'd
say. Also, you might want to use PAE on 32bit platform because of some
better load balance and so.


On Fri, Dec 6, 2013 at 12:07 AM, Thomas Dinges <blender at dingto.org> wrote:

> Hi Mitchell,
> this is strange, maybe its due to some compiler flags?
>
> Maybe Visual Studio 2013 doesn't suffer from this, otherwise it would be
> bad. Needs investigation.
>
> Best regards,
> Thomas
>
> Am 05.12.2013 18:10, schrieb Mitchell Stokes:
> > The last time I checked, vc11 created slower Blender builds than vc9 for
> > the game engine. Not that I would like to stick to vc9, but vc11 isn't
> > always faster. For a specific example, I've found OpenMP to be rather
> slow.
> > It's been a while since I last ran some tests, but I seem to remember the
> > difference being at least 20%.
> >
> > --Mitchell Stokes
>
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> http://lists.blender.org/mailman/listinfo/bf-committers
>



-- 
With best regards, Sergey Sharybin