Re: Performer 2.0 Problems

New Message Reply Date view Thread view Subject view Author view

Brian Furtaw (brian++at++sgi.sgi.com)
Thu, 14 Mar 1996 09:03:25 -0500

I would like to address question number five from below, one of the replies to
this message suggested that IGLOO was a translation of OpenGL to IRISGL on VGX
systems, this is not true IGLOO is a translation of IRISGL to OpenGL which is
currently availiable on the IMPACT product, IGLOO will be used for backward
compatiblity of IRISGL on future releases of IRIX, this is necessary because
the IMPACT and now Infinite Reality have hardware which is specifically
designed to run OpenGL not IRISGL. The problem you are seeing on the VGX is
because OpenGL has limited support on that platform `grelnotes gl_dev` should
tell you the limitations of the VGX platform and OpenGL.

Brian

On Mar 13, 8:46am, David H. Whittington wrote:
> Subject: Performer 2.0 Problems
> I'm having difficulties getting started with Performer 2.0
> and would like some feedback from those who are doing ports.
>
> 1) When running the makefile provided with perfly, all sorts of
> multiply defined symbols are spewed out at the link stage - if there
> are any real errors I would not easily see them. There is also a
> linker command "ignore unresolved" in the makefile that hides any
> real problems once you start to modify perfly to your own desires.
>
> We have several machines around of different speed and we were able
> to easily take a performer based executable around from machine to
> machine but now, even when I try to link "as static as possible",
> when I move to other machines, some shared objects are missing
> and it vomits stuff to the screen and then dies - what a pain!
>
> 2) I was dismayed that perfly_2.0 is considerably changed in areas
> from perfly_1.2 and I am unaware of any code comments, man pages,
> or other documentation that provides understanding to the user about
> the need for the architectural changes.I tried using the tool port2.0
> on our previously working 1.2 based source and stopped after
> frustration - too many significant changes, etc.
>
> 3) I will ask again - why the OPENGl version of perfly so SLOW on
> a VGX machine compared to the irisgl version. I see a factor of 5
> times slower on a "5-span" VGX Crimson. I could understand a 30-50%
> slowdown, but this??
>
>-- End of excerpt from David H. Whittington

-- 
o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o


Brian Furtaw			(brian++at++sgi.com)
RSE Graphics/Communications	Office:	(301) 572-3293
Silver Spring, MD		Fax:	(301) 572-3280

New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:52:33 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.