Returned mail: Host unknown (Name server: darkstar.diginext.fr: host not found)
Mail Delivery Subsystem (MAILER-DAEMON++at++relay-smtp.imaginet.net)
Mon, 8 Mar 1999 07:37:35 +0100 (MET)
The original message was received at Mon, 8 Mar 1999 07:37:35 +0100 (MET)
from sophocle.imaginet.fr [195.68.0.10]
----- The following addresses had permanent fatal errors -----
<diginext++at++darkstar.diginext.fr>
----- Transcript of session follows -----
550 <diginext++at++darkstar.diginext.fr>... Host unknown (Name server: darkstar.diginext.fr: host not found)
- message/delivery-status attachment: stored
attached mail follows:
Date: Sat Mar 6 02:00:05 PST 1999
From: info-performer++at++sgi.com (info-performer Mailing List)
To: allan++at++holodeck.engr.sgi.com
Subject: info-performer Mar 05 1999
List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
Send Submissions to: info-performer++at++sgi.com
Add/Remove requests: info-performer-request++at++sgi.com
Message Subjects:
performance on lock phase.
Re: performance on lock phase.
Re: PBuffer on Onyx2 Reality
buffer requirment for glPolygonOffsetEXT
Re: performance on lock phase.
Re: performance on lock phase.
Re: saving rgb files (pfuSaveImage)
Ask for help!
Re: Ask for help!
Re: buffer requirment for glPolygonOffsetEXT
3ds Loading with pthread and Rgba from PC
******************************************************************************
From: "VSM 13" <vsmsa++at++club-internet.fr>
Date: Fri, 5 Mar 1999 09:24:30 +0100
Subject: performance on lock phase.
Message en plusieurs parties et au format MIME.
------=_NextPart_000_0013_01BE66E9.F90ACBC0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
I made a bench with perfly, on oNYX2 IR2 tripipe
When I use free mode : the application runs at 30 Hz.
When I use lock mode on 30 Hz: the application runs between 20/30Hz.
(I lock the cull and draw process),
How you can explain that?
HALLAKOUN Yoel
VSM sa
------=_NextPart_000_0013_01BE66E9.F90ACBC0
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN">
I made a bench with perfly, on oNYX2 =
IR2=20
tripipe
When I use free mode : the =
application runs at=20
30 Hz.
When I use lock mode on 30 Hz: the =
application=20
runs between 20/30Hz.
(I lock the cull and draw =
process),
How you can explain =
that?
HALLAKOUN Yoel
VSM sa
------=_NextPart_000_0013_01BE66E9.F90ACBC0--
******************************************************************************
From: Rob Jenkins
Date: Fri, 05 Mar 1999 00:03:43 -0800
Subject: Re: performance on lock phase.
>I made a bench with perfly, on oNYX2 IR2 tripipe
>When I use free mode : the application runs at 30 Hz.
>When I use lock mode on 30 Hz: the application runs between 20/30Hz.
>(I lock the cull and draw process),
>How you can explain that?
This is an illustration from the PF course notes:
Video refresh 60Hz( 16.7 mS /frame )
Frame Rate 30Hz ( 33.3mS /frame )
Frame 1 overrruns
| ! | ! | ! | ! |
|-F0-|~>|-F1-------------|~~~~~>|-F2-|~>!-F3-----|~~~~~>! | FREE
| ! | ! | ! | ! |
|-F0-|~~~~~~~~~>|-F1-------------|~~~~~~~~~~~~~>|-F3-----|~~~~~>| LOCK
| ! | ! | ! | ! |
In FREE run, the draw can start as quickly as possible so no frame get
dropped ( in this case though frame rate will vary )
In LOCK F0 has to wait for a fixed frame boundary, then as F1 overruns
it hs to wait for another fixed frame boundary so gets even further
behind.
This isn't exactly what you are seeing but you get the idea, in your
test FREE run is managing to always get everything done in 33ms whereas
in LOCK the wait for a frame boundary is causing drops.
Cheers
Rob
--
________________________________________________________________
Rob Jenkins Silicon Graphics mailto:robj++at++sgi.com
******************************************************************************
From: reiners++at++igd.fhg.de
Date: Fri, 5 Mar 1999 10:35:19 +0100 (MET)
Subject: Re: PBuffer on Onyx2 Reality
On Mar 4, 7:37pm, Christian Ernst wrote:
> Subject: PBuffer on Onyx2 Reality
> Hi all,
>
> I'm trying to use the SGI PBuffers on different machines. I have read
> the old messages from Dec 98, so that I know about the "Bad Alloc "
> problem.
> When I get the fbconfig list, I'm trying all configs till I'm
successful
> in
> creating a pbuffer.
> This works perfect on an Octain, an Indigo2 and an Onyx. But on an
Onyx2
> Reality I get no suitable pbuffer. On this machine I can only
allocate
> pbuffers with no depth buffer and R>=10/G>=10/B>=10, or with depth
> buffer
> and R>=10/G=0/B=0.
>
>-- End of excerpt from Christian Ernst
Apparently iR/Reality won't give you a usable pbuffer if you only have
small pixel depth. When we reduced the managed area so that we got
medium pixel depth a usable configuration became available.
It's a bit strange, as AFAIK small pixel depth is 128 bit/pixel and
Impact has, again AFAIK, 3*36=108 bit/pixel. There are probably some
other magic constraints about blocking different groups of bits, but I
would have expected more.
As you only have a Reality you have a problem there, as even 1024x768
will only get you small pixel depth. You would have to go down to
800x600 or cook up your format in the area of ~960x680. Not a pretty
sight in any case, but good enough for testing.
But then again I just might be wrong and there is a magic button that
will bring us eternal peace and infinite pbuffers. Can anybody tell me
which one it is?
Dirk
--
--
-- Dirk Reiners reiners++at++igd.fhg.de, Dirk.Reiners++at++gmx.net
-- IGD - A4 http://www.igd.fhg.de/~reiners
-- Rundeturmstrasse 6
-- D-64283 Darmstadt All standard disclaimers apply.
-- Truth is stranger than fiction because fiction has to make sense.
******************************************************************************
From: Liu Xiaoyan
Date: Fri, 05 Mar 1999 18:36:28 +0800
Subject: buffer requirment for glPolygonOffsetEXT
Hi, all,
I'm trying to do the same light mapping effect as that in
Bumplogo on our VR platform (CAVE). Two coplanar textured
polygons are offsetted using glPolygonOffsetEXT before they
are blended. I tested this very simple part with different
framebuffer setup. It works in Performer but not in CAVE. In
CAVE, I can only see the front polygon without any blending.
I suspect it's due to different framebuffer configurations.
What's the buffer requirement for gl-polygon_offset
extension and glTexEnvfv(ENV_BIAS_SGIX)? Does it need
stencil?
static int PreBlend(pfTraverser *, void *)
{ pfTransparency(PFTR_BLEND_ALPHA);
pfOverride(PFSTATE_TRANSPARENCY, PF_ON);
glBlendFunc(GL_ZERO, GL_SRC_COLOR);
glPolygonOffsetEXT(-0.5f, -0.00001f);
glEnable(GL_POLYGON_OFFSET_EXT);
glTexEnvfv(GL_TEXTURE_ENV, GL_TEXTURE_ENV_BIAS_SGIX,
biasparams);
glTexEnvfv(GL_TEXTURE_ENV, GL_TEXTURE_ENV_COLOR,
colorparams);
return PFTRAV_CONT;
}
static int PostBlendC(pfTraverser *, void *)
{
glBlendFunc(GL_ONE, GL_ZERO);
pfOverride(PFSTATE_TRANSPARENCY, PF_OFF);
pfTransparency(PFTR_OFF);
glDisable(GL_POLYGON_OFFSET_EXT);
glPolygonOffsetEXT(0.0f, 0.0f);
return PFTRAV_CONT;
}
Thanks.
Liu
--
***********************************************************************
Liu Xiaoyan Institute of High
Performance Computing
Research Engineer National University
of Singapore
Data Visualisation Group http://www.ihpc.nus.edu.sg
Tel:(65)7709267
***********************************************************************
******************************************************************************
From: Phil Keslin
Date: Fri, 05 Mar 1999 07:17:33 -0800
Subject: Re: performance on lock phase.
> VSM 13 wrote:
>
> I made a bench with perfly, on oNYX2 IR2 tripipe
> When I use free mode : the application runs at 30 Hz.
> When I use lock mode on 30 Hz: the application runs between 20/30Hz.
> (I lock the cull and draw process),
>
> How you can explain that?
>
> HALLAKOUN Yoel
> VSM sa
Which OS? With IRIX 6.5 there is a bug that can cause the draw sync
logic to miss a frame boundary. We're working on a fix.
- Phil
--
Phil Keslin
******************************************************************************
From: "VSM 13"
Date: Fri, 5 Mar 1999 17:55:29 +0100
Subject: Re: performance on lock phase.
Thank you all to answer.
I'm in ONYX2 IR 2 on 6.5 with 3 pipes.
Phil Keslin said to me that it's a bug, I think so because I'm not this g=
ap
on my old RE2 on 5.3
Marc Simon said that I have to genlock my pipe and use the swapready. I h=
ave
genlocked with horizontal rate but I haven't use the swapready ( how you
configure it ?)
Rob Jenkins said that the frame boundary can cause this, does exist some
optimisation for that ?
-----Message d'origine-----
De : Phil Keslin
=C0 : VSM 13
Cc : info-performer
Date : vendredi 5 mars 1999 16:17
Objet : Re: performance on lock phase.
>> VSM 13 wrote:
>>
>> I made a bench with perfly, on oNYX2 IR2 tripipe
>> When I use free mode : the application runs at 30 Hz.
>> When I use lock mode on 30 Hz: the application runs between 20/30Hz.
>> (I lock the cull and draw process),
>>
>> How you can explain that?
>>
>> HALLAKOUN Yoel
>> VSM sa
>
>
>Which OS? With IRIX 6.5 there is a bug that can cause the draw sync
>logic to miss a frame boundary. We're working on a fix.
>
>- Phil
>
>--
>Phil Keslin
>
******************************************************************************
From: Colin Bannister
Date: Fri, 05 Mar 1999 16:58:33 +0000
Subject: Re: saving rgb files (pfuSaveImage)
Thanks to all those who replied to my posting on saving images from a
fly-through. This is now working well, including the use of shared memory
to pass data into the appropriate callback routines. A lot of this was in
the 'perfly' code, but its not always so easy to spot at first glance.
Colin
--
Dr. Colin Bannister
Academic Computing Services
Cripps Computing Centre Tel: 0115 9513326
The University of Nottingham Fax: 0115 9513358
Nottingham NG7 2RD
www: http://www.nottingham.ac.uk/~cczcb/
email: Colin.Bannister++at++nottingham.ac.uk
******************************************************************************
From: yunda++at++mail.sc.cninfo.net (Publishing User)
Date: Fri, 5 Mar 1999 09:14:56 +0800
Subject: Ask for help!
Dear Performer Experts,
I am a new Performer user.I want to use Performer
to develop visual simulation program in locomotive
simulator.I use MultiGen to build model,when I use
perfly to drive it,the texture on the ground plane
exhibit blurr.I want to know how to resolve this problem,
maybe there are some tricks.
I have another question: whether it is fit for to use
the clipmaping in visual simulation
of locomotive simulator or not.
Thank you very much!
Qiujing
---------------------------------
QiuJing
Tel:(86)28-7600132(o)
Email:qiujing5++at++263.net
yunda++at++public.sc.cninfo.net
qiu_jing++at++usa.net
Chengdu,Sichuan
P.R.CHINA
******************************************************************************
From: Angus Dorbie
Date: Fri, 05 Mar 1999 16:20:35 -0800
Subject: Re: Ask for help!
The problem is isotropic MIP mapping.
See the "Anisotropic" section on www.dorbie.com for an explanation and
sample code which can fix this in Performer. You have a very constrained
problem so it should be easy for you to unblur your tracks & approaching
objects like buildings and hedge rows.
Cheers,Angus.
Publishing User wrote:
>
> Dear Performer Experts,
>
> I am a new Performer user.I want to use Performer
> to develop visual simulation program in locomotive
> simulator.I use MultiGen to build model,when I use
> perfly to drive it,the texture on the ground plane
> exhibit blurr.I want to know how to resolve this problem,
> maybe there are some tricks.
> I have another question: whether it is fit for to use
> the clipmaping in visual simulation
> of locomotive simulator or not.
>
> Thank you very much!
> Qiujing
>
> ---------------------------------
> QiuJing
> Tel:(86)28-7600132(o)
> Email:qiujing5++at++263.net
> yunda++at++public.sc.cninfo.net
> qiu_jing++at++usa.net
> Chengdu,Sichuan
> P.R.CHINA
>
> =======================================================================
> List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
> Submissions: info-performer++at++sgi.com
> Admin. requests: info-performer-request++at++sgi.com
--
"Only the mediocre are always at their best." -- Jean Giraudoux
For advanced 3D graphics Performer + OpenGL based examples and tutors:
http://www.dorbie.com/
******************************************************************************
From: Angus Dorbie
Date: Fri, 05 Mar 1999 16:29:10 -0800
Subject: Re: buffer requirment for glPolygonOffsetEXT
There is no requirement, other than a depth buffer.
Watch your z precision as the EXT version will vary depending on z bits.
Also multisample on iR triggers a different z storage mechanism which
improves precision in the distance and will affect the required offset.
You can choose to use pfLayers for this, watch that performer may be
clobbering your layter information, it makes it's own offset calls.
If using pfLayers you could try using the stencil layering instead of
offset, the default pf offset values chosen don't work very well on iR
because they are excessively large and cause punch through (too much
offset), this should be fixed in the 2.2.5 release, and OpenGL 1.1
platforms will use the precision independent 1.1 OpenGL ARB offset
instead of the old EXT offset in the next release.
Cheers,Angus.
Liu Xiaoyan wrote:
>
> Hi, all,
>
> I'm trying to do the same light mapping effect as that in
> Bumplogo on our VR platform (CAVE). Two coplanar textured
> polygons are offsetted using glPolygonOffsetEXT before they
> are blended. I tested this very simple part with different
> framebuffer setup. It works in Performer but not in CAVE. In
> CAVE, I can only see the front polygon without any blending.
> I suspect it's due to different framebuffer configurations.
>
> What's the buffer requirement for gl-polygon_offset
> extension and glTexEnvfv(ENV_BIAS_SGIX)? Does it need
> stencil?
>
> static int PreBlend(pfTraverser *, void *)
> { pfTransparency(PFTR_BLEND_ALPHA);
> pfOverride(PFSTATE_TRANSPARENCY, PF_ON);
> glBlendFunc(GL_ZERO, GL_SRC_COLOR);
> glPolygonOffsetEXT(-0.5f, -0.00001f);
> glEnable(GL_POLYGON_OFFSET_EXT);
> glTexEnvfv(GL_TEXTURE_ENV, GL_TEXTURE_ENV_BIAS_SGIX,
> biasparams);
> glTexEnvfv(GL_TEXTURE_ENV, GL_TEXTURE_ENV_COLOR,
> colorparams);
> return PFTRAV_CONT;
> }
>
> static int PostBlendC(pfTraverser *, void *)
> {
> glBlendFunc(GL_ONE, GL_ZERO);
> pfOverride(PFSTATE_TRANSPARENCY, PF_OFF);
> pfTransparency(PFTR_OFF);
> glDisable(GL_POLYGON_OFFSET_EXT);
> glPolygonOffsetEXT(0.0f, 0.0f);
> return PFTRAV_CONT;
> }
>
> Thanks.
>
> Liu
> --
> ***********************************************************************
>
> Liu Xiaoyan Institute of High
> Performance Computing
> Research Engineer National University
> of Singapore
> Data Visualisation Group http://www.ihpc.nus.edu.sg
> Tel:(65)7709267
> ***********************************************************************
>
> =======================================================================
> List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
> Submissions: info-performer++at++sgi.com
> Admin. requests: info-performer-request++at++sgi.com
--
"Only the mediocre are always at their best." -- Jean Giraudoux
For advanced 3D graphics Performer + OpenGL based examples and tutors:
http://www.dorbie.com/
******************************************************************************
From: "ÀåÀç¿ø"
Date: Sat, 6 Mar 1999 10:23:53 +0900
Subject: 3ds Loading with pthread and Rgba from PC
This is a multi-part message in MIME format.
------=_NextPart_000_0014_01BE67BB.6F4746C0
Content-Type: text/plain;
charset="euc-kr"
Content-Transfer-Encoding: base64
SGl+ICANCg0KSSdtIHByb2dyYW1taW5nIHdpdGggcGVyZm9ybWVyLg0KDQpUZXN0IHRoaXMgOiBj
b21waWxlICYgbGluayAgbW90aWYuQyAoaW4gQysrKSB3aXRoIHB0aHJlYWQgKCAtbHB0aHJlYWQg
KSBMaWIuDQoNCmFuZCBMb2FkaW5nICouM2RzICggaW5jbHVkZSAgdGV4dHVyZSBtYXAgaW1hZ2Ug
KQ0KDQooIGNoZWNrIGltYWdlIGNvbnZlcnRpbmcgcHJvY2VzcyBqdXN0IGxpa2UganBnLT5yZ2Ig
KQ0KDQpPbmUgbW9yZS4uLg0KDQpJIHdhbnQgdG8ga25vdyB0aGlzLi4uDQoNClVtbS4uLiAgSG93
IGRvIEkgaW1wb3J0IGFscGhhICggbWFrZSAqLnJnYmEgKSBmcm9tIG15IFBDICggd29ya2luZyB3
aXRoIDNkc01BWCApLg0KDQpJbiBteSAzZHMsIHRyYW5zcGFyZW5jeSwgb3BhcXVlIGF0dHJpYnV0
ZXMgaGF2ZSBub3QgYmVlbiBpbXBvcnRlZCBjb3JyZWN0bHkuLi4NCg0K
------=_NextPart_000_0014_01BE67BB.6F4746C0
Content-Type: text/html;
charset="euc-kr"
Content-Transfer-Encoding: base64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------=_NextPart_000_0014_01BE67BB.6F4746C0--
******************************************************************************
This archive was generated by hypermail 2.0b2
on Sun Mar 07 1999 - 22:42:12 PST