On May 23, 11:08 am, Michael Ben-Yosef <septa...@[EMAIL PROTECTED]
> wrote:
> I have in the past been interested in whether the RenderMan Interface
> Spec would be a good standard to adopt for a new renderer or modeller.
For a renderer, I would say it depends a lot on your project. There
are certainly upsides to using it: you get to use a lot of existing
tools and get some guidance as to how things should be done. But it
could also be overly broad for your renderer or become a burden to
sup****t.
For a modeller, sup****ting a single spec is probably a bad idea. The
way to go is to have a fairly neutral API on top of which plug-ins can
be written for various standards. Even if you're only going to sup****t
RenderMan, it's a good idea to not let the rendering API "crawl" too
deeply in your app.
> My first question is: in the opinion of the readers of this newsgroup,
> did this change improve the viability of the RI as a public standard,
> signaling a change in Pixar's attitude towards it?
I don't think use of the RI spec was ever *the* major point of the
lawsuit. Now I'm not a lawyer and I haven't read the actual filings,
but from what I heard and read, it had more to do with Larry leaving
Pixar with lots of knowledge about how PrMan works and starting his
own competing company. There are a lot of trade secrets in a renderer,
many of which are quite simple but hard to obtain, so I could
understand if some people at Pixar got pissed. Now that's just my
reading between the lines so take it for what it's worth ;-)
> To me it seems that there are still dangers for the (non-PRMan) users
> of RenderMan:
> 1. Pixar asserts numerous times in the RI Spec 3.2.1 (On p. i, in the
> Preface on p. x, and on p. 214) that the RI Spec may not be copied.
> The statement repeated on pp. i and x is, "No part of this publication
> may be reproduced, stored in a retrieval system, or transmitted, in
> any form or by any means, electronic, mechanical, photocopying,
> recording, or otherwise, without the prior permission of Pixar."
> Version 3.2.1 appears to have added the sentence (p. 214), "The
> foregoing statements of permission expressly do not include permission
> to reproduce the specifications for the RenderMan Interface." This
> means that any modeller or renderer following the RI Spec cannot
> reproduce the standard governing its use other than to link to Pixar's
> online copy of the Spec
athttps://renderman.pixar.com/products/rispec/risp=
ec_pdf/RISpec3_2.pdf
> . Should Pixar decide to take that do***ent off its site, a conforming
> program is left completely stranded in terms of do***entation.
There are printed versions of the spec, probably a truckload of online
copies, etc. But it's still a valid point. I read that as "don't copy
our do***entation for your own software", which brings us to the next
point...
> 2. Pixar can decide to close up subsequent versions of the RI Spec,
> making them available only to licensed PRMan users. Is it the case
> that this is already happening with the draft version 3.3, or has
> Pixar stated that it will become publicly available?
I don't think they have (or ever will) make any statement about that.
And later versions of the "spec" are pretty much the do***entation of
PrMan. I don't know if there will ever be an effort to make later
versions public. Pixar certainly has very few good reasons to do that.
Places which buy "the whole pipeline" have little reason to ask for it
either.
> Also, have there been any attempts to make a more "open" standard in
> the same mould as the RI? In the do***ent "Ten Things About NVIDIA
> Gelato that May Surprise and Delight You" (http://www.nvidia.com/
> object/gelato_ten_things.html), I see the following statements made:
> "Gelato=92s APIs are also truly 'open.' The APIs and how they are used,
That's marketing :-)
> Do the authors of Gelato envisage it's API as a possible alternative
> standard to the RI?
I'm sure they did (it is very similar) but right now it doesn't seem
to be getting anywhere. Nvidia, being squeezed from all sides by
integrated video solutions, is probably quite busy enough finding a
future for itself.
Olivier
|
