Bruce,

Thanks for your post, forgive me for not giving it proper
attention, but certainly I'll bite on major points.

> I'm only playing a little bit of Devil's advocate here, and that's 
> not meant to derail your concept, but an attempt to understand 
> what the benefits are. So on with the games <g>

That's fine.  

> Ummm... With the complete source code to the EXACT POV parser 
> available,

Which BTW is quite hard to maintain, and can't be incorporated
into a commercial tool because of the POV license.  This means
that only POV can read POV files.

> but I suspect your purpose
> is not to be able to quickly make money off the honest efforts of
> other's.

I don't have the attitude that commercial application of open
projects is a bad thing.  Linux, X Windows, Gimp or KDE have not
suffered because they've been "ripped off" by companies like
Red Hat, IBM, SGI, etc...  In fact, these companies are investing
nice things into the core technology, besides making their money.
My interest here is non-commercial, although I could use it
commercially, if the license was more generous.

At the end of the day, I want funkier tools, not tools that
I own and control.  I don't care about owning POV - I wouldn't
even care if I'd invested 1000 hours of my time into it.  But
that's me, not the POV team.

> My only comment is "just because it uses the modernized Greek alphabet,
> doesn't make it 'human legible'! 

It certainly isn't meant to be hand edited, which is an assumption
that many pov users here have a hard time suspending.

> Ask yourself when the last time was that while reading a
> book you ran across the <aTag> ... </aTag> construct?

Very often actually, but in all the parsers I maintain commercially,
they are formulated in different, inconsistent ways.

> HTML and VRML are horrible languages to parse! They include all sorts of
> forward referencing, allow limitless addition of application-specific
> tags

They are still easier to parse than POV, and are at least standardised
so I can buy a commercial library that can handle the VRML parsing for
me.  XML I think is easier to parse because it is more constrained than
HTML or VRML.

>(God forbid you accidentally choose the same 'tag-name' as someone
> else)

This is the whole point of XML.  You define the tags that are relevant
to your problem domain, specify this to the XML parser, and the rest
is mainly automatic.

> a large number of other inconsistencies, and are far from being
> 'human legible'!

VRML is far more legible than say, 3ds or lightwave.  In some ways
comparibly legible to POV-script, considering the difference
between the architectures.

> So basically, I'm wondering what benefit this would really provide?

A means of properly exchanging 3D data between POV ray, 3rd party
tools, database systems, and your own applications.  A standardised
means of encoding POV scenes that can be more easily translated
to later versions of POV.  A encoding scheme that is more oriented
to editing tools, because the structure is more explicit.
 
> I fully endorse the POV-teams wish that POV NOT be a library. It is
> completely user supported, out of the sheer goodness-of-heart of a VERY
> few people, and to turn it into a library would open the door for
> unscrupulous people to use it in their own commercial products.

I respect the right of the POV team to do things this way - but
I honestly think it's time for POV to evolve to a rendering
architecture, rather than a rendering application.  It's a two
way street with commercial products, you can't expect anything
back in POV if you shut out the professional 3D graphics
community.

> So. To convert the POV language to make it easier to use for people who
> want to make POV tools, seems like a bit of wasted effort, since the
> language parser is already available in source code.

It's not.  It is restricted by licenses, and the interface to
POV is not well defined.  I've not heard of anyone using the
existing parser this way.

> but does make the get-rich dweebs look
> for another approach.

?!?  Have you had some kind of nasty experience !?!

> I've heard the "This is the newest-greatest-bestest language ever" 

Me too, but I think XML is a very interesting and useful technology.

> XML may actually provide useful, but it ain't pretty! And it's a long
> way away from human legible.

It's not even meant for human editing in most cases, but since
POV users are so used to hand editing, it may be too tough a
case to make that other methods may have different strengths
and weaknesses.

> So tell me, what benefit would this actually provide?

There is quite a good list on this thread for you to ponder.

(Hope I don't sound too blunt, I'm in a hurry!)


--
Nigel Stewart (nig### [at] nigelscom)
Research Student, Software Developer
Y2K is the new millenium for the mathematically challenged.

Post a reply to this message