Multi-Surfaced Blobs

Here it is! The "alpha" code for supporting multiple surfaced blobs
in Rayshade.  I say "alpha" because it, the code, has been
tested, but not rigourously.  It seems simple enough to implement and try
that I feel it should be available to those who want to use it.  Above
all, I want everyone interested in this to have some source with which
to experiment and have some fun!

Please send me any comments or suggestions for improving the code,
speeding it up, or just general style issues.

I want to thank Mark Podlipec for all his help in getting this running
(read: If it weren't for him I might still be thrashing about
helplessly).  Kudos, of course, to everyone who has helped make
Rayshade available to even the most novice developers, like myself.

I have tar'red and compressed only the "necessary and sufficient"
files for making multi-surfaced blobs.  All you need to do is
uncompress and untar the file 'CBlob.tar.Z' in the top level Rayshade
directory and you should have a subdirectory 'cblob' with the
necessary files in it.  Next you need to put them into the their
proper directories(see below) and recompile, but please back up the
original files first! :-) The names of the files in CBlob.tar are
identical to the Rayshade distribution files.

		      Using Multi-Surfaced Blobs

Multi-surfaced blobs have been implemented, at this time, as a new
object.  This object is called a "cblob".  Anything you can do to a
"blob" you can do to a "cblob".  Here's the catch, the syntax for a
"cblob" is different, but only _VERY_ slightly.  The "cblob" has, after
each coordinate triple for a metaball, the metaball's particular
surface specification.  For example, below is the "cblob"
specification derived from Mark's blob example file "blob.ray": 

/*
 * Example blob input file, courtesy of Mark Podlipec
 * Modified for multi-surfaced blob (i.e. cblob) by George McGregor
 */
eyep  0.0 -6.0 1.0
lookp 0.0 0.0 0.0
fov 45. 45.
screen 240 240
light 1.4 point 15.  -20. 25.
light 0.7 point 5. -25. 10.

surface white ambient .02 .02 .02 diffuse .5 .5 .5 specular .8 .8 .8 specpow 32
surface bland ambient .02 .02 .02 diffuse .5 .5 .5 specular .8 .8 .8 specpow 32

/* Note: Seperate surfaces for each metaball... which may not be necessary */
surface crystal diffuse .03 .03 .08
        reflect 0.03 transp 0.96 index 1.5
surface greensmoke diffuse .01 .05 .01 reflect .10 specpow 30 specular .8 .85 .8
surface blueglass diffuse .09 .15 .45 specular .5 .5 .6 reflect .5
surface gold  diffuse 0.22 0.18 0.05 specular .8 .8 .4 ambient .04 .02 .0
	reflect 0.93 specpow 58
surface ceramic diffuse 0.6118 0.6794 0.7471 ambient 0.2039 0.2265 0.2490
  specular .5 .5 .5 specpow 25
surface plaster diffuse 0.6735 0.6059 0.4465 ambient 0.2245 0.2020 0.1388
surface silver diffuse .02 .02 .04 specular .71 .71 .7 reflect .95 specpow 70

/*   surface  treshhold strength radius    x      y      z      surface */
/*      |         |         |       |      |      |      |      |   |   */
/*      V         V         V       V      V      V      V      V   V   */
cblob bland      0.6       1.0     1.0   -0.900  0.000  0.000  plaster
                           1.0     1.0    0.900  0.000  0.000  greensmoke
                           1.0     1.0    0.000  0.000 -0.900  ceramic
                           1.0     1.0    0.000  0.000  0.900  silver

plane white 0. 0. -3.77   0 0 1



Please note the multiple surface specifications, one for each metaball 
in the "cblob".  As Mark pointed out to me, surfaces are global and it
would be a good idea to make the "cblob"'s surface specification
unique.

I am certain that with a better understanding of the yacc parser and a
little more work I can make the "blob"/"cblob" objects into one
description. e.g. the surfaces for the metaballs would be optional and
thus make the single surface "blob" the default case.  Also, it would be
nice to be able to specify a surface once and have that surface used as a
default for subsequent metaballs until a new surface specification is
given.  I have attempted these cases, but the shift/reduce errors and
parsing glitches forced me to getting the "cblob" working first.  I am
hoping Mark, or someone who is more familiar with these sorts of
things can point me in the right direction.

I apologize if I have left anything out, and I look forward to any
constructive commentary about this code.  I will be placing this in
our anonymous ftp directory under pub/rayshade as well as at the
weedeater site.

Ftp'ing details for ncifcrf.gov:

	ftp fcs280s.ncifcrf.gov   -or- ftp 129.43.2.10

        [login as "anonymous" and password in user@host form.] 

	cd pub/rayshade

Now you may get the files. Oh, yes, be sure to set to binary to before
you grab the .Z files.  They should all have the CBlob prefix.

The weedeater details have been covered elsewhere.  I will place the
files in the incoming directory. Again, they should all have the CBlob
prefix. 

If you have any problems you can reach me at the adresses below.

Enjoy!

George


  George McGregor    @	NCI Biomedical Supercomputer Center
  c/o PRI, Inc.		Phone: (301) 846-5789                  
  PO Box B, Bldng. 430  mcgregor@ncifcrf.gov
  Frederick, MD 21702	mcgregor@fcs280s.ncifcrf.gov

NOTE:

The original locations for these modified, source files are as follows:

[relative to the top rayshade directory]

File                  original location
----                  -------- --------

blob.c                libray/libobj/blob.c

blob.h                libray/libobj/blob.c

yacc.y                libshade/yacc.y

lex.l                 libshade/lex.l

cblob.ray             can go into Examples/cblob.ray

color.h               libray/libcommon/color.h