To LUGNET HomepageTo LUGNET News HomepageTo LUGNET Guide Homepage
 Help on Searching
 
Post new message to lugnet.cad.devOpen lugnet.cad.dev in your NNTP NewsreaderTo LUGNET News Traffic PageSign In (Members)
 CAD / Development / 741
740  |  742
Subject: 
Our Modeling Future (Was: Feature requests for LDraw/LEdit v. 2)
Newsgroups: 
lugnet.cad.dev
Date: 
Tue, 16 Mar 1999 15:00:26 GMT
Viewed: 
947 times
  
In lugnet.cad.dev, Paul Gyugyi writes:
The *strength* of LDRAW is the Type 2 and Type 5 lines.
They are key to making rendered instruction sheet
output look good.  AC3D is a modeller that supports
both triangles and lines, with POV and DXF output.
I encourage you to check it out at:
http://www.comp.lancs.ac.uk/computing/users/andy/ac3d.html
The downside is you have to register ($40) to use custom
.dat import and export routines.

Another approach is to switch over completely to POV.
I'd recommend using Rayshade instead of POV, but that's
just because I know how to extend the lex/yacc parser
to handle .dat syntax directly.  One could probably hack
POV to support a syntax very similar to .dat files, with a
few braces thrown in here and there.  We could then toss
away all of the .dat files, and start working exclusively in
a Raytracer-enhanced syntax.  Time and again, this has
been shown to generate really nice output (L3G0, L2P, L3P,
the POV-port of L3G0, the .dat to VRML conversions, etc),
but some people are happy to stick with the old .dat format.
And, they happen to be the really productive parts authors.
And the simplicity of the .dat language is not to be ignored.

-gyug

i see the good *and* the bad in the .DAT format... i like the idea that the
same parsing routines handle everything from the model down to the line type...
all transformations can be uniformly applied to every sub element, element,
part, and model... it is just plain *perfect* recursion...
my problem is that we have painted ourselves into a corner of limitation...
even with great converters there is still too much time spent tweaking things
so that LDraw gives us the output we strive for... hmmm... as far as type 2 and
type 5 lines are concerned... routines could be implemented in LDraw2 that
could process the "visiblity" of mesh edges created in a modeler (i know how
all this works in MAX, i can't say for other apps... but in MAX you can set an
edge's visibity on or off... i wrote a routine in MAX2DAT that can then be set
to read this info and use it to generate type 2 lines only from edges that are
set to "visible" in the MAX mesh)
this type of processiung built into LDraw2 would deal with the type 2 line
issue...
as far as type 5 lines... we could perhaps have LDraw2 process edge information
differently for edges of a particular color... or name...
MAX2DAT currently will create type 5 lines from any cylinder-based object named
"optionalxxxxxxxx..." while this has nothing to do with what LDraw does to
process the type 5 line in a .DAT file, it is something to look at in terms of
how we could have a *real* modeler output objects meeting a certain spec and
then have LDraw2 process these objects so that their output is the same as
LDraw's type 5 lines...

basically what i'm after is a file format that is easily transferrable bewteen
decent modeling programs, preferably *native* to each program's list of
exportable files, that way we don't have to each write converters for our
favorite modeler... i think LDraw2 should inherently support this file format
so that we don't spend time reinventing the wheel, writing converters and
hacking math in spreadsheets... let *one* person hack the math.. that person:
the programmer of LDraw2... once you're outside of that app you should model in
a modeler, export to a file format, have no tweaking necessary, then let LDraw2
handle all the processing of the information the modeling app provided... i
guess i'm saying that we don't need to replace LDraw (although we, YOU, sort
of already have with LDLite), or even the .DAT format... but we *do* need to
replace LEdit... this replacement is actually what i feel would be known as
LDraw2... and whatever it is replaced with needs to support a more robust file
format, that file format can then be converted by LDraw2 internally... it just
seems that we have too many people working in too many different directions
right now... let's get a central hub agreed upon, one that allows us each our
own tools, but also one that helps us end up in the same place, with a little
more standardization on the quality/methods that new elements adhere to.

i think the world already has a file format that would be great, that most
modelers recognize and that contains everything we would need...

DXF

most CAD/3D proggies use it, import, export... we could define standards for
objects/layers, etc that meet a spec that LDraw2 would know how to process as a
type 2 kinda line or a type 5 kinda line... and it's scalable in terms of what
people can afford... if you have access to ACAD, 3DS or MAX then great... jump
on board... if you can only afford freeware or shareware, then there's the
behemot editor or AC3D... just about anything that has CAD in its name is going
to handle DXF... the hard part is going to be breaking down the DXF format so
that LDraw2 can read it.... if only we had people in our community who knew the
DXF format intimately... hmmm there was that one guy... with the funny
nickname... hmmm... jooj or something like that???... :)

J



Message is in Reply To:
  Re: Feature requests for LDraw/LEdit v. 2
 
The *strength* of LDRAW is the Type 2 and Type 5 lines. They are key to making rendered instruction sheet output look good. AC3D is a modeller that supports both triangles and lines, with POV and DXF output. I encourage you to check it out at: (URL) (...) (26 years ago, 16-Mar-99, to lugnet.cad.dev)

16 Messages in This Thread:







Entire Thread on One Page:
Nested:  All | Brief | Compact | Dots
Linear:  All | Brief | Compact
    

Custom Search

©2005 LUGNET. All rights reserved. - hosted by steinbruch.info GbR