Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Some beginner's questions & some objects

10 views
Skip to first unread message

UD14...@ndsuvm1.bitnet

unread,
Feb 25, 1988, 12:31:56 AM2/25/88
to

[a line for the line eater]

After getting a frantic message about my posting, I figured I'd better post
this followup to clear up some confusion. What I'll (eventually) be sending
out across the net won't be 3 meg worth of data--only after it is uudecoded and
dearc'ed will it equal (almost) 3 meg. I also won't be sending it out at once,
but in many smaller chunks. The person who wrote me chided me for not waiting
for replys on my posting before posting all of it--well, I actually am but
didn't mention it. I'm sorry if I've caused any consternation among netters--
I thought readers of comp.graphics really wanted more object postings, but if
I'm wrong I won't send the rest of the files. Please let me know what you'd
like--if you'd prefer, I can send the stuff out over a period of weeks
(although it would take a long time to get a complete file for decoding). It
will not be easy for me to send out individual objects or even sets of objects
(other than the ones I mentioned in my article) because I have limited on-line
storage and only a 1200baud uplink to my accout. Please let me know your
preferences.

Sincerely,
Scott Udell
UD14...@NDSUVM1.BITNET

[a line for the line eater]

Leo 'Bols Ewhac' Schwab

unread,
Feb 27, 1988, 11:05:18 PM2/27/88
to
In article <618UD140469@NDSUVM1> Scott Udell writes:
>I thought readers of comp.graphics really wanted more object postings, but if
>I'm wrong I won't send the rest of the files. Please let me know what you'd
>like--if you'd prefer, I can send the stuff out over a period of weeks
>(although it would take a long time to get a complete file for decoding). It
>will not be easy for me to send out individual objects or even sets of objects
>(other than the ones I mentioned in my article) because I have limited on-line
>storage and only a 1200baud uplink to my accout. Please let me know your
>preferences.
>
I grabbed the objects you sent, and after writing an extremely
disgusting conversion program, got them into VideoScape 3D format, and
looked at them. For those who didn't bother, here's what they were:

bigpyras:
A couple of pyramids on rectangular slabs.

city1:
An arrangement of buildings. Those with object editors may be able
to pick out stuff they like.

hudbldg:
A semi-oval "building" with a notch cut through the middle.

inset2, insets:
Two of the same pair of objects, located at different positions in
space. Their profile looks something like this (sideways):
___ _____________
|__===____________|

oval:
Extruded oval.

pentagon:
Pentagonal prism.

pointop:
Rectangular prism with a pyramid on top.

tallgrp:
Four longish rectangular prisms arranged in an "office block"
arrangement.

twotri:
A pair of three triangular prisms of differing sizes stacked on top
of each other, smallest on top (best way to verbally describe it).

van2, van3, van4, van5, van6, vanilla:
Rectangular pyramids of differing sizes and locations in space.

If you load all these up at once, you get a semi-reasonable city.
Why it was split up into several seperate files, I don't know. Perhaps the
individual objects have different colors? No color information was in the
files. BTW, it didn't look like San Francisco to me.

Judging from the way the object polygons and points were arranged,
I'd say that most of them were lathed out. In other words, most of the
objects were very simple, and could be generated using any self-respecting
object editor.

I now have the basis for a city, which I can hack around, lay
streets down, put windows in the building, etc.

I don't know whether or not other readers of this newsgroup would be
interested in the remainder of the objects you have. Judging from the list
in your first posting, I'd say that most of them are relatively simple and
wouldn't bother. However, if the shuttle or Enterprise model are any good,
I wouldn't mind seeing them.

Does CAD-3D allow you to dump objects in ASCII format? If so, I can
borrow an ST, dump the objects, port them over to my Amiga, and use this
sleazy program I hacked up to convert them into VideoScape, which I can edit
from there.

I'd hold off on sending them out for a week or so, and wait to see
what the Net response is like. What you've sent so far is, by comparison,
primitive, and people may scoff at you (don't let this bother you, though;
it's The Way of The Net). People may be willing to sit through the shuttle
or Enterprise, though...

I don't know; it's a judgement call. Sit back and see what opinion
forms.

_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_
Leo L. Schwab -- The Guy in The Cape ihnp4!ptsfa -\
\_ -_ Recumbent Bikes: dual ---> !{well,unicom}!ewhac
O----^o The Only Way To Fly. hplabs / (pronounced "AE-wack")
"Work FOR? I don't work FOR anybody! I'm just having fun." -- The Doctor

UD14...@ndsuvm1.bitnet

unread,
Mar 1, 1988, 5:15:19 PM3/1/88
to

[a line for the line eater]

Thank you for the input. Based on Leo's response and the response of
others, I've decided to keep posting objects but in a more limited fashion, and
over a longer period of time. Below I've listed the objects I think I'll send
(deleting some of the simpler ones from my original list). If there are any
among these that you don't think you want (or any that you really do want, or
any that I deleted from the list), let me know. Among the objects I've
dropped is the space shuttle--I finally looked at it "up close" and decided
that it wasn't as good as the one posted earlier. Also, the SRBs and the tank
aren't that hard to make, so I won't post those. Before I get on with the
list, let me answer a couple of other questions that were raised and pose a
couple of my own.
First: yes, CAD 3D does have an ascii output--in fact the listings you'll
be getting ARE the CAD 3D ascii outputs. I suspect it listed each of the
buildings separately because the artist (cadist?) drew the buildings
individually and then loaded and placed them in their spots in the CAD
universe. They could have been "added" together to form a single object, but
this process takes a long time in CAD 3D, and just loading/placing individual
objects gives the same effect to CAD 3D users. I don't particularly like the
ascii output it generates (it leaves in all the 'construction lines' it uses
for solid modeling), but haven't figured out a way to clean them up. One
possibility is to hack at the data files it outputs, or to write a 'desk
accesory' program that could communicate with the program directly: either of
these, however, are beyond my means at this time, so we'll have to make due
with what CAD 3D gives us. Another thing--several people have asked me for the
Lavi fighter object. I don't have that--I said that it exists for CAD 3D and
if you have access to that program, it can be gotten from Antic Software ($12
for "shipping and handling"). If I do get it, I'll post it (it's PD).
One last thing before I get on with the list: how many out there in
netland have trouble with the 'ARC' format? It seems that while many have
no trouble UUdecoding my files, there are some that can't dearc them (in fact,
some that have never even heard of ARC--mostly mainframe users, it seems). If
this is going to be a big problem, I can post the original ascii listings
instead of the compressed ARChive files. This would really stretch out my
postings because that's a lot more data.
Okay, on to the listing. In cases where an object's name doesn't describe
it, I've put a little description (if it's really describable). The objects
are:

arch, bishop (a chess piece), chalice1, cone_10, cone_20, cuberoom (used in
a CAD 3D animation demo, it kind of reminds me of a space docks in Star Trek),
cyl_16, cyl_3, cyl_4, cyl_6, cyl_8, disk (once you take the construction lines
out, it looks like a 3 1/2" disk), ellipse, F-14, hexagon, hollow, hyprbola,
jack1, king (another chess piece), knight (another chess piece), octagon,
parabola, pentagn2, pentagon, queen (last chess piece), saturn, saucer (looks
like a flying saucer), sphere1, sphere2, sphere3, star1, starspn, parabola,
torus3d (a really nice torus), torus_10, torus_14, torus_5, USS1701, cruiser
(a space ship), and some new ones: neat1, coolship (a really nice 'star
ship'), monitor (a computer monitor when you remove the construction lines),
and strwrs (a spherical death star with some tie fighters--I might just post
the tie fighters because the 'death star' is just a plain sphere).

If there are any from my original list that you REALLY WANT, I'll post those
too. Also, I'm going to see about posting of objects from our E&S Digistar--
I'll talk this week or next with a company rep. who will be visiting us on
what is and isn't public domain (some really nice stuff is up there).


One last note. Many have suggested that I set up some kind of file server for
these files. For those on arpa and usenet/UUCP, I can't--I'm on bitnet and
don't have any direct access to those nets (as far as I know). I could
possibly get a bitnet file server (but not likely for a fairly lowly undergrad)
but from my expieriance with them people on other networks have lots of
trouble accessing them. So for now I'll continue posting to comp.graphics, but
at a much reduced rate over what I initially suggested. If you have any
suggestions or comments, please let me know. It may take me a while to get
back to you because of the dual problems of spring break and the fact that
a new computer is being installed on my node and my system will be down quite
a bit over the next week or so. Despite this, I'll try to return any messages
that make it to me. Thanks a lot for your help.

Stuart H. Ferguson

unread,
Mar 3, 1988, 8:19:20 PM3/3/88
to

Scott Udell asks about the "work lines" in CAD 3D and how to remove them.

The binary format for CAD 3D, the "3D3D" format files, have 3 bits in them
that signal the lines of each triangular face as being "real" or "work" lines.
For some reason this is lost when using the ascii output format. If you
don't mind bashing on the binary format you could extract the data sans
work lines.

--
Stuart Ferguson (s...@well.UUCP)
Action by HAVOC (s...@Solar.Stanford.EDU)

Jim Parry

unread,
Mar 5, 1988, 12:55:23 PM3/5/88
to

I'm finishing up a program to convert Atari ST CAD-3D data files (.3D2)
files into ASCII data files - including the details such as which construction
lines are invisible, the color palette, lighting positions, etc.
In a while (say two weeks), I should have it ready for posting to the net...

-----------
Kibo (Jim Parry)
userfe0n%mts.r...@itsgw.rpi.EDU
user...@rpitsmts.BITNET
or wox me on AmonNet!
--
Kibo (with a long "i") kibo%mts.r...@itsgw.rpi.edu
Jim Parry user...@rpitsmts.bitnet
"Let's wox!"

Leo 'Bols Ewhac' Schwab

unread,
Mar 5, 1988, 6:10:43 PM3/5/88
to
In article <53...@well.UUCP> s...@well.UUCP (Stuart H. Ferguson) writes:
>Scott Udell asks about the "work lines" in CAD 3D and how to remove them.
>
>The binary format for CAD 3D, the "3D3D" format files, have 3 bits in them
>that signal the lines of each triangular face as being "real" or "work" lines.
>For some reason this is lost when using the ascii output format. If you
>don't mind bashing on the binary format you could extract the data sans
>work lines.
>
Yes, but this doesn't address the other question: How do you
take a set of (possibly randomly arranged) edges, and reconstruct them into
reasonable polygons? You could play all sorts of games by comparing edge
endpoints to each other, but that seems like a messy solution, possibly
unreliable, and if your object is at all complex, it could take a long time
to figure out. (Is this an n^2 problem?)

UD14...@ndsuvm1.bitnet

unread,
Mar 7, 1988, 6:48:37 PM3/7/88
to

[a line for the line eater]

With all the responses I've gotten back from various locations, it's
fairly obvious that my previous method of posting my objects isn't acceptable
to a large chunk of those people wanting the objects (either I'm not using
their compression format or the objects are too big any way). Since all of you
seem to have UUdecode capabilities I've decided to just post the CAD 3D
binaries in UUencoded format and let you hack on them yourselves. These
binaries should be much smaller than even the arced ascii listings I was going
to post. This should be a resonable solution considering Jim Parry's
CAD 3D-to-ascii conversion program (sources? binaries?) and Stuart Ferguson's
file format information (Stuart, could you post a little more detail on CAD
3D's file formats?). If I don't get any complaints, I'll start posting in a
couple of days (finally!). I will still spread it out to keep traffic loads
down.
Scott Udell
UD14...@NDSUVM1.BITNET

A private message to "j" at the Astronautics Technology Center
Needless to say, I forgot your name--sorry about that! My message to you
got bounced back (it said you didn't exist, or something like that). Anyway,
I'd love to get anything you have (including your Amiga! :), and thank you
muchly for the offer.

Jim Parry

unread,
Mar 9, 1988, 8:48:51 AM3/9/88
to
In article <646UD140469@NDSUVM1>, UD14...@NDSUVM1.BITNET writes:
> to post. This should be a resonable solution considering Jim Parry's
> CAD 3D-to-ascii conversion program (sources? binaries?) and Stuart Ferguson's
> file format information (Stuart, could you post a little more detail on CAD
> 3D's file formats?). If I don't get any complaints, I'll start posting in a


I'll be posting the conversion utility here, as source -- posting it as
binary would be fairly useless, since then it would only convert from CAD-3D
to MTS XYZPlot format. Also, it's got a major bug at the moment. If I can't
work around it, at least then someone else will.
The program will be more of a 'how to convert' tutorial than an actual
utility, since I'm sure nobody out there is using the same soft/hard-ware
I am...

Ralph Martin

unread,
Mar 15, 1988, 6:31:27 AM3/15/88
to
Posting-Front-End: GNU Emacs 18.47.1 of Mon Nov 23 1987 on v1 (berkeley-unix)


In general it is not possible to take a set of edges and from those make a
unique set of polyhedral faces - a wireframe can represent more than one solid
object. See any introductory text/paper on solid modelling.

0 new messages