french@caip.RUTGERS.EDU (01/13/86)
From: "french robert%c.mfenet"@LLL-MFE.ARPA
Before I start on the problems I've been having, let me thank RJ Mical for
turning my first attempt at Amiga programming into an incredible graphics
system. It's amazing how much you can learn from looking thru code that
stiil bears your name somewhere! The DPAINT interface is really neat.
(While I'm at it, I congratulate Electronic Arts for DPAINT, easily their
magnus opus to date. It ought to sell a few Amigas by itself!)
RJ - I wish you the best of luck in the future, and await your next
software releases.
Now for the problems...
I have been having the same problems as Bernie Cosell <cosell@BBN-PROPHET.ARPA>.
The Lattice bug with consecutive floating point assignments was mentioned
in Bob P's fast floating point demonstration program, but I never really
realized what he meant until now. I wonder if 3.03 has this problem?
Also, I had the same problems with enumerated data types in workbench.h
and performed almost exactly the same surgery. Everything seems to work
OK now, except that the computer crashes immediately after writing the
screen in IFF format. It doesn't close the files, but they may be recovered
by copying them to another file. Finally, ever since Mandelbrot 2.00, I
have been unable to draw a picture in INTERLACED mode. Looking thru the
code, I noticed the distinct absense of the line:
if (color_mode & INTERLACE_MODE) {
ns.Height = 400;
ns.ViewModes |= INTERLACE;
(^^^^^ this line missing ^^^^^)
}
in open_winds() in mand3.c. Apparently this was dropped in my original
transmission, and was never replaced.
Thanks again RJ for all of your improvements. BTW, why ARE the colors
GRB?
Your fractal and mine,
Robert French
ARPA: French#Robert%d@LLL-MFE
Well: french
UUCP: ihnp4!ptsfa!well!french
CC : info-amiga@rutgers.
cosell@bbn-proppapa@uscvax.UUCP (Marco Papa) (01/16/86)
> Now for the problems... > > I have been having the same problems as Bernie Cosell <cosell@BBN-PROPHET.ARPA>. > The Lattice bug with consecutive floating point assignments was mentioned > in Bob P's fast floating point demonstration program, but I never really > realized what he meant until now. I wonder if 3.03 has this problem? > Also, I had the same problems with enumerated data types in workbench.h > and performed almost exactly the same surgery. Everything seems to work > OK now, except that the computer crashes immediately after writing the > screen in IFF format. It doesn't close the files, but they may be recovered > by copying them to another file. Finally, ever since Mandelbrot 2.00, I > have been unable to draw a picture in INTERLACED mode. Looking thru the > code, I noticed the distinct absense of the line: > > if (color_mode & INTERLACE_MODE) { > ns.Height = 400; > ns.ViewModes |= INTERLACE; > (^^^^^ this line missing ^^^^^) > } > > in open_winds() in mand3.c. Apparently this was dropped in my original > transmission, and was never replaced. ... > Your fractal and mine, > Robert French > I just applied the surgery suggested by Bernie Cosell and Robert French and got the thing working (I mean the program compiles, Presets() works, and Interlace mode works also). But as Robert suggested the Save function does not work at all. The behavior on my machine when invoking Save (for a 320x200 image) is as follows: 1. the disk starts spinning and spins for about 20 minutes !!! 2. sometimes I get back the CLI. sometimes the machine locks and I have to reboot. 3. When I look at the directory two files are there: MandPic.xxxx.x and MandPic.xxxx.x.info (the x's are numbers depending on the Mandelbrot window, I think). 4. Both files are "Empty". I tried, as suggested by Robert, to copy the files to other files. The first one got copied in about 15 min (and ended up being ~40K, and the second one (the .info file) ended up being more than 100K!!! I tried to load them back into the Mandelbrot program, but with no success. Now, I would really suggest that anybody that submits a program to USENET or other public networks, first checks whether his version will run properly with the version of the compiler that everybody has (that is, 3.02). I have been waiting for the documentation and software from Commodore for 3 months now, and got my Lattice C 3.02 disks by showing my developer acceptance letter from Commodore. Naturally I still have no documentation, and have bad feelings about it when I consider that its has been 3 months since I saw the message from somebody at CA saying that the 1.1 manuals were at the printers, and a week ago I see from another guy at CA that "maybe" the 1.1 manuals are at the printers. I know of no dealer in the LA area that has the Lattice C version 3.03, and all developers that I know also have 3.02, and have sent out their update letters long ago and received nothing. By the way, I love the Mandelbrot program. It is full of great examples on how to use Intuition, the workbench, the various graphics modes, gadgets, anything one can think of. Thanks to Robert and RJ Mical for a great job. Marco Papa USC - Computer Science Dept. UUCP: ...!{decvax,ucbvax,hplabs}!sdcsvax!sdcrdcf!uscvax!papa ...!{allegra,trwrb,trwspp,ism780,cithep,oberon}!uscvax!papa CSNET: papa@usc-cse.csnet ARPA: papa%usc-cse@csnet-relay.arpa
papa@uscvax.UUCP (Marco Papa) (01/17/86)
Another problem that it is still present in the latest version of the
Mandelbrot program is that, when invoked by the workbench, the Quit command
locks the machine. No Guru meditation, but after a few seconds the cursor
blocks and one has to reboot. Does this happens also under Lattice C 3.03?
Marco Papa
USC - Computer Science Dept.
UUCP: ...!{decvax,ucbvax,hplabs}!sdcsvax!sdcrdcf!uscvax!papa
...!{allegra,trwrb,trwspp,ism780,cithep,oberon}!uscvax!papa
CSNET: papa@usc-cse.csnet
ARPA: papa%usc-cse@csnet-relay.arpa