CSB - The Movie

[Paul Stevens]

I have finally succeeded in recording an entire game of CSB so that
it can be played back. Before I managed to get the record and playback
precise enough I had played the entire game three times without any saves.
This is because there is no way to interrupt and restart the recording.
You can imagine how many times I had to face those three worms in the
starting room before accomplishing this feat. The last time took me about
12 hours. I always play with a single Champion - Toadrot, in this case.

At any rate, the recording is available if anyone wants it. It, the savefile,
graphics, and the executable take about one MB compressed. And of
course the executable can be used to make your own CSB movie. There
is alxo an option to replay at about 16 times normal speed so that the
entire game is played in 45 minutes. Exciting stuff. ; - )

I would like someone with a machine faster than mine ( 166 MHz) to give it
a try because timing is so very critical in getting the playback precisely
the same as the original. ( I also discovered a Microsoft compiler bug.)

I f anyone could try it out, let me know and I will find a place to put it
temporarily to download.

PAul ------- prsteven@facstaff.wisc.edu

[ChristopheF]

I can put the file on my site if you want.

[DjArcas]

I haven't seen it, but my friend has, he said the ending was very predictable, and that the acting was poor...

[ChristopheF]

You can download the file here:
http://dmweb.free.fr/CSBForWindowsLinux.htm

I tried it on my Athlon 1.2 GHz, no timing problems. But I only watched a few minutes (with and without QuickPlay).

[beowuuf]

Same here, but on slower machine than that : )
I've let it run mostly on fast mode twice now to the end, and absolutely no timing problems

[Paul Stevens]

I should explain that any timing problems will not be seen by
the movie-goer. They simply crash the playback process. I built
a lot of error checking into the playback to catch any differences between
the original and the copy. And, believe me, there were a lot of problems.
Some of them were sensitive to whether I was using the 'debug' or 'release'
version of the compiler, for example. But I think the result was worth the
effort because it did catch so many subtle problems in the original
Atari code, in my translation of it, and in the Microsoftt compiler itself.

There is what is referred to as the 'butterfly effect' in chaos theory.
If a single monster on any level turns once in a different direction or delays
his movement by one tick of the clock then the entire dungeon changes very
quickly. This is because almost everything is decided by the random number
generator and when a monster is in a slightly different state then he is likely
to soon cause one fewer or one more calls to the random number generator.
After that happens, everone else gets different random numbers than they
got in the original. And all is lost. So the copy must be perfect to work at all.

And thanks for testing it for me,

PAul

[ChristopheF]

I also let the game run until the end with quickplay twice. No problem encountered, and it took less then 30 minutes to complete.
Good work!