It's not going to be technically amazing, it may not even be possible with the package I am using to write on the C64 BUT I think it will be fun to start something new

A longgggggggg time ago I was starting to work on Berzerk for the C64, but this was being programmed in secret by others. So I abandoned this project as it was much further completed. But when I was working out how to do Berzerk my mind was on another Atari VCS favourite I had that the C64 never had an arcade conversion of.

So anyway Vanguard is not even unofficially ported to the C64, nothing like this game exists for the C64 and this is a shame because I think the horizontal, vertical and diagonal scrolling adds a lot to the basic Scramble type gameplay.

Recently those very helpful people at Lemon64 finally got Laser Basic transferred for use with VICE, so I can now technically start coding this in a system where I don't have to worry about corrupt 5.25" disks destroying my progress.

The speech is not going to happen, but I think apart from that I can actually do this game reasonably well after compiling the Basic listing with the Laser Compiler.

But what's the best way to start doing a game? I heard that back in the early 80s they always did the code for the game engine first with meaningless graphics and the did the graphics last. Do you think that's the best way? Map out the game logic and just dive in and write the code until it works and only after this is done do the graphics?

This be a coding thread, so i've moved it to the coding section. =-)



i'd have thought not to be honest; even if you ignore the long build cycle (you won't be able to properly test the code until it's compiled each time) there's not many compilers can crank out code fast enough to drive a scroller and all the ones i've seen it done with to date have been cross dev tools.



i tend to do something like that personally; most of my stuff starts life as a test to see if something specific is possible, which then has some dummy graphics and game logic strapped in to see how much CPU time is available for everything else.

I think you can have three extra tasks under interrupt which helps, and IIRC you can set sprites on a flight path independently as you can with music. 19kb free memory is a possible issue though!

I will write uncompiled game engine backbone code until it works then compile it.

Either that or massively improve my machine code skills

The speech, using only a really really (really!) simple compression[1], at 3900Hz fits into 34K, and 3900Hz is a reasonable bandwidth given the crappy bandwidth of the original synthesis.. Okay, it's not blinding, but there's a fair bit of it though, sounds better at both 7900Hz, and beautiful at 15600Hz, but there's not a lot of memory left at those rates

Hence, why I've found the urge to muck around with a speech synthesiser that can recreate the SC-01a sound, which would require a few mere K to do all the speech in Vanguard.. But that's nothing more than a few experiments in matlab right now.. If you fancy doing the game, I'll quite happily wedge the speech in somehow if anything useful comes of these experiments..

As for LaserBasic.. Good luck

[1] The one I've rabbited on about here and on AA that also reduces the interrupt rates as well, so even the 15.6KHz playback doesn't use anywhere near 15.6KHz interrupts per second.. Likewise for 7.8KHz and 3.6KHz

I'm curious to see if it is possible, should be similar speed to white lightning (older version and Forth based so compiled). Bet the code output is pretty horrible though to hack!

I'll just try level 1 and see how fast it is first