Hi there!
I fear I'll end up programming 20 separate games, one for each level.
You know, it might not work out given the different level of skill we
all have but this might be an ideal project for a distributed approach.
Sure, sounds good to me!
Perhaps you can get a level done and then farm out levels which are
similar and move on to some of the more challenging levels.
I'm already thinking a lot about the framework for the whole game. With
this turning into a 16/32/64K project, one should really be very careful
with the design here. I'm thinking about one "master" 4K ROM here handling
the game loop and distributing tasks to the other banks.
Stuff like Jumpmans movement for example will be in the "master" bank, but
for the display kernel or level logic, it'll branch out.
Once there is a stable idea behind that with one level working, the doors
are open for others to join the project I guess...
Right now I'm just evalutating everything. My first task is gathering all
structure data for all levels, just to see how much data is in there and
also to eliminate a first bunch of undoable levels.
Just finished the structure for "Robots I". The top kernel is again
mirrored to save a few bytes. While the bottom kernel is assymetrical, it
has still a mirrored CTRLPF state, I'm just timing register rewrites here.
I also started packing all those kernels into macros, so I can quickly
reuse all learned techniques and it also will help a lot when it's getting
more complex (sprites, missiles), as all code has to be touched only once
- per macro...
Perhaps in a larger group we could tackle more of the game. I have to
admit that Jumpman style games are my favourite.
Maybe not "more", but definitely "quicker"...
All help is appreciated. If someone is interested, I can try to explain
how the data is transfered from screenshot to sourcecode.
My, it'd even help if someone would just play savestates for all levels of
the two games... ;-)
Greetings,
Manuel
Attachment:
robotsi.bin
Description: application/macbinary