Post by rmartins on Mar 14, 2016 19:39:35 GMT
After using 6 vacation days, and 4 weekend days (in a total of 10 days,sleeping less than 6 hours per day), dedicated to give a boost to Steel Ball completion, I'm afraid I will not be able to provide a playable version by 15 March.
I knew I started with some handy-caps:
- only found out about contest in mid August 2015
- had no previous solid experience with Z80 Assembly in particular
- was not aware that ZX CPU frequency limits were that severe.
- use no pre-packaged libs
- use direct assembly
- have a big goal ( Physics based game with almost full screen scrolling )
CPU limits, eat most of my time, that had to be dedicated to optimizations and refactoring lots of code.
Using Zeus was fun, and good for quick tests, but severely limited development, once I got about 20KB source code into ZX Memory.
Since early February, I switched to using Pasmo, which increased productivity and removed the memory limits I had (with Zeus).
Tracking some obscure bugs in assembly, takes ages, and sometimes, is a very silly thing, like deleting an instruction in editor, by mistake.
Sometime it's not a "bug" at all, but interference from an external source, like ROM interrupt screwing memory data because I used IY register.
I also had to implement a Music player (implemented 3 distinct versions), to understand the limitations, which took some valuable time, investigating and coding.
I would like to public thank UTZ for his feedback and composing music for it.
I also used lots of time, to make and improve graphics, which I hope will show the effort put into them.
All these things had to be learned and done, but all this extra efforts skewed my initial estimate for game completion of about 6 months, in part time.
I'll be sure to finish it, but not for tomorrow (15 March).
Most of all, I want to make a good game.
Have to get back to coding now
I knew I started with some handy-caps:
- only found out about contest in mid August 2015
- had no previous solid experience with Z80 Assembly in particular
- was not aware that ZX CPU frequency limits were that severe.
- use no pre-packaged libs
- use direct assembly
- have a big goal ( Physics based game with almost full screen scrolling )
CPU limits, eat most of my time, that had to be dedicated to optimizations and refactoring lots of code.
Using Zeus was fun, and good for quick tests, but severely limited development, once I got about 20KB source code into ZX Memory.
Since early February, I switched to using Pasmo, which increased productivity and removed the memory limits I had (with Zeus).
Tracking some obscure bugs in assembly, takes ages, and sometimes, is a very silly thing, like deleting an instruction in editor, by mistake.
Sometime it's not a "bug" at all, but interference from an external source, like ROM interrupt screwing memory data because I used IY register.
I also had to implement a Music player (implemented 3 distinct versions), to understand the limitations, which took some valuable time, investigating and coding.
I would like to public thank UTZ for his feedback and composing music for it.
I also used lots of time, to make and improve graphics, which I hope will show the effort put into them.
All these things had to be learned and done, but all this extra efforts skewed my initial estimate for game completion of about 6 months, in part time.
I'll be sure to finish it, but not for tomorrow (15 March).
Most of all, I want to make a good game.
Have to get back to coding now
