Hello Ruud, * On Fri, Jan 09, 2015 at 02:50:52PM +0000 Baltissen, GJPAA (Ruud) wrote: > So I started to think of an overlay system. You start with one main file, > version 1. And the rest is superposed on that one. So the files for the other > versions only contain the changes. The first disadvantage: an assembler will > need more time to come the the end result. Do you really want to consider the time to assemble with ca65 on a multi-Gigahertz machine? I would neglect the time needed. > And what about readability? It is clear to me that I should be able to generate > various files again as I did with the 'one-source-code-file'. Unfortunately > this could be a problem. When overlaying bytes, you can pinpoint the address > where the overlay start using the '.org' or '*=' code'. But that isn't possible > with text. I had the same problems. That's why I came up with the approach I outlined before, using the linker of ca65. With that, you can do it rather nicely. The linker implements what you call the overlay approach. Regarding the addresses: Here, I agree with Kajtár: The listing file should be enough. Unfortunately, here is currently the problem with my approach: The listing file only shows addresses relative to the segment. That is, for every segment, it starts at 0. However, I already started modifying the assembler to generate an additional file with position information. Next step would be to generate an additional file with the linker, and a third program to combine the two files and the assembler output into a "real" output which shows correct addresses. If I only had more time. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ Message was sent through the cbm-hackers mailing listReceived on 2015-01-09 20:00:03
Archive generated by hypermail 2.2.0.