Thank you gentlemen!
The question is ... what next?
I know that I need to give you a list of all of the text messages, so that you can translate them into Hungarian and Spanish.
Apart from that, should we think about adding any new capabilities to IS-DOS?
We cannot add compatibility with CP/M 3, because some of the CP/M 3 BDOS calls conflict with the existing MSX-DOS 1 BDOS calls that are already in IS-DOS.
Apart from that, CP/M 3 is very much tied to the CP/M disc format, and to using user-numbers instead of directory names.
We could look at adding MSX-DOS 2 compatibility, but I would need considerable help with that, because a lot of the extra MSX-DOS 2 functions deal directly with the disc, and so would need to be translated into EXDOS FISH calls ... and I have absolutely no experience with the EXDOS FISH system!
On top of that, MSX-DOS 2 really relies on the addition of "Environment" variables, similar to those in MS-DOS.
Adding some buffer space for those environment variables would mean needing to allocate another segment in EXOS, or adding some extra memory requirement to one of the existing drivers.
Either way, this will likely use another EXOS segment. There is still one segment free for use on a 128KB Enterprise, but is this how we want to use that memory?
For the environment to be useful, I think that it would need to exist all of the time that IS-DOS is running.
If we do add MSX-DOS 2 compatibility, it would make IS-DOS a nicer system for programmers, but I don't think that it would make any classic MSX applications available on the Enterprise, because there are very few MSX-DOS 2 utilities worth having, and MSX applications rely on the existance of the MSX BIOS and hardware.
Another possibility is to add some limited compatibility with the ZCPR3 system.
Because ZCPR3 was built on CP/M 2.2, it would not conflict with our existing IS-DOS BDOS and BIOS, and it would allow for some commercial CP/M programs to activate built-in enhancements for ZCPR3 (such as Z80ASM, SLRNK, and WordStar 4).
But just like CP/M 3, a lot of the ZCPR3 improvements are very reliant on the old CP/M concept of disc user-numbers, rather than directory names.
Does anyone have any thoughts?