Hi Eric, Just a thought on using “undocumented” function calls.
> On Jul 28, 2021, at 6:40 AM, Eric Auer <e.a...@jpberlin.de> wrote: > > > Hi! > >> #### reserved to condition for example yes, no, quit ####> 0.0:Y> 0.1:N >> #### Space reserved to file diskcopy.c #### >> (...) >> 1.29:image file (Y/N)? >> 1.30:disk (Y/N)? >> 1.31:Copy another disk (Y/N)? > > Note that yes/no questions actually have kernel support. This > will not help for more complex choices, but it is good to know: > >> INT 21 U - DOS 4.0+ - DETERMINE IF CHARACTER REPRESENTS YES/NO RESPONSE >> AX = 6523h >> DL = character >> DH = second character of double-byte character (if applicable) >> Return: CF set on error >> CF clear if successful >> AX = type >> 00h no >> 01h yes >> 02h neither yes nor no > > This is controlled by your COUNTRY config sys settings. > It only helps with parsing, not with showing the question. > > Cheers, Eric During those days when DOS was the primary user operating system, I always felt it was a really bad idea to lean on any undocumented functions. But nowadays, I think I probably have changed my mind. All the “big” DOS platforms of that time are long dead and no longer get any form of updates or changes. Which brings up the minor philosophical point… Should new programs even worry about supporting those dead DOS platforms? While I feel it is nice to be able to run a new DOS program on old PC/MS/DR/*-DOS, I no longer feel it is absolutely necessary. FreeDOS needs to be able to run MS-DOS programs. But, do programs for FreeDOS need to run on *-DOS? It’s nice. But, I don’t feel it is mandatory. I’ve seen “bug” reports on SourceForge that go something like… “program x doesn’t run on MS-DOS”. I find myself thinking “thats not a bug, it’s a feature request. so what? why should I care?” :-) Jerome _______________________________________________ Freedos-devel mailing list Freedos-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freedos-devel
