Linus_Sch wrote: ↑Wed Sep 28, 2022 5:45 pm
o me the spirit of the programming model of the 32sii assumes that you only really have one program in it, and most of what we're discussing with regards to this significantly steps away from that. A big change no matter how it is done.
Well, the 32S/SII were designed to have a few programs (saying one program would be exaggerated). In 384 bytes, i.e. 256 steps max., less if you use variables, you can't do much more.
For instance, a 32S user (like me) could reserve LBL A to LBL F to 6 possible main programs (short of course) , LBL X, Y, Z to scratch pad programs, and the rest for 'local' labels (loops, jumps, local sub inside the programs).
Now, we have much more program space, and (hopefully) a way to backup them. So we can extend (a bit) this model to a maximum of 26 (or ideally 52 with lower case LBL) programs, each using internally local numeric labels for loops, jumps, and local subroutines.
Local labels 0 to 9 and .0 to .9 sounds good to me. Note that my vote is on displaying the latter set with the dot and not as 10 to 19, so that it matches what I key in.
I disagree, it should be displayed as 10 .. 19 ,as are already doing FIX/ENG/SCI and SF/CF/FS?.
For line numbers definitely let go of using the dot and the comma though, just add the third digit when needed and let them go to 999.
SM will have to find a solution to fit a 3-digit line number in the same 12-place display as the 32S/SII
A option may be to get rid of the alpha character (the name of the program) when the line number is >99. It will not happen so often, I guess, the DM32 typical use should be for writing short programs.
I.e. A00 A01 ... A99 100 101 .. 199 200 ..
It was almost never happening to go beyond 99 steps on the 32SII (except to deliberately test the feature), so it would not really break anything, and it could now happen for DM32 power users.
J-F