I think we have a slight change in opinion but let me fully explain my thoughts and maybe we can come to a compromise.
0-all of the x86 programs I compiled the last few days work on x86 except for Phyllis but I was reserving judgment until/if IDFB has a chance to test on her device.
1-In my experience, most programs I compile that are fairly simple, not an emulator or otherwise dependency heavy about 80% of the time just indeed do work
(obviously assuming they compile successfully
). I only compile for cpu architectures or versions not previously available. For example x86 or trying the CE2.0 EVT SDK for MIPS, etc.
I have stopped trying to do the complicated programs because it’s too much effort and beyond my knowledge/scope.
2-I have been testing the builds on the devices I have but some I’ll never have
(for example when I add builds for SH3/SH4/or CE4+ MIPS
). Generally speaking if it works on x86 it should most likely work on those others. But I will delete them if they don’t work. I’ve been adding to the SCL with the verbiage you gave me “ProgramName-CE300-X86-beta”.
3-generally speaking, there’s three situations for compiling the code on new architectures
3-i : code compiles successfully with no modifications other than “adding a new configuration”
3-ii: code compiles successfully after adding a new configuration plus a minor code / line change
(like the HPEN issue you helped me solve
)
3-iii : code compiles successfully after anything more than minor work. Example what @thenzero did to get Hot Death Uno running.
So in short:
I agree that situations 3-ii and 3-iii should have their source code both be in the Github. In which case I’ll start working on those this week.
I agree that situation 3-iii actually warrants a new “beta release version tag” separate from the original version in the scl. Just like Hot Death Uno. Or like that npopuk.next.
Or I could going forward have both situations 3-ii and 3-iii go into a separate scl version listing. I was just worried it would clog up the scl if the versions are effectively 99% the same
I always respect you and will follow your final say. I just wanted to explain my perspective on the matter.
Edited by torch 2024-03-16 6:45 PM