richmond62 wrote: ↑Sat Oct 21, 2023 8:23 pm
I am wondering about the legality of putting patched versions inside DMG images and making them publicly available.
BUT I will defer to
FourthWorld first.
I am no lawyer but I don't see why a patched executable of a GPL'd released binary should be a problem, since we have here exactly what the changes do to the execution and 'source' hex patch for how to reproduce said changes to that binary, a binary which is reproducible from GPLv3 source, and which we should be able to update to produce the same outcome in future binaries. Does it matter that the same techniques used are also employed for reverse engineering copy protections? I mean this isn't 'cracking' a commercial app here.
I had actually thought maybe patching could be used to avoid the damn embedded registration stack in a similar way, changing a pointer somewhere, or stripping the whole embedded stack out of the binary completely (and we have the 'source' stack for pattern match), but then I thought working towards re-compiling from source would be better use of time (if I had any).
NOTE to self, I really need to read up on memory debuggers and dissemblers (probably said that like 20 years ago too, lol).
Anyway, the new executable would not be signed with LiveCode Ltd. sig. I've stripped all of the 'meta data' type info from these binaries (the windows binarie(s) too). As separate as can possibly be made (so far. still chipping away) from LCs thing (and they seem to be headed in a different direction anyway)... but as you touched on, there's not really any avoiding all of the baked-into-the-code references anyway, like all of the reverse-dns-style identifiers on most ever extension/widget/script/resource library that make up the IDE, and there is indeed references to RunRev and MetaCard in the source / stack scripts as well.
Officially, LC say they're proud of their 'open source' legacy, and we're pretty much that right here! They've even publicly pointed out the existence of OXT recently...righteous!