16:45 <lisbeths> I am working on my tokenized forth in c still. I am not ready to move it over into asm.

16:48 <APic> A big Problem of Forths is Stack Overflow 😉

16:48 <lisbeths> My three biggest problems: c doesn't want to work in words of memory, it is hard to make a syntax where every feature of forth is one letter, and also the runtime to c on my system is 16 kilobytes but the rest of my code is only 360 bytes

16:49 <APic> Good Luck

16:54 <lisbeths> I wish there was something a bit more minimal than c in order to put my forth in. I have considered fortran which might be faster but it isn't any smaller. I want something that is portable.

17:00 <APic> You could try one of https://esolangs.org/ 😉

17:03 <lisbeths> The way I am handling my stack is that it is a character array and if I pop an int I pop sizeof(int) bytes and if I push a void* I push sizeof(void*) bytes. Although this means I can not use swap.

17:07 <APic> ☹

17:08 <APic> Probably You need a union?

17:09 <lisbeths> I have been told unions are slower under the hood.

17:13 <APic> You cannot say that generally. Just get Your Project to work, then profile, then optimize

17:18 <lisbeths> I find character array to be more elegant than using some piece of code I don't understand.

17:24 <APic> lol

19:19 <crc> re runtime size: just write in assembly and/or write your own libc and startup code...

19:22 <crc> what systems do you plan to run your forth on? is the ~16k overhead significant on them?