2026-04-26 16:39:32 <atami       > hello. for people who have written a forth system, how do you handle multiple source files? (like in the case of an include statement). im working on a forth system, and im unsure how to go about it
2026-04-26 16:41:34 <atami       > a block file system would make things easier, but i'm targeting 32-bit linux, and i'd rather avoid reading whole source files into memory
2026-04-26 17:01:52 <smlckz      > atami: 'avoid reading whole source files into memory' why?
2026-04-26 17:15:52 <atami       > currently, i read source code from an input file page-by-page. i read a line from that page buffer into the TIB, and if that line intersects a page-boundary in the file, i read the next page from the file, and continue accumulating the line into the TIB. whether i'm reading from stdin, a regular file, or a stream socket, or a bunch of files and
2026-04-26 17:15:53 <atami       > streams concatenated together, it's all treated the same
2026-04-26 17:17:39 <atami       > and while realistically source files wont be very large, i also dont have to impose any file size limit, which is another advantage
2026-04-26 19:41:43 <atami       > is it better to read whole files into memory? are input streams then treated differently, since you dont know how much data will be read in ahead of time?
2026-04-26 21:59:38 <tpbsd       > https://techtrenches.dev/p/the-west-forgot-how-to-make-things