2026-08-04 13:16:08 thinking about how to keep track of garbage collected structures where they contain pointers to other garbage collected things 2026-08-04 13:17:21 maybe I can register them centrally somehow then sweep everything and readjust pointers based on a stored starting point 2026-08-04 13:18:50 pointers in the struct have to be tagged and registered somehow 2026-08-04 13:21:00 Yeah that is the big problem with any kind of 'smart' memory management, pointers to pointers 2026-08-04 13:21:03 Good luck 2026-08-04 13:22:32 Personally I would try and have unified data structures, so array, table, etc, so you can just walk those structures and know the layout and what's managed 2026-08-04 13:23:07 Thinking with my Lua hat 2026-08-04 15:39:08 MrMobius: there are a lot of different approaches to that problem 2026-08-04 15:53:50 ya the problem is figuring out what's "managed" and not. it would work fine to list those pointers specifically until you forget to list one and disaster ensure 2026-08-04 15:54:08 seems best to find a way where declaration tags the pointer as managed 2026-08-04 16:17:31 There are compact memory management available with garbage collectors available in Forth. 2026-08-04 16:17:50 -available 2026-08-04 16:35:10 Source? 2026-08-04 16:44:07 MrMobius: you might want to check out the Boehm-Demers-Weiser GC for C 2026-08-04 16:50:44 RPN in the wild: tektronix 7854 waveform oscilloscope uses RPN https://w140.com/tekwiki/images/c/ce/42AX-4682.pdf 2026-08-04 17:15:23 the X, Y, Z, T nomenclature is clearly HP-inspired 2026-08-04 17:18:09 of course. 2026-08-04 17:18:20 hp and tektronix both made 'scopes. 2026-08-04 17:18:30 they had nerds in common, I bet. 2026-08-04 17:18:53 engineers who used HP RPN calculators, no doubt. 2026-08-04 18:10:10 yes, certainly 2026-08-04 20:38:52 veltas https://github.com/AntonErtl/garbage-collection