2023-05-02 11:06:56 is there any sort of C FFI for RetroForth (or Konilo)? Specifically I'd like to have something where I could push a pointer to a C function and some arguments onto the data stack, execute a RetroForth word to call the C function and get the result back on the stack. Looking at using RetroForth to write some games for the PlayDate: https://sdk.play.date/1.13.4/Inside%20Playdate%20with%20C.html 2023-05-02 11:50:30 not well documented, but retroforth has an optional ffi 2023-05-02 11:53:36 http://forth.works/temp/ffi.txt 2023-05-02 11:54:23 http://forth.works/temp/ffi-testing.tar (some old test code) 2023-05-02 12:02:26 great, thanks! will check it out 2023-05-02 12:06:07 I'll work on some better documentation this evening 2023-05-02 12:08:00 btw, I like the tweaks you made to the site. Is Konilo the evolution of the two non-nga VMs? 2023-05-02 12:40:15 yy 2023-05-02 12:41:37 ilo started as napia, which was a mulitcore vm. The initial work proved very buggy, so it was simplified into ilo, then expanded back into a multicore system 2023-05-02 12:42:16 konilo is the forth that runs on ilo & napia (and later arks; another ilo-derived vm one of my kids is working on) 2023-05-02 12:43:33 at this point konilo on ilo is my primary daily system for note taking, small programs, and interactive use 2023-05-02 12:43:53 alongside retroforth & nga for bigger things like my http servers, database work, and system scripts 2023-05-02 12:45:01 I'm also using ilo for expermi 2023-05-02 12:45:40 experiments with smaller targets like microcontrollers as it has the smallest requirements of my various systems 2023-05-02 12:58:03 I see, interesting. I've been thinking of porting one of your VMs to ARM assembly (my main machines are a Reform laptop (ARM A53) and a Mac Mini (M1)). 2023-05-02 13:11:18 it may need some adjustment for A53, but http://fossils.retroforth.org:8000/ilo/file?name=vm/ilo-aarch64-linux.s&ci=tip is an AArch64/ARMv9 implementation of ilo that I run on my phone, under Termux; I've not tried adapting it for Mm 2023-05-02 13:11:23 M1 yet 2023-05-02 13:12:45 (this did start as a disassembly of the compiled c implementation, so probably not optimal vs a hand written implementation) 2023-05-02 14:26:56 ah, neat. that'll be a good starting point