2023-03-05 11:24:35 ae_chep: the guix .scm file works for me. I made some changes (setting to current release version, adding the url for the tarball, and proper hash value & license identifier); my copy is currently at http://forth.works/retroforth.scm 2023-03-05 12:13:59 :) Hey, awesome 2023-03-05 12:14:35 The hash doesn't matter when pulling from local directory but when fetching data remotely it is checked against 2023-03-05 12:14:50 Do you want me to submit it for it to appear in the main channels? 2023-03-05 12:20:46 and does retro-repl have a better error handling than `retro -i` ? I would like it to not crash when I create fatal errors (of which I do plenty) 2023-03-05 13:59:04 might be better to wait until the next release (about 2 weeks out) 2023-03-05 14:03:44 `retro-repl` has less error checking than `retro -i` 2023-03-05 14:04:24 fatal errors should mostly be caught and reported on exit (at least for stack over/underflow or things triggering signals) 2023-03-05 14:07:09 I was doing something like "#3 #2 / + 1 / 200" and had to restart the REPL about 5 times until I got the prefixes for values and arithmetics-words right 2023-03-05 14:08:15 And basically my question is; is this the intended experience or just the state of things as they are? 2023-03-05 14:09:13 shutting down on errors is what I've done for years. Not sure it's the best way, and it's something I'm open to changing. 2023-03-05 14:43:43 I was _imagining_ a forth development experience to be like a conversation had with the interpreter. In which case to be hung up on is rather unpleasant 2023-03-05 15:14:05 makes sense. I've setup an issue (at https://todo.sr.ht/~crc_/retroforth/94 ) for gathering feedback on this. I'm quite willing to make changes here to make the overall experience better 2023-03-05 16:15:51 And I think I'll pause a bit with the practical exercises and study forth theory in the mean time 2023-03-05 16:16:18 have a good (next) week! bye