2021-06-16 05:09:45 Cool 2021-06-16 09:32:51 crc So I imagine that creating a distinction between a device ID and a device handle was so that there could be multiple instances of the same hardware device class, yeah? 2021-06-16 09:33:03 Eg: a VM with two block storage devices 2021-06-16 09:34:57 that, and it allows for a VM to selectively enable/disable devices without needing the image to hardcode the order 2021-06-16 09:50:05 Ah interesting. 2021-06-16 09:50:30 Also, I just raised a ticket on srht. Mostly opinion based. 2021-06-16 09:51:05 I have a hunch maybe the `-i` behavior is intentional to simplify usage with tools like `cat` and `echo`? 2021-06-16 10:19:12 I don't think it should be quitting on word not found in interactive mode 2021-06-16 10:24:47 It's not exiting for me in the repl when I enter an unknown word, either with or w/o -i 2021-06-16 10:34:24 Oh derp, actually, it exits on syntax errors, sorry. 2021-06-16 10:35:19 Sometimes when in REPL mode, I will forget a sigil on a data type (Missing `#` or `"` instead of `'` on a string) 2021-06-16 10:36:03 Oh wait... now it's not doing it. Maybe there are more steps needed to reproduce this one. 2021-06-16 10:36:12 It's done it a few times for me. One sec. 2021-06-16 10:36:37 Ah you know what? It's not on `err:notfound`. It's crashing on stack underflow. 2021-06-16 10:37:07 The syntax error triggers a word not found error, the REPL recovers, then the proceeding word is executed on an empty stack. 2021-06-16 10:38:05 I def. misunderstood the root cause, but the behavior is different than other Forths. 2021-06-16 10:38:17 I am heading out but I can update that ticket when I get back 2021-06-16 10:38:32 sounds good, thank you! 2021-06-16 10:38:48 ACTION is at work, so can't do much on it right now