When running newLISP-GS, the newLISP monitor process in the lower part of the IDE is controlled using STD I/O. This is why 'read-line' and other built-in functions processing STD I/O channels should only be used when running newLISP in a terminal shell.
Thanks - I think I understand now what is going on.
Since my interest in newLisp is for natural language interaction, I guess I will work in a terminal until or unless I have a simple graphical interface for the user.
> (set 'z (parse (read-line)))
this is a test
("this" "is" "a" "test")
>
I confess I don't use newlisp-gs - i prefer the full-on editing environment of a text editor like BBEdit, using the terminal for interaction. I would - not very convincingly and only if pushed - argue that it's better to meet newLISP first of all in the simplest environment possible (the terminal), then move up to editors and stuff later. Something to do with cognitive loading... :)
OK, techdir0, I understand why it didn't worked for you and worked for me.
1. Write (parse (read-line)) and press enter in editor part of IDE. 2. Select (with mouse) and copy (parse (read-line)) together with enter from editor part of IDE. 3. Paste in dialog part of IDE. 4. Write the statement you want parsed and press enter. 5. Goto 3.
---
Lutz will maybe fix it when he came back from travelling. I also use GS IDE only for short programs, and Scite for longer programs. I described how it can be customized for Newlisp in one post in blog.
I think I will go along with the majority opinion, and stick to the terminal view. I expect I shall end up using TextPad as my
editor, since I'm used to it, and I can modify the syntax files to suit newLisp.