Correctly or incorrectly, Tratt concluded that backtracking in Icon was more difficult to use than one might hope, and reduced its scope in his language.
It’s a very clearly written article, I did not know about the goal directed execution, and the description is very approachable. It’s an interesting variant on handling control flow.
I think I first read about Icon via Laurence Tratt’s Converge, which borrowed the idea of goal directed execution: http://tratt.net/laurie/research/pubs/html/tratt__experiences_with_an_icon_like_expression_evaluation_system/.
Correctly or incorrectly, Tratt concluded that backtracking in Icon was more difficult to use than one might hope, and reduced its scope in his language.
It’s a very clearly written article, I did not know about the goal directed execution, and the description is very approachable. It’s an interesting variant on handling control flow.
A more full-featured language based on Icon. Icon itself was interesting for goal-directed execution. So, I’m submitting them together in this thread.