Think of it as similar to this: Fear what goes bump in the dark.
A pretty dependent clause?
gopher://verisimilitudes.net/12022-01-31
I was careful not to read any commentary before I had written my own. If I follow the approach is one of structuring the state-space from the top down, and using lookup tables at run-time. You must understand I'm a mathematician and learned to program from SICP. As such the conversion of a rule into ad hoc cases does not naturally sit well with me. It even depressed me slightly to read this; where has the spirit that lives in the computer gone? I must agree however that your argument is consistent.
To refine my thoughts: I've no opposition to the use of lookup tables, especially as an alternative to parsing. I agree that analyzing a state-space can be a useful tool, but disagree that it should be dominant. Having multiple perspectives on a problem tends to be beneficial. More dramatically I can't stomach ad hoc structuring of the state-space. I suppose my fear is that a large number of ad hoc modifications would be difficult to know, and unless reality dictates otherwise knowing is more important than computation.
I wasn't too keen on joining just to shill my work.
I doubt they would mind if it was relevant to academic PLT, but it might not be a good fit for that reason.
>>133
I would prefer he didn't. There was some discussion about this website in the beginning of the thread and it wasn't very interesting.