On Apr 6, 12:58 pm, "Edward K. Ream" <[email protected]> wrote:

> Anyway, you should be aware of how I am spending my time lately...

Pass1, which includes code generation, takes 0.06 sec to process
leoNodes.py.

Given the number and simplicity of the generated data structures, I
suspect local analysis for this file will take less than 0.1 sec.

Speed of global analysis depends on the number of source files.  I
expect linear performance, so it might be possible to fully analyze
Leo's 30+ sources in under 10 sec. We shall see...

The code now seems completely natural to me. I could put it on the
shelf indefinitely, should that seem desirable, and pick it up easily
again much later.

Edward

P.S.  I have been making a mental catalog of the places in Leo's
sources that could cause problems for type analysis.  Presumably,
these will be the places for lint_assert statements.  It does seem
like such assertions cut the Gordian knot of type inference.

EKR

-- 
You received this message because you are subscribed to the Google Groups 
"leo-editor" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/leo-editor?hl=en.

Reply via email to