[cap-talk] Joe-E

Mark Miller markm at cs.jhu.edu
Mon Nov 1 21:35:17 EST 2004

Jonathan S. Shapiro wrote:
> From conversations with MarkM at the time -- and I may have
> misunderstood him -- the "original E" project (which appears to have
> morphed into Joe-E) was one of those asymptotic convergence problems. It
> was easy to get the first 80%, then feasible to get 80% of the rest,
> then there was hard stuff, and then it all came apart at the seams.
> MarkM, if I understood him, felt that it could not (in a practical
> sense) be done.

That's a correct summary of my opinions from that time.

> If I misunderstood, or if new insights have emerged, I would be very
> interested to hear about them.

New insights have indeed emerged, as has new code (thanks to Chip). I would 
indeed like to explain these, but not this week.

In summary, I believe Joe-E to be immensely practical and, given the 
investment Chip's already made, fairly easy to do.

On e-lang, Mark van der Kraan has proposed another approach that sounds promising.

Text by me above is hereby placed in the public domain


More information about the cap-talk mailing list