Re: Better support for transition to guile-1.6

[Tom Lord]

      On the other hand, achieving such freedom has a large impact on
      client programs.

Not necessarily -- in the sense that something close to the existing 
API can certainly be implemented on top of the cleaner API, for those
projects where GC robustness and fast call/cc are not important.

My question is about the number and degree to which people are relying
on the current API.

        We can't wait until next year.

That's understandable.  I'm still interested in how far the current
API has spread.

        I definitely agree that this has been a weak point in Guile
        development: We should have put more thought into the API from
        the start (many times).

Absolutely.

-t