Re: [Clorb-devel] clorb-0.2 on CMUCL and CLISP

[Eric M. <ema...@ma...>]

>>>>> "ls" == Lennart Staflin <le...@ly...> writes:

  ecm> ok, I think the problem is in the CMUCL definition of
  ecm> read-sequence on socket streams. After replacing it by a simple
  ecm> version, everything works as expected.

  ls> Does it depend on the CMUCL version. The CMU Common Lisp release
  ls> x86-linux 2.4.19 8 February 2000 build 455, works fine for me on
  ls> Linux x86.

I don't know; I only have access to CMUCL on Solaris. I reported the
problem to the CMUCL maintainers, who have confirmed that it is a bug.
You may not see it on Linux: the problem seems to be that
#'read-sequence is mapped to a read(), so if the platform read() is
able to keep up with the socket you won't see the problem.

-- 
Eric Marsden                          <URL:http://www.laas.fr/~emarsden>