Re: [MIT-Scheme-devel] scheme/edwin build problem?

[Taylor R Campbell]

   Date: Mon, 7 Nov 2011 17:49:29 -0700
   From: Matt Birkholz <address@hidden>

   The error it was signaling came from the set-float-environment
   primitive, which was not happy to be applied to a 16 byte vector-8b.
   Apparently sizeof(fenv_t) was 16 when all.com was dumped.

   Unfortunately sizeof(fenv_t) is 28 bytes here.

Chris recently changed Scheme to preserve the floating-point
environment across band save/restore, but floating-point environment
objects are OS-dependent.  If we want to preserve things that way,
we'll need to invent an OS-independent representation.