From octave-maintainers-request at bevo dot che dot wisc dot edu  Thu Feb  1 16:09:19 2001
Subject: Re: ISO C++ and Octave
From: Mumit Khan <khan at NanoTech dot Wisc dot EDU>
To: octave-maintainers at bevo dot che dot wisc dot edu
cc: khan at xraylith dot wisc dot edu
Date: Thu, 1 Feb 2001 16:09:17 -0600 (CST)

fyi, does look like I finally have 2.1.33 working with gcc-2.97, except 
for of course anything that uses formatting, and as a result, you'll 
see 2 regressions at the very least -- sscanf-1, hex2dec (and perhaps 
something else). Btw, John, you may want to reword the comment in 
octave_scan (oct-stream.cc) ...

Turns out the only way I could find the problems, given that gcc was
being of no help whatsoever, was to first make Octave buildable under 
Sun Workshop 5.0, and then use Sun Pro's much better diagnostics to 
track and fix. My head hurts.

Speaking of Sun Workshop, I get the following:
  
  octave> sqrt (-1)
  ans = 6.1232e-17 + 1.0000e+00i

Obviously this doesn't tests like ``sqrt (-1) == 1i'' and tests fail.

  octave> eps
  2.2204e-16

Ideas?

I'll submit the patches after I update my tree from Octave CVS repo.

Regards,
Mumit