rat 4.0.4

[Tom Coffin]

>From: Stuart Levy <slevy at ncsa.uiuc.edu>
>Date: Thu, 1 Jul 1999 12:03:53 -0500 (CDT)
>To: olson at mcs.anl.gov, tcoffin at ncsa.uiuc.edu
>Subject: rat 4.0.4
>X-md5sum: fd45e0b6af52af4341db1383cf2c627c
>X-md5sum-Origin: mx1.ncsa.uiuc.edu
>
>I was finding similar troubles with rat 4.0.3, *both* on the SGI
>(compiled from the UCL distribution after some tweaking) and on Linux
>(RH 6.0).  In both cases, it tended to print a message (from mix.c)
>about appending too many zeros, and then either seg-fault or encounter an
>assertion failure and abort.
>
>But since yesterday I've been playing with rat 4.0.4 -- compiled from
>the current anonymous-CVS copy -- and it seems to be working OK so far,
>both on linux and on my O2.    That is, I *think* it's OK.  If I transmit
>from one rat set at 16khz, and listen on another set to 8khz, I get
>no audible signal at all, even though the reception-quality matrix shows it's
>receiving something.  But that's better than dumping core.
>
>I'm leaving the two binaries up on the web at
>
>   http://niri.ncsa.uiuc.edu/slevy/rats/rat-sgi+linux.tar.gz
>
>The Linux one was compiled under Red Hat 5.2, so it seems to work on systems
>with either the older or newer glibc (i.e. both RH 5.2 and 6.0).
>
>