From bug-octave-request at che dot utexas dot edu  Wed Oct 12 10:10:03 1994
From: "James B. Rawlings" <jbraw at che dot utexas dot edu>
To: bug-octave
CC: jcc at che dot utexas dot edu
Date: Wed, 12 Oct 1994 10:09:59 -0500


John,

Here is a serious bug, i.e. if we use Octave at some future MPC short
course, we are not going to be happy when it behaves in the following
way.

jcc tells me he already mentioned this one to you.  I am mentioning it
again in case it fell through the cracks.

By the way, as I read through your reporting bugs description, it
strikes me as somewhat excessive.  I guess I am to report the type of
machine I am using, OS and version number, the command line arguments
I gave to configure, etc.  Most people are either not going to know
all of these things or are not going to be willing to put forth the
effort to find them all out.

--STOP-- OK, now that I actually read the documentation of max, I see
we are not using it correctly.  Boy, this is turning into a poor bug
report, but I will send it anyway.  Max is not supposed to find the
max of a list of reals.  Jcc is trying to trick it into doing this.
He wants max to do comparison on two vectors, i.e.

v1=[a;b;c];
v2=[d;e;f];
max(v1,v2)

is supposed to return a vector

v3=[jmax(a,d);jmax(b,e);jmax(c,f)]

in which jmax is the max of a scalar list.  Maybe jwe should let jcc
know of a proper way to do this, rather than us keep picking on the
poor max function?

I would try

max([v1';v2'])'

to get what jcc is after.  Is there a better solution?  Do we need
more max function capabilities?

jbr


wally:3> octave
Octave, version 1.0.  Copyright (C) 1992, 1993, 1994 John W. Eaton.
This is free software with ABSOLUTELY NO WARRANTY.
For details, type `warranty'.

octave:1> max(1,2)
ans = 1
octave:2>