beowulf performance with MPI

[Tony Skjellum]

Greg, you are obviously right about knowing what you meant.  I too know
what I meant.  So, I concede your point, so you can go on to other
postings.

_Tony

Anthony Skjellum, PhD, President (tony at mpi-softtech.com) 
MPI Software Technology, Inc., Ste. 33, 101 S. Lafayette, Starkville, MS 39759
+1-(662)320-4300 x15; FAX: +1-(662)320-4301; http://www.mpi-softtech.com
"Best-of-breed Software for Beowulf and Easy-to-Own Commercial Clusters."

On Sun, 25 Jun 2000, Greg Lindahl wrote:

> > Let me clarify even further before we waste more bandwidth on this,
> > and people get upset by Greg...
> >
> > 1) I was speaking about something that is totally free
> 
> Tony, if you knew this community better, you would be aware that "free beer"
> is different from "free software", and that calling any software "totally
> free" is a really bad idea.
> 
> If your product is now free in the liberty sense, that would be news, but I
> think you're probably just confusing the two.
> 
> > 2) My comments were directly aimed at the fact that free-to-free
> >    comparison suggests that people use the existing free software from us
> >    for TCP+Linux, rather than wait and hope that someone updates some
> >    other free software for them.
> 
> My comments were about modifying mpich, not about waiting or hoping. I said
> nothing about either. If you wish to reply to my comments in the future, I
> would suggest replying to what I said, not what you read into it.
> 
> -- g
>