🚀 go-pugleaf

In article <727@mips.UUCP>, John Mashey (mash@mips.UUCP) writes:

> Comments? What sorts of metrics are important to the people who read
> this newsgroup? What kinds of constraints?  How do you buy machines?
> If you buy CPU chips, how do you decide what to pick?

1. Reputation of the manufacturer -- will they be there next year, and
   five years from now?  Will the family grow, or will it dead-end?
   Will existing software run on new members of the family with minimal
   changes and at the same time gain the performance advantages of the
   upgraded hardware?

2. Software development environment.  Can the desired end result be
   developed with available tools, or will we have to create our
   own?

3. Performance of the system that will result.

4. Cost of the CPU and FPU chips, and cost of the hardware required to
   make it all work.

The ranking is relative, but not rigid -- for example, failing #3 might
require changing #4 or accepting more pain in #2 in order to gain
acceptable performance.  Fudging on #1 could be hazardous to your mental
health, though!

					Steve Rice

----------------------------------------------------------------------------
{decvax | hplabs | ihnp4 | uw-beaver}!tektronix!videovax!stever