Support for Reti's argument.

[Peter Van Sickel <pvs@JAZZ.ALCOA.COM>]

I can support Kalman Reti's point regarding the openness of the
Symbolics system and how that allows its users to leverage their
time and effort on the huge body of existing code that comes with
Genera.

Recently, we needed to develop a server for each of a bunch of Mac's 
that are going to a Symbolics for Statice records.  Rather than think too 
hard about the details of what writing a server entails, we spent a day poking 
around in the guts of the Printer Manager.  It was our first time to spawn
stand-alone processes that need to do LMFS operations, etc. The code in
the various files associated with the Printer Manager, while not directly
usable, certainly made enlightening reading and allowed us to develop
the server we needed in a couple of days.  And we continued to go back to
the Printer Manager code as we encountered funny little things we hadn't 
anticipated initially, e.g., fs:with-automatic-login-to-sys-host is a
nifty form to wrap around code that uses LMFS if you want it to run 
without someone being logged in.

When I used to do C programming on UNIX or even now when I write C code for
the Mac II there was/is nothing like the availability of sources and 
on-line documentation that come with the Symbolics.  Programming on
the Symbolics is certainly a pleasant, productive experience for a wide 
variety of reasons and one of them is the openness of Genera.

Peter Van Sickel