I'm not convinced that it makes sense to keep the FreeVMS idea up. It just
does not make sense:
- OpenVMS is a mature environment. Maturity and obsolescense are tightly
coupled, but one thing is sure right now: VMS is 24 years old, many of the
ideas which influenced its design where intimately coupled to the computing
style of the late 1970ies (Minicomputers, CISC, large systems etc.)
- OpenVMS is complete. Reimplementing even only a limited subset of VMS would
be a real ambitious effort. Reimplementing it in a portable way would be even
harder, as the operating system is tightly coupled to the design of the DEC
CPUs.
- OpenVMS is completely different. It is not Unix with a funny command shell.
It has logical names, asynchronous I/O and generic traps available to user
programs, it has fat processes and a command interpreter in supervisor mode,
it has logical names which are visible process-, job-, group or system-wide,
it has fine-grained privileges, cluster-wide lock management, a file system
with structured files, it has a structured command line with unified command
and argument parsing, structured error messages etc etc. The whole power of
VMS becomes visible only if the complete system is at your disposal.
Porting Unix applications to VMS is usually a pain and I can see no sensible
reason why one would want to write applications for an operating system which
could, at best, reach the maturity of VMS 1.x in a few years from now.
I can understand the urge to write an own operating system. Re-writing a
proprietary or obsolete OS from the past would be the wrong way to enhance the
world. We already see the pain one has to experience when being faced with
Unix rewrites.
If what you want is use the coolest operating system of the world, get a VAX
or an Alpha and run OpenVMS. It's ready to use, it is complete and it is
pretty cool.
-Hans
--
finger hans_at_huebner.org for details
Received on Fri Jan 04 2002 - 13:14:25 GMT