On Fri, 16 Mar 2001 17:41:35 -0500, "Manny Bhuta"
<[log in to unmask]> wrote:

>It has been over 11 years since I last did anything with HP3000/MPE.  I am
>now considering buying either HP3000 or HP9000 computer to do some software
>development.  I know next to nothing about HPUX but I am assuming that the
>future is with HPUX rather than MPE.

A reasonable assumption, but there are too many things going on to
know how things will pan out. One thing is for sure ... no single O/S
will be "the future".

MPE's future is as dependent on HP's practices as our commitment to
it. By choosing to not develop on MPE, then you are becoming part of
the self-fulfilling prophecy that MPE has no future.

Talk to some of the porting experts on 3000-L and see if you could
develop for both HP-UX and MPE. You may need to write an abstraction
layer for file handling (so you can let MPE do the things it's good
at, while holding HP-UX's hands for the same functions). Or you could
perhaps write the whole project in Perl, which runs on HP-UX and MPE
and just about anything short of a TRS-80 Model 1.

>Reading several Unix books, I am totally confused
>about its file system.

That's because it doesn't have one. Or rather, it has three or four
filesystems, none of which we MPE folks would consider real
filesystems. (UNIX programming challenge 1: place a non-advisory read
lock on file "a".)

>I would most appreciate some help with the following
>two questions.
>
>1.  Is C++ language available for MPE?  Will it work with KSAM, Message
>files and record oriented flat files?
>
>2.  Is it possible to get a very efficient file system for HPUX that
>supports Index sequential files and record oriented (rather than byte
>addressed) files.  Though not ideal, I can get away with using pipes in
>place of the message files.

As someone else earlier mentioned, the UNIX architects in the 1690s
(or was that 1960s?) that it was a Good Thing to consider all files as
a collection of single bytes; if the programmer wanted anything more
sophisticated, they could implement it themselves. Now, on most UNIXes
there are a couple of "standard" file tools, one of which is "dbm"
which is kinda like KSAM. Do a "man dbm" on your HP-UX system and see
what you think.
>
>My emphasis on record oriented rather than byte oriented file access is
>because, I was told that the Unix file system with its byte address, is not
>very efficient for randomly accessing records from very large files.

There are ways around that. As you learn more about what's available
on HP-UX, you might change this design criteria.

And all inefficiencies can always be sidestepped by faster hardware.
:)