HP3000-L Archives

May 2000, Week 1

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
"Leonard S. Berkowitz" <[log in to unmask]>
Reply To:
Date:
Thu, 4 May 2000 11:58:51 -0400
Content-Type:
text/plain
Parts/Attachments:
text/plain (119 lines)
MPEX "compiles" command scripts. This might be a model to the approach being
suggested here.

===================
Leonard S. Berkowitz
Perot Health Care Systems
(Harvard Pilgrim Health Care account)
voice: 617-509-1212
fax:   617-509-3737







John Backus <[log in to unmask]> on 05/03/2000 03:29:33 PM

Please respond to John Backus <[log in to unmask]>








 To:      [log in to unmask]

 cc:      (bcc: Leonard Berkowitz/CORP/HPHC)



 Subject: [HP3000-L] A "CI" compiler?








In reply to John Krussel's post:

> I was thinking about this list last night and realized that several of
these have been done by users using CI programming (System-level CI
variables, and Newacct). And how it would be great if HP would just include
these with the next release so that everyone could use them and HP could
then support them.>

As in, if HP had HPCMD556, HPCMD557, HPCMD600, HPCMD601, HPCMD602, HPCMD650,
etc. groups in TELESUP, where scripts that pass reasonable testing could be
placed?  Perhaps though it is better that these be made available as
freeware via the internet where Jeff Vance and others at HP can quickly make
adjustments or corrections.

> The problem is that CI scripts are too easy to modify by the user and
there is no way HP could support them.>

Use at your own risk.

> But what if they made a compiler for these Scripts, and a facility to take
these compiled scripts, or any compiled program, and add it to the system
as a  regular command. This way if a user came up with a great script that
everyone wanted, HP could just take the script, compile it, and add it to
the OS.  It would keep the engineers from re-inventing (no pun intended on
HP's new logo) the wheel and give the users a new tool.>

The Denkor website http://www.denkor.com/hp3000/command_files/index.html
which Tim Ericson has maintained has been a useful resource for HPe3000
scripts!

Tim already has asked numerous times for such contributions.  A few things
might make finding and choosing scripts (programs, etc.) to utilize much
easier is...

Perhaps script segregation by OS level and 'assurance level' as in
directories like HPCMD557, HPCMD600, HPCMD601, etc. where 'proven' scripts
and programs are kept and then perhaps TCMD557, etc. directories where
scripts could be submitted for the HP3000 community to use.  Once scripts
have been tested or as a certain number of HP3000-L members give a "tests
out ok here vote", they could be moved to the 'HP' directories for 'use at
your own risk, others have ok but YMMV'.

> So I guess the bigger question is, has anyone thought of, or made an
attempt at creating a CI compiler?>

Since the "I" in "CI" is Interpreter, instead of a compiler, perhaps some
functionality that allows for script pre-execution syntax checking.  Perhaps
functionality that would step through a script and where any of certain ???
'destructive' commands are executed, the user is prompted for "Echo this" or
"Do this"?

How about a script building utility (compiler) which might be called
"CIUTIL"?  This might be making CI programming into something it was never
meant to be but what I have thought of is...(and this might be what you were
thinking of)...

  CIUTIL would understand a code that would consist of CI commands and
CIUTIL language ('Pseudo CI'?) that could be 'compiled' and which would
produce actual CI scripts. For example,  if we wrote a script in 'Pseudo
CI', a script had line numbers as in BASIC, then this pseudo language could
utilize PERFORM THROUGH's and GOTO like language which the 'CI compiler'
would 'explode' into a linear script with sections of lines repeated as
'directed'.  Some common scripting techniques such as dumping data "X" into
message file "Y" and then processing line by line could be denoted by some
verbiage that this CI 'compiler' would then turn into CI lines with the
correct syntax.  Other shortcuts and script building ideas could be
incorporated and this 'CI compiler' made as smart as desired.

With such wishful thinking having been said, what other useful tasks could
a utility like this do, has anyone else thought of something along these
lines?


John Backus
Beechglen Development, Inc.
beechglen.com

ATOM RSS1 RSS2