This is G o o g l e's cache of http://www.lp.se/ftp/mailinglists/FREE-VMS.1998-09.
G o o g l e's cache is the snapshot that we took of the page as we crawled the web.
The page may have changed since that time. Click here for the current page without highlighting.
Google is not affiliated with the authors of this page nor responsible for its content. |
| These search terms have been highlighted: | levitte | programming |
|
|
Archive-Date: Mon, 28 Sep 1998 22:54:56 +0200
Date: Mon, 28 Sep 1998 16:54:45 -0400
From: Michael Goumans <mgoumans@uunet.ca>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
Subject: Re: Anything happening
In-Reply-To: <3.0.5.32.19980928140239.0090da10@ebsmmfs001.servicenet.net>
Message-ID: <Pine.GSO.3.96.980928164535.3928L-100000@troll.uunet.ca>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
> Anybody notice the following on INFO-VAX list. It seems he has a different
> name :)
Why would an employee of a company actually do this? I cant find
information regarding this on their web site, and I doubt that it will
ever end up there. Im pretty sure that nobody here will ever purchase any
product or service from sector 7 anymore, and it seems that basically
because his ideas were not accepted, he seems to want to try and do
whatever he can to destroy the entire project.
There arent too many Linux people that accept products without source code
though. It has happened, but isnt typical.
Perhaps he will actually have to follow through on it now. And even
better, those people who end up running it, will like the idea of the
FreeVMS project and join the team.
Hey Jon, if we come close to having a working system, will you anonymously
offer source on usenet?
Mike
Mike Goumans mgoumans@uunet.ca(work)
Unix Systems Manager mike@sage.irix.org(home)
UUnet WorldCom 416-216-5137(tel)
416-368-6701(fax) 416-719-1984(pgr)
>
> >From: Dave Bowman <reknaw@IBM.NET>
> >Description: REALLY FreeVMS? OpenVMS for Intel? Etc..
> >
> >X-Newsgroups: comp.os.vms
> >Lines: 11
> >X-Newsreader: Microsoft Outlook Express 4.72.3115.0
> >X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3
> >X-Trace: 25 Sep 1998 12:30:54 GMT, 129.37.52.229
> >X-Notice: Items posted that violate the IBM.NET Acceptable Use Policy
> >X-Notice: should be reported to postmaster@ibm.net
> >X-Complaints-To: postmaster@ibm.net
> >X-Gateway-Source-Info: USENET
> >Message-ID: <360b8cfe.0@news1.ibm.net>
> >Date: Fri, 25 Sep 1998 07:30:56 -0500
> >Reply-To: INFO-VAX@MVB.SAIC.COM
> >Sender: INFO-VAX Discussion <INFO-VAX@UGA.CC.UGA.EDU>
> >From: Dave Bowman <reknaw@IBM.NET>
> >Organization: IBM.NET
> >Subject: REALLY FreeVMS? OpenVMS for Intel? Etc..
> >Comments: To: Info-VAX@Mvb.Saic.Com
> >To: INFO-VAX@UGA.CC.UGA.EDU
> >
> >I have heard that Sector 7 are making their VMS environmental subsystem
> >available for linux. I was emailed to see how many of us would actualy use
> >it. Its going to be free for non commercial users. The downside, is no
> >source code access. I dont know the Sector 7 guys name to contact, but it
> >would be great to have VMS on my linux box !!!
> >
> >reknaw.
> >
>
>
================================================================================
Archive-Date: Mon, 28 Sep 1998 23:04:15 +0200
Date: Mon, 28 Sep 1998 17:04:05 -0400 (EDT)
Message-ID: <199809282104.RAA12310@antiochus-fe0.ultra.net>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: Free-VMS@lp.se
From: noahp@altavista.net (Noah Paul)
Reply-To: Free-VMS@lp.se
Subject: Re: Offer From Sector 7
What license is this under. And Jon, please use you REAL NAME!!!!
>Jon Power has made the following offer to the freeVMS group:
>
>
>I will provide a substantial subset of my VMS environmental subsystem
>(tools) (including DCL and a BATCH/PRINT queue manager) on linux free of
>charge for home (not commercial use). I just need to know how many of you
>out there are interested.
>
>My initial idea is:
>
> DCL
> BATCH & PRINT services
> SYS$, STR$, LIB$, MTH$, CONV$, FDL$, CLI$, OTS$
> RMS (keyed, sequential, relative [fixed and variable length])
> CLD compiler
> SMG services (not the client server GUI version)
> MESSAGE compiler
> VAX BASIC compiler (transpiler) - converts to C first and
> then compiler the C.
> beta's of our VTxxx emulator for NT
> beta's of our VTxxx DLL console app for NT (make a NT console look and
>behave like a VTxxx and interprets VT escape sequences without having to
>recode the programs)
>
>At this time I dont know if linux supports threads so I may NOT be able to
>offer the AST version. The I/O will be synchronous to Terminals and
>Mailboxes with a additional system service call SYS$SELECT_CHAN [ select() ]
>for I/O detection on multiple channels. if threads work the full AST driven
>version will be made available.
>
>The support will be on an ad hoc basis and only through email, and I will
>release BETA's from the web site early for you guys to hack with.
>
>At this time QIO to TCP/IP, DECNET, X.25 and X.29 will not be supported on
>linux.
>
>I will not be supplying:
>
> DEC C to ANSI C
> VMS BASIC to C++
> VMS PASCAL to C/C++
> VMS MACRO to C
> VMS COBOL to MF COBOL
> DEC FORMS
> TDMS
> FMS
> RTR
>
>Please find out how may of the FreeVMS group are interested, if its enough,
>Ill do the linux port.
>
>----------------------End of offer from Jon ---------------------
>
>People who are interested should let Jon know.
>His E-mail address is Jon_Power@sector7.com.
>
>Hopefully, having these tools available will increase participation in the
>project.
>
>Gary Burner
>g_burner@clark.net
>
>
>
---------
``There is no reason for any individual to have a
computer in their home''
-- Ken Olson, who lives within walking
distance of my house. Yeah, Kenny. How many do you
have in there?
Regards,
Noah Paul <noahp@altavista.net>
================================================================================
Archive-Date: Mon, 28 Sep 1998 23:22:33 +0200
Date: Mon, 28 Sep 1998 17:22:23 -0400 (EDT)
Message-ID: <199809282122.RAA15075@strato-fe0.ultra.net>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: Free-VMS@lp.se
From: noahp@altavista.net (Noah Paul)
Reply-To: Free-VMS@lp.se
Subject: Re: Anything happening
Probably would break the f---ing NDA he holds as his bible.
(not directed at mike)
JON YOU ARE WORTHLESS AS A MARKETOID AND PROGRAMMER. GO BACK UNDER YOUR
BRIDGE, TROLL!!!
(sorry, couldn't resist)
>> Anybody notice the following on INFO-VAX list. It seems he has a different
>> name :)
>
>Why would an employee of a company actually do this? I cant find
>information regarding this on their web site, and I doubt that it will
>ever end up there. Im pretty sure that nobody here will ever purchase any
>product or service from sector 7 anymore, and it seems that basically
>because his ideas were not accepted, he seems to want to try and do
>whatever he can to destroy the entire project.
>
>There arent too many Linux people that accept products without source code
>though. It has happened, but isnt typical.
>
>Perhaps he will actually have to follow through on it now. And even
>better, those people who end up running it, will like the idea of the
>FreeVMS project and join the team.
>
>
>Hey Jon, if we come close to having a working system, will you anonymously
>offer source on usenet?
>
>Mike
>
>Mike Goumans mgoumans@uunet.ca(work)
>Unix Systems Manager mike@sage.irix.org(home)
>UUnet WorldCom 416-216-5137(tel)
>416-368-6701(fax) 416-719-1984(pgr)
>
>
>>
>> >From: Dave Bowman <reknaw@IBM.NET>
>> >Description: REALLY FreeVMS? OpenVMS for Intel? Etc..
>> >
>> >X-Newsgroups: comp.os.vms
>> >Lines: 11
>> >X-Newsreader: Microsoft Outlook Express 4.72.3115.0
>> >X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3
>> >X-Trace: 25 Sep 1998 12:30:54 GMT, 129.37.52.229
>> >X-Notice: Items posted that violate the IBM.NET Acceptable Use Policy
>> >X-Notice: should be reported to postmaster@ibm.net
>> >X-Complaints-To: postmaster@ibm.net
>> >X-Gateway-Source-Info: USENET
>> >Message-ID: <360b8cfe.0@news1.ibm.net>
>> >Date: Fri, 25 Sep 1998 07:30:56 -0500
>> >Reply-To: INFO-VAX@MVB.SAIC.COM
>> >Sender: INFO-VAX Discussion <INFO-VAX@UGA.CC.UGA.EDU>
>> >From: Dave Bowman <reknaw@IBM.NET>
>> >Organization: IBM.NET
>> >Subject: REALLY FreeVMS? OpenVMS for Intel? Etc..
>> >Comments: To: Info-VAX@Mvb.Saic.Com
>> >To: INFO-VAX@UGA.CC.UGA.EDU
>> >
>> >I have heard that Sector 7 are making their VMS environmental subsystem
>> >available for linux. I was emailed to see how many of us would actualy use
>> >it. Its going to be free for non commercial users. The downside, is no
>> >source code access. I dont know the Sector 7 guys name to contact, but it
>> >would be great to have VMS on my linux box !!!
>> >
>> >reknaw.
>> >
>>
>>
>
>
---------
``There is no reason for any individual to have a
computer in their home''
-- Ken Olson, who lives within walking
distance of my house. Yeah, Kenny. How many do you
have in there?
Regards,
Noah Paul <noahp@altavista.net>
================================================================================
Archive-Date: Mon, 28 Sep 1998 23:23:10 +0200
MIME-Version: 1.0
Date: Mon, 28 Sep 1998 16:30:26 -0500
Message-ID: <00009782.eval@advocatehealth.com>
From: David.Dachtera@advocatehealth.com (David Dachtera)
Reply-To: Free-VMS@lp.se
Subject: Re[2]: writing DCL
To: "'Free-VMS@lp.se'" <Free-VMS@lp.se>
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Description: cc:Mail note part
Specifically, you're looking at the LIB$FIND_FILE
routine.
No, DCL itself does not process command-line parameters
much beyond some very basic verification. DCL does NOT
expand wildcarded filespec.'s as the UN*X shells do.
David J. Dachtera
dba DJE Systems
______________________________ Reply Separator _________________________________
Subject: RE: writing DCL
Author: Chris Kaler <ckaler@microsoft.com> at PO_EXTERNET
Date: 9/28/98 3:38 PM
VMS does not expand these symbols like UNIX. However, there
is a LIB$xxxx function that takes a pattern and optional /BEFORE
and /SINCE arguments and callers can walk the list
of matched files in an iterative fashion. As well, the $PARSE service will
walk the raw
list of file matches.
Chris
-----Original Message-----
From: Pradeep.Bashyal@ebs.ac.com [mailto:Pradeep.Bashyal@ebs.ac.com]
Sent: Monday, September 28, 1998 1:22 PM
To: Free-VMS@lp.se
Subject: writing DCL
>The solution to #1 is to go browse the DCL Dictionary (if memory
>serves, it is on the web somewhere) and pick a tiny subset to start
>with. One implementation strategy which might be interesting (not the
>only option, mind you, pick any one you like) is to start with a unix
>shell
As I understand the unix shells expand the shell metacharacters eg. *, ? etc
to filenames before passing it to the application so each utility does not
have
to. How does DCL handle this ? Should this functionality be in some library
that
each utility such as COPY, DELETE get from ? I went through the DCL
dictionary,
but it doesn't hint how DCL works. Excuse me for the ignorance, but can
anybody
provide pointers (hopefully non-null) or describe how this beast works ?
Anybody have Vax/Vms Internals gathering dust that they want to give a new
home to?
Thanks
Pradeep
================================================================================
Archive-Date: Tue, 29 Sep 1998 00:13:23 +0200
To: <free-vms@lp.se>
Subject: Re: Anything happening
Message-ID: <LEVITTE.98Sep29000315@nic.bofh.se>
From: levitte@lp.se (Richard Levitte - VMS Whacker)
Reply-To: Free-VMS@lp.se
Date: 28 Sep 1998 22:03:15 GMT
References: <199809282122.RAA15075@strato-fe0.ultra.net>
In-Reply-To: noahp@altavista.net's message of Mon, 28 Sep 1998 17:22:23 -0400 (EDT)
MIME-Version: 1.0
Content-Type: Text/Plain; Charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
In article <199809282122.RAA15075@strato-fe0.ultra.net> noahp@altavista.net (Noah Paul) writes:
Probably would break the f---ing NDA he holds as his bible.
I doubt it anyone holds an NDA as his or her bible. However, it's
usually a good idea to stick to those you have signed unless you enjoy
risking legal problems (pretty serious ones, most often).
OK, about Jon bashing...
(not directed at mike)
JON YOU ARE WORTHLESS AS A MARKETOID AND PROGRAMMER. GO BACK UNDER
YOUR BRIDGE, TROLL!!!
(sorry, couldn't resist)
I will simply say THAT IS FUCKING ENOUGH. There's a limit to this
kind of behavior, and I think it just got overstepped! Just leave
him alone, OK?
*sigh*
--
R Levitte, Levitte Programming; Spannv. 38, I; S-168 35 Bromma; SWEDEN
Tel: +46-8-26 52 47; Cell: +46-708-20 09 64; Fax: +46-708-20 06 13
PGP key fingerprint = 35 3E 6C 9E 8C 97 85 24 BD 9F D1 9E 8F 75 23 6B
http://richard.levitte.org/pubkey2.asc for my public key. levitte@lp.se
"price, performance, quality. Choose any two you like"
================================================================================
Archive-Date: Tue, 29 Sep 1998 00:33:58 +0200
Date: Tue, 29 Sep 1998 00:33:53 +0200
Message-ID: <6041-Tue29Sep1998003353+0200-levitte@lp.se>
From: Richard Levitte - VMS Whacker <levitte@lp.se>
Reply-To: Free-VMS@lp.se
To: free-vms@lp.se
Subject: Apologies...
MIME-Version: 1.0
Content-Type: Text/Plain; Charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
Sorry for the rude language a moment ago. Bad mood, I guess...
--
R Levitte, Levitte Programming; Spannv. 38, I; S-168 35 Bromma; SWEDEN
Tel: +46-8-26 52 47; Cell: +46-708-20 09 64; Fax: +46-708-20 06 13
PGP key fingerprint = 35 3E 6C 9E 8C 97 85 24 BD 9F D1 9E 8F 75 23 6B
http://richard.levitte.org/pubkey2.asc for my public key. levitte@lp.se
"price, performance, quality. Choose any two you like"
================================================================================
Archive-Date: Tue, 29 Sep 1998 00:36:53 +0200
Date: Mon, 28 Sep 1998 18:41:43 -0400
From: "Brian Schenkenberger, VAXman-" <system@TMESIS.COM>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
Message-ID: <009CCE98.19F6F799.5@TMESIS.COM>
Subject: Re: Anything happening
> Probably would break the f---ing NDA he holds as his bible.
>
>I doubt it anyone holds an NDA as his or her bible. However, it's
>usually a good idea to stick to those you have signed unless you enjoy
>risking legal problems (pretty serious ones, most often).
... and don't forget about the one's that you don't actually ink! There
are real sleaze-balls out there that will, without blinking an eye, show
up in a court of law with a forged document to bleed and extort money in
the guise of legally sanctioned extortion. ... speaking from the exper-
ience!
>OK, about Jon bashing...
>
> (not directed at mike)
> JON YOU ARE WORTHLESS AS A MARKETOID AND PROGRAMMER. GO BACK UNDER
> YOUR BRIDGE, TROLL!!!
> (sorry, couldn't resist)
>
>I will simply say THAT IS FUCKING ENOUGH. There's a limit to this
>kind of behavior, and I think it just got overstepped! Just leave
>him alone, OK?
To all reading this list...
Jon is obviously trying, in an immature and selfish manner, to promote his
warez and hinder the altruistic goals of this list. His "Gib Sinep" post
the other day was a perfect example! I'll offer the following suggestion:
IGNORE HIM! But be wary as he may come back to taunt this list using yet
another childish or feigned address.
Obviously, for Richard to respond with his uppercase castigation replete
the casual obscenity, Jon's flagrant marketeering and cajoling of Sector
7 is getting under everybody's skin. Chill!
--
VAXman- OpenVMS APE certification number: AAA-0001 VAXman@TMESIS.COM
================================================================================
Archive-Date: Tue, 29 Sep 1998 00:37:42 +0200
Date: Mon, 28 Sep 1998 18:38:09 -0400
Message-ID: <199809282238.SAA31772@harvey.cyclic.com>
From: Jim Kingdon <kingdonc@cyclic.com>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
In-Reply-To: <3.0.5.32.19980928152132.0090e100@ebsmmfs001.servicenet.net> (Pradeep.Bashyal@ebs.ac.com)
Subject: Re: writing DCL
References: <19980927014634.A15750@coyote.org> <001b01bde99d$45e7d340$ea782581@margarita.sector7.com> <19980927014634.A15750@coyote.org> <3.0.5.32.19980928152132.0090e100@ebsmmfs001.servicenet.net>
> As I understand the unix shells expand the shell metacharacters eg. *,
> ? etc to filenames before passing it to the application
Yes, although you can also turn this off by quoting them (for example:
ls a* # list files starting with a
ls 'a*' # list a file whose name is "a*", that is it contains "*"
).
As for VMS, I would suggest that applications expand wildcards
(presumably we'll be adding the relevant LIB$, RMS, &c calls to the
Kevin Handy library or similar ones, so they can do it the same way as
on VMS), and that our DCL-clone just not expand them (if having the
DCL-clone invoke unix programs seems important, could think about
various ways to make expansion optional, but this is a secondary goal.
If I were writing the code I'd probably play around with various ways
to do it and not break DCL compatibility too badly, but since I'm not
I should shut up now).
> Anybody have Vax/Vms Internals gathering dust that they want to give a
> new home to?
You don't need internals for this one. It may not be in the DCL
Dictionary, but the various interfaces we need to support should be in
the manuals for LIB$ and RMS and so on.
Reading the published information about VMS internals is fine (with
emphasis on the word published), but keep in mind that the internals
of FreeVMS can and should be different.
================================================================================
Archive-Date: Tue, 29 Sep 1998 01:04:22 +0200
Date: Mon, 28 Sep 1998 19:03:57 -0400 (EDT)
Message-ID: <199809282303.TAA09877@antiochus-fe0.ultra.net>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: Free-VMS@lp.se
From: noahp@altavista.net (Noah Paul)
Reply-To: Free-VMS@lp.se
Subject: Re: Apologies...
Apology accepted.
>Sorry for the rude language a moment ago. Bad mood, I guess...
>
>--
>R Levitte, Levitte Programming; Spannv. 38, I; S-168 35 Bromma; SWEDEN
> Tel: +46-8-26 52 47; Cell: +46-708-20 09 64; Fax: +46-708-20 06 13
> PGP key fingerprint = 35 3E 6C 9E 8C 97 85 24 BD 9F D1 9E 8F 75 23 6B
> http://richard.levitte.org/pubkey2.asc for my public key. levitte@lp.se
>
> "price, performance, quality. Choose any two you like"
>
>
---------
``There is no reason for any individual to have a
computer in their home''
-- Ken Olson, who lives within walking
distance of my house. Yeah, Kenny. How many do you
have in there?
Regards,
Noah Paul <noahp@altavista.net>
================================================================================
Archive-Date: Tue, 29 Sep 1998 01:09:44 +0200
Date: Mon, 28 Sep 1998 19:09:31 -0400 (EDT)
Message-ID: <199809282309.TAA22761@antiochus-fe0.ultra.net>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: Free-VMS@lp.se
From: noahp@altavista.net (Noah Paul)
Reply-To: Free-VMS@lp.se
Subject: VMS
I am very interested in OS programming and free software. However, I
don't know all that much about VMS. I tried to read 'Vax/VMS Internals and
Data Structures' but I couldn't understand it, because it assumed knowledge
of VMS. Is there any place that could help me learn these basics so I can be
useful to the team? I could probalby program VMS stuff if given a detailed
enough description of what it is, what syscalls to use, etc.
---------
``There is no reason for any individual to have a
computer in their home''
-- Ken Olson, who lives within walking
distance of my house. Yeah, Kenny. How many do you
have in there?
Regards,
Noah Paul <noahp@altavista.net>
================================================================================
Archive-Date: Tue, 29 Sep 1998 01:25:58 +0200
From: "Ashley Betts" <ashleyb@mincom.com>
Reply-To: Free-VMS@lp.se
To: <Free-VMS@lp.se>
Subject: RE: VMS
Date: Tue, 29 Sep 1998 09:20:05 +1000
Message-ID: <003d01bdeb36$8790ff70$405211ac@augusta.mincom.oz.au>
MIME-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
In-Reply-To: <199809282309.TAA22761@antiochus-fe0.ultra.net>
There is another book that is a little lighter. I think it is
"VAX/VMS Operating System Concepts". I don't have any other
details at the moment as the book is at home (with my IDS).
Regards,
AshleyB
-------
| | | | | | | Ashley Betts | Internet: ashleyb@mincom.com
| | | | | | | Technology Engineer | Voice : +61 7 3303 3120
| | | | | | | Mincom Pty Ltd | Fax : +61 7 3303 3232
| | | | | | | PO Box 72, Stones Corner |
|M|I|N|C|O|M| Brisbane, 4120, Q Australia |
> -----Original Message-----
> From: Noah Paul [mailto:noahp@altavista.net]
> Sent: Tuesday, September 29, 1998 9:10 AM
> To: Free-VMS@lp.se
> Subject: VMS
>
>
> I am very interested in OS programming and free
> software. However, I
> don't know all that much about VMS. I tried to read 'Vax/VMS
> Internals and
> Data Structures' but I couldn't understand it, because it
> assumed knowledge
> of VMS. Is there any place that could help me learn these
> basics so I can be
> useful to the team? I could probalby program VMS stuff if
> given a detailed
> enough description of what it is, what syscalls to use, etc.
>
> ---------
> ``There is no reason for any individual to have a
> computer in their home''
> -- Ken Olson, who lives within walking
> distance of my house. Yeah, Kenny. How many do you
> have in there?
>
> Regards,
>
> Noah Paul <noahp@altavista.net>
>
>
>
================================================================================
Archive-Date: Tue, 29 Sep 1998 01:49:10 +0200
Message-ID: <19980929004858.A18318@avalon.no46.wombat.ie>
Date: Tue, 29 Sep 1998 00:48:58 +0100
From: Kenn Humborg <kenn@iol.ie>
To: Free-VMS@lp.se
Subject: Re: VMS basics
Reply-To: Free-VMS@lp.se
References: <199809282309.TAA22761@antiochus-fe0.ultra.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
In-Reply-To: <199809282309.TAA22761@antiochus-fe0.ultra.net>; from Noah Paul on Mon, Sep 28, 1998 at 07:09:31PM -0400
On Mon, Sep 28, 1998 at 07:09:31PM -0400, Noah Paul wrote:
> I am very interested in OS programming and free software. However, I
> don't know all that much about VMS. I tried to read 'Vax/VMS Internals and
> Data Structures' but I couldn't understand it, because it assumed knowledge
> of VMS. Is there any place that could help me learn these basics so I can be
> useful to the team? I could probalby program VMS stuff if given a detailed
> enough description of what it is, what syscalls to use, etc.
Try and dig up a copy of Part I of the System Services manual. (I'm
working from memory here, so I can give volume numbers or order
codes.) This covered concepts such as ASTs, event flags, logical
names, processes, etc and the system services (aka syscalls) used
to manipulate them. Part II is the System Service Reference - not
quite as interesting to read (for most people, anyway... :-)
Also useful is the Run-Time Library (LIB$) manual. I'm pretty sure
there was a 'tutorial' section in this as well.
The I/O User's Manual gives all the info necessary to program
the main device types (aka ioctl()'s).
Older manuals are useful, too. While there have been some major
changes since V4.x, pretty much anything you read in a V4 manual
will be still usable today. Most of what I know came from V4
manuals while programming V5.4 & V5.5. (Then I left college
and had to program Windows :-(
It's actually not very difficult to learn VMS system programming.
The API is wonderfully consistent and you'll find the same
concepts appearing again and again. Try and find some user-level
freeware programs and take a look at them. Something like BOSS
(the session manager, not the DCL clone) which contains lots of
terminal I/O stuff might be a starting point.
Also, Wayne Sewell's 'Inside VMS: The System Manager's and System
Programmer's Guide to VMS Internals' is a sort of simplified
IDSM. It's out of print, but amazon.com have it listed at:
http://www.amazon.com/exec/obidos/ASIN/0442004745/002-9048972-7242043
Good luck,
Kenn
================================================================================
Archive-Date: Tue, 29 Sep 1998 07:14:34 +0200
Message-ID: <v04003a02b2361c1624f2@[140.186.88.14]>
In-Reply-To: <199809282309.TAA22761@antiochus-fe0.ultra.net>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Tue, 29 Sep 1998 01:13:34 -0400
To: Free-VMS@lp.se
From: Dick Munroe <munroe@acornsw.com>
Reply-To: Free-VMS@lp.se
Subject: Re: VMS
> ``There is no reason for any individual to have a
> computer in their home''
> -- Ken Olson, who lives within walking
> distance of my house. Yeah, Kenny. How many do you
> have in there?
Fewer than you might think. One of the reasons he came to that conclusion
was that he ordered a small PDP-11 for his home. It came on the same
pallet and in the same boxes that Decsystem 10s (at the time) were shipped
in. After hitting the roof about overpackaging and not being able to get
his computer in the door (like all his customers) he waited for Dec to do
something about the "problem".
Nothing ever happened.
So, if you can't get them through your door you can't use them in your
home, I guess.
Dick Munroe
(I have about 15 OpenVMS system hot all the time here, so I really AM on
the side of the angels but KO had his head wedged in a lot of ways and
stayed at Dec a MINIMUM of 10 years too long!)
--
Richard Munroe Internet: munroe@acornsw.com
Acorn Software, Inc.
267 Cox St. Office: (888) 226-7679
Hudson, Ma. 01749 USA FAX: (508) 562-1133
================================================================================
Archive-Date: Tue, 29 Sep 1998 10:36:55 +0200
Message-ID: <19980929083642.12540.qmail@hotmail.com>
From: "Steve Lindsey" <stevelindsey@hotmail.com>
Reply-To: Free-VMS@lp.se
To: free-vms@lp.se
Subject: Ramble on
Content-Type: text/plain
Date: Tue, 29 Sep 1998 01:36:40 PDT
I apologize in advance for this rather eclectic collection of points but
what does one do? Where is the faq?
I'm coming to this project late, about 8 years late according to my
machine which tells me its 29th feb 2004 at the moment. Please accept my
apologies if i've missed something or tread on somebody's toes.
I have read the mail archives (i'm a sad person), they contain alot of
useful information but its all over the place eg theres quite a bit
about wendin but people are still asking about it. Where is the faq?
When you've used DCL you don't want anything else, but you see the real
beauty of VMS when you do an internals course and/or read the big black
book. It is no coincidence that it's called internals and data
structures. VMS is a bunch of well thought out data structures with a
bunch of algorithms that use them. We (I said we not you. Oh shit) need
to nail these down, btw i haven't seen any mention of SDA which is now
on the freeware cd , we should be using this i guess.
I used to be a fortran bigot, then i discovered C and became a Pascal
bigot. Gentlemen what you do in the privacy of your own homes is up to
you but i just can not join in. I am prepared to do research, admin,
testing (yes you read that correctly) & docs. If we get pascal i'll
hack, i've used virtually every rtl routine there is. I'm sure this mail
list is not the place for all of this but what else is there, pray tell
me.
Looking ahead we have to have java, and when we do we'll get all the
apps we can handle.
Richard if you want some web pages put together let me know, i'm also
prepared to reread the mail archives and extract the useful bits, we
need pages on each executable,library,file formats, data structures etc
etc etc , if we do them well they can also become the foundation for the
documentation.
I have written a number of utilities that have their own cli, a la ncp,
authorize etc, it strikes me that dcl is just the daddy utility. it
would be a good idea to write a bare bones utility that supports
spawn
show version
help
define key
recall /all/load/save
exit
but does nothing real, this can be used by all of the following
utilities. I have such a beast but it's written in Pascal and it also
supports command files,comments and symbols
Would a web ring be a good idea?
Geocities give you 11mb free space, i took some and nothing strange has
happened to me.
How about freevms in white letters in reddy brown boxes as a logo.
Original.
Why doesn't everybody mail everybody they know about freevms
What is compaq's stance on this,where is the faq.
All that is required for evil to succeed is for good men to do nothing.
Somebody tell me who wrote that
I wrote that
Tot ziens
sys$steve
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
================================================================================
Archive-Date: Tue, 29 Sep 1998 13:39:20 +0200
From: Bill/Carolyn Pechter <pechter@shell.monmouth.com>
Reply-To: Free-VMS@lp.se
Message-ID: <199809291138.HAA01186@shell.monmouth.com>
Subject: Re: VMS
To: Free-VMS@lp.se
Date: Tue, 29 Sep 1998 07:38:53 -0400 (EDT)
In-Reply-To: <v04003a02b2361c1624f2@[140.186.88.14]> from "Dick Munroe" at Sep 29, 98 01:13:34 am
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
> > ``There is no reason for any individual to have a
> > computer in their home''
> > -- Ken Olson, who lives within walking
> > distance of my house. Yeah, Kenny. How many do you
> > have in there?
>
> Fewer than you might think. One of the reasons he came to that conclusion
> was that he ordered a small PDP-11 for his home. It came on the same
> pallet and in the same boxes that Decsystem 10s (at the time) were shipped
> in. After hitting the roof about overpackaging and not being able to get
> his computer in the door (like all his customers) he waited for Dec to do
> something about the "problem".
>
> Nothing ever happened.
Sure did. See the Micro PDP11 and MicroVax 8-)
Technology made system smaller.
>
> So, if you can't get them through your door you can't use them in your
> home, I guess.
>
> Dick Munroe
>
> (I have about 15 OpenVMS system hot all the time here, so I really AM on
> the side of the angels but KO had his head wedged in a lot of ways and
> stayed at Dec a MINIMUM of 10 years too long!)
Well, I actually installed an 11/750 in a carpeted living room in a
home in a development about a half hour away from here for a guy who
was a software developer. A little too much machine for the living room
if you ask me. But in 1983 that's what he had to choose from.
I disagree on the last point. I think KO just stayed about 5 years
too long and hired too many ex-IBM'ers in the sales force.
DEC should've avoided chasing IBM and kept itself in the midrange
and realtime market... when the PC showed up they could've kept the
move to Vaxes going by coming out with 32 bit Vax based PC's if they
didn't support the overblown bureaucracy that DEC built in the early
'80's.
One thing about KO and Jack Shields (who should've succeeded KO)
the services guys would've crawled uphill over broken glass nude
for them.
--Bill
Here's an exerpt from my motd. I wish it said FreeVMS though.
KO wasn't clairvoyant... but he was innovative. I wonder what
would've happened with the AT&T DEC merger of 1984.
FreeBSD i4got.pechter.org 2.2.7-STABLE - Wed Jul 22 23:50:44 EDT 1998
--------------------------------------------------------------------------------
| Quotations from chairman Ken |
--------------------------------------------------------------------------------
One of the questions that comes up all the time is: How
enthusiastic is our support for UNIX?
Unix was written on our machines and for our machines many
years ago. Today, much of UNIX being done is done on our machines.
Ten percent of our VAXs are going for UNIX use. UNIX is a simple
language, easy to understand, easy to get started with. It's great for
students, great for somewhat casual users, and it's great for
interchanging programs between different machines. And so, because of
its popularity in these markets, we support it. We have good UNIX on
VAX and good UNIX on PDP-11s.
It is our belief, however, that serious professional users will
run out of things they can do with UNIX. They'll want a real system and
will end up doing VMS when they get to be serious about programming.
With UNIX, if you're looking for something, you can easily and
quickly check that small manual and find out that it's not there. With
VMS, no matter what you look for -- it's literally a five-foot shelf of
documentation -- if you look long enough it's there. That's the
difference -- the beauty of UNIX is it's simple; and the beauty of VMS
is that it's all there.
-+-+=-+-
"TCP/IP is OK if you've got a little informal club, and it doesn't make
any difference if it takes a while to fix it."
-- Ken Olsen, in Digital News, 1988
+---------------------------------------------------------------------------+
| Bill and/or Carolyn Pechter | pechter@shell.monmouth.com |
| Bill Gates is a Persian cat and a monocle away from being a villain in |
| a James Bond movie -- Dennis Miller |
+---------------------------------------------------------------------------+
================================================================================
Archive-Date: Tue, 29 Sep 1998 14:23:07 +0200
Message-ID: <020201bdeba3$bd7993a0$083693d0@hmiller>
Reply-To: Free-VMS@lp.se
From: "Henry W. Miller" <henrym@sacto.mp.usbr.gov>
To: <Free-VMS@lp.se>
Subject: Re: VMS
Date: Tue, 29 Sep 1998 05:21:31 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Dick,
-----Original Message-----
From: Dick Munroe <munroe@acornsw.com>
To: Free-VMS@lp.se <Free-VMS@lp.se>
Date: Monday, September 28, 1998 22:20 PM
Subject: Re: VMS
>> ``There is no reason for any individual to have a
>> computer in their home''
>> -- Ken Olson, who lives within walking
>> distance of my house. Yeah, Kenny. How many do you
>> have in there?
>
>Fewer than you might think. One of the reasons he came to that
conclusion
>was that he ordered a small PDP-11 for his home. It came on the same
>pallet and in the same boxes that Decsystem 10s (at the time) were
shipped
>in. After hitting the roof about overpackaging and not being able to
get
>his computer in the door (like all his customers) he waited for Dec to
do
>something about the "problem".
>
>Nothing ever happened.
>
>So, if you can't get them through your door you can't use them in your
>home, I guess.
>
>Dick Munroe
>
Actually, I think that it was a PDP-8 based word processor, that was
supposed to be user-installable using merely a small screwdriver. KO
could not get it working. He was incensed and made a video tape that
was required viewing for all DEC employees about product responsibility.
>(I have about 15 OpenVMS system hot all the time here, so I really AM
on
>the side of the angels but KO had his head wedged in a lot of ways and
>stayed at Dec a MINIMUM of 10 years too long!)
>
>
>--
>Richard Munroe Internet: munroe@acornsw.com
>Acorn Software, Inc.
>267 Cox St. Office: (888) 226-7679
>Hudson, Ma. 01749 USA FAX: (508) 562-1133
>
>
>
-HWM
================================================================================
Archive-Date: Tue, 29 Sep 1998 14:53:18 +0200
From: Bill/Carolyn Pechter <pechter@shell.monmouth.com>
Reply-To: Free-VMS@lp.se
Message-ID: <199809291252.IAA20631@shell.monmouth.com>
Subject: DEC in the home
To: Free-VMS@lp.se
Date: Tue, 29 Sep 1998 08:52:29 -0400 (EDT)
In-Reply-To: <020201bdeba3$bd7993a0$083693d0@hmiller> from "Henry W. Miller" at Sep 29, 98 05:21:31 am
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
>
> Actually, I think that it was a PDP-8 based word processor, that was
> supposed to be user-installable using merely a small screwdriver. KO
> could not get it working. He was incensed and made a video tape that
> was required viewing for all DEC employees about product responsibility.
>
> >(I have about 15 OpenVMS system hot all the time here, so I really AM
> on
> >the side of the angels but KO had his head wedged in a lot of ways and
> >stayed at Dec a MINIMUM of 10 years too long!)
> >
> >
> >--
> >Richard Munroe Internet: munroe@acornsw.com
> >Acorn Software, Inc.
> >267 Cox St. Office: (888) 226-7679
> >Hudson, Ma. 01749 USA FAX: (508) 562-1133
> >
> >
> >
>
> -HWM
>
>
Hmm... is that where the "the DEC PC's must be installable with no tools
other than a standard screwdriver... and we'll include the screwdriver"
rule came from back in the Rainbow/Pro/DecmateII days...
I thought the test of having a DEC department manager install a machine
with no training was a good way of going through the idiot user
acceptance test.
Boy was KO ever right on this one.
Bill
+---------------------------------------------------------------------------+
| Bill and/or Carolyn Pechter | pechter@shell.monmouth.com |
| Bill Gates is a Persian cat and a monocle away from being a villain in |
| a James Bond movie -- Dennis Miller |
+---------------------------------------------------------------------------+
================================================================================
Archive-Date: Tue, 29 Sep 1998 15:40:30 +0200
Reply-To: Free-VMS@lp.se
From: "Christopher Chiesa" <cfchiesa@servtech.com>
To: <Free-VMS@lp.se>
Subject: RE: writing DCL
Date: Tue, 29 Sep 1998 09:34:19 -0400
Message-ID: <000201bdebad$dc610fa0$412b9780@franklin.lle.rochester.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
In-Reply-To: <199809282238.SAA31772@harvey.cyclic.com>
Jim Kingdon (kingdonc@cyclic.com) writes:
> As for VMS, I would suggest that applications expand wildcards
> (presumably we'll be adding the relevant LIB$, RMS, &c calls to the
> Kevin Handy library or similar ones, so they can do it the same way as
> on VMS), and that our DCL-clone just not expand them
I agree! Let's keep the fundamentals of FreeVMS programming as close
as possible to the fundamentals of VMS programming! If we don't, it
won't BE FreeVMS, it'll be FreeSomethingElse.
> (if having the
> DCL-clone invoke unix programs seems important, could think about
> various ways to make expansion optional, but this is a secondary goal.
> If I were writing the code I'd probably play around with various ways
> to do it and not break DCL compatibility too badly, but since I'm not
> I should shut up now).
IMHO this should be the responsibility of the C RTL, and should be an
OPTIONAL behavior selectable at compile or link time. Specifically,
the behind-the-scenes routine (is it VACX$MAIN? That doesn't look quite
right. Oh, woe is me; my VMS knowledge flees before the onslaught of
my now-daily use of <cringe> Unix... :-( ) that establishes the "stan-
dard Unix/C"-looking "argc/argv" environment and calls "main()". It
would be both EASY and APPROPRIATE to have that code expand wildcarded
filespecs and insert the resulting tokens in argv[]. By providing this
extension, we would greatly ease the porting to FreeVMS of C programs
which rely on the assumption that wildcards have ALREADY BEEN EXPANDED
upon entry to the program, and by making the extended behavior OPTIONAL
we would also preserve the traditional VMS behavior for the sake of
backward compatibility.
> > Anybody have Vax/Vms Internals gathering dust that they want to give a
> > new home to?
>
> You don't need internals for this one. It may not be in the DCL
> Dictionary, but the various interfaces we need to support should be in
> the manuals for LIB$ and RMS and so on.
The one that worries ME is SYS$CLI -- the underpinning of all the DCL-
callback services and RTL routines. More generally, the THINGS that
worry me are the "undocumented" parts of VMS on which much of the OS
relies but about which we know next to nothing in detail except by means
which would tend to "poison" the purity of a re-creation -- i.e., having
seen the source, or having disassembled/reverse-engineered real VMS code.
> Reading the published information about VMS internals is fine (with
> emphasis on the word published), but keep in mind that the internals
> of FreeVMS can and should be different.
Personally, I find this particular restriction infernally frustrating.
For me, the whole POINT of running something that claims to be "VMS," is
that I can program in Macro and use the VMS internals I know-and-love.
Without those two things, you don't have "FreeVMS;" you have "the VMS
APIs on [pick your platform]." It is VMS' internals that make it VMS;
the whole debate (e.g. with Jon Power of late) re whether-or-not QIOs
and ASTs can be properly implemented atop Linux or NT or Mach or what-
ever, becomes moot if ALL we're going to implement is the "look" (and,
if we're lucky, SOME of the "feel") of a particular API. Sure, it'll
then become convenient to port VMS-specific source code to FreeVMS (so-
called) -- but what's the POINT if the VMS-specific-*LOOKING* features
aren't really layered atop VMS-equivalent UNDERPINNINGS? The whole
thing just loses about 90% of its worth-doing-ness, in my eyes, if it's
only going to be allowed to go "so far, and no further."
My two cents...
Chris Chiesa
cfchiesa@servtech.com
================================================================================
Archive-Date: Tue, 29 Sep 1998 16:01:31 +0200
From: Bill/Carolyn Pechter <pechter@shell.monmouth.com>
Reply-To: Free-VMS@lp.se
Message-ID: <199809291401.KAA11766@shell.monmouth.com>
Subject: Re: writing DCL
To: Free-VMS@lp.se
Date: Tue, 29 Sep 1998 10:01:02 -0400 (EDT)
In-Reply-To: <000201bdebad$dc610fa0$412b9780@franklin.lle.rochester.edu> from "Christopher Chiesa" at Sep 29, 98 09:34:19 am
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
> Personally, I find this particular restriction infernally frustrating.
> For me, the whole POINT of running something that claims to be "VMS," is
> that I can program in Macro and use the VMS internals I know-and-love.
> Without those two things, you don't have "FreeVMS;" you have "the VMS
> APIs on [pick your platform]."
> My two cents...
>
> Chris Chiesa
> cfchiesa@servtech.com
>
The only thing that will run VAX Macro is a Vax... Are you proposing
writing a new macro assembler that looks like DEC's Macro32 for
other processors -- or do you want to run FreeVMS on a real Vax.
If the latter -- you'd be better off with the DEC Hobbiest use license.
There's no way an intel (or Sparc, or Mips etc) box can be a
completely Vax compatible box without writing a VAX simulator for
Intel and running the real VMS on it.
(Which will probably be slower than running a VMS-Like OS on another
processor.)
Bill
+---------------------------------------------------------------------------+
| Bill and/or Carolyn Pechter | pechter@shell.monmouth.com |
| Bill Gates is a Persian cat and a monocle away from being a villain in |
| a James Bond movie -- Dennis Miller |
+---------------------------------------------------------------------------+
================================================================================
Archive-Date: Tue, 29 Sep 1998 16:23:21 +0200
MIME-Version: 1.0
Date: Tue, 29 Sep 1998 09:20:42 -0500
Message-ID: <00009E03.eval@advocatehealth.com>
From: David.Dachtera@advocatehealth.com (David Dachtera)
Reply-To: Free-VMS@lp.se
Subject: Re: Ramble on
To: free-vms@lp.se, "Steve Lindsey" <stevelindsey@hotmail.com>
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Description: cc:Mail note part
Steve,
From where I sit, if you want to do something - JUST DO
IT! Richard (Levitte) has always expressed that attitude,
one way or another - except when he gets annoyed, then
he's as human as anyone.
I agree C is an attrocity - so is Neophyte Trash as is
most Micro$haft shoveware - but, it's what the "world"
has embraced to a large degree. Such is life, I guess.
Geocities tends to inflict a great deal of advertising
and other screen cluttering, cookie-depositing stuff on
folks who visit Geocities sites. That's too high a price
to pay in my book. I'd avoid it, personally.
IMHO...
David J. Dachtera
dba DJE Systems
================================================================================
Archive-Date: Tue, 29 Sep 1998 16:42:37 +0200
Date: Tue, 29 Sep 1998 09:41:31 -0600 (CST)
From: YEAR 2000 Means Writing Four Digits To Raise Consciousness of the MILLENIUM <DPENKALA@qgraph.com>
Reply-To: Free-VMS@lp.se
Subject: VAX/DEC BASIC Learning ?
To: Free-VMS@lp.se
Message-ID: <01J2DHF03W6G9I9TL0@hop.qgraph.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
Does anyone know where I could learn VAX or DEC BASIC from; meaning what book
is available as the closest means of a learning exchange or what inexpensive
classes would be preferrable to learning the language to the fullest.
I'm going to be working Y2K date changes and conversion programs for a while
and would like to become fluent in the language.
Any help ?
Dave Penkala
Quad Graphics Corp.
================================================================================
Archive-Date: Tue, 29 Sep 1998 16:46:21 +0200
Date: Tue, 29 Sep 1998 10:46:29 -0400
Message-ID: <199809291446.KAA27743@harvey.cyclic.com>
From: Jim Kingdon <kingdonc@cyclic.com>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
In-Reply-To: <000201bdebad$dc610fa0$412b9780@franklin.lle.rochester.edu> (cfchiesa@servtech.com)
Subject: Re: writing DCL
References: <000201bdebad$dc610fa0$412b9780@franklin.lle.rochester.edu>
> The one that worries ME is SYS$CLI -- the underpinning of all the DCL-
> callback services and RTL routines. More generally, the THINGS that
> worry me are the "undocumented" parts of VMS
Shrug. If you want to write something which attempts to support
undocumented interfaces, programs which SYS$CMKRNL and then poke
around in the OS's address space, &c, no one is stopping you. I mean,
that's what I've been saying about Bazaars and such - if you want it,
write it and the rest of us shouldn't stop you. Of course it would be
specific to the VAX (such programs, generally, don't even run on Alpha
VMS).
However, I second the suggestion that you look into the Hobbyist's
License. Trying to implement the documented interfaces for VMS is a
_huge_ task; trying to implement the undocumented ones is that much
harder. Running real VMS on a VAX (or VAX emulator) strikes me as
much more realistic (just my opinion, of course).
================================================================================
Archive-Date: Tue, 29 Sep 1998 16:59:41 +0200
Date: Tue, 29 Sep 1998 10:59:51 -0400
Message-ID: <199809291459.KAA27772@harvey.cyclic.com>
From: Jim Kingdon <kingdonc@cyclic.com>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
In-Reply-To: <19980929083642.12540.qmail@hotmail.com> (stevelindsey@hotmail.com)
Subject: Re: Ramble on
References: <19980929083642.12540.qmail@hotmail.com>
> I apologize in advance for this rather eclectic collection of points but
> what does one do? Where is the faq?
Do you want to write one? I assume you have already found
www.free-vms.org and some of the other web sites it links to....
> it would be a good idea to write a bare bones utility that supports [a
> few DCL commands] but does nothing real, this can be used by all of
> the following utilities. I have such a beast but it's written in
> Pascal and it also supports command files,comments and symbols
Where do I download it?
This is exactly the sort of thing I've been talking about on this list
lately. What Pascal compilers does it compile with?
> Would a web ring be a good idea?
We sort of have an informal one, in the sense that www.free-vms.org
links to my page and a few others, and I link to some of the others
too. If you want to join in, by all means write a web page and tell
the list about it.
================================================================================
Archive-Date: Tue, 29 Sep 1998 17:22:30 +0200
Message-ID: <FFACFADB9236D1118F2700805FFE65E50143F9B6@AMCLVX4>
From: "Canaday, Susan L (CTG)" <CanadaSL@BP.COM>
Reply-To: Free-VMS@lp.se
To: "'Free-VMS@lp.se'" <Free-VMS@lp.se>
Subject: RE: VAX/DEC BASIC Learning ?
Date: Tue, 29 Sep 1998 11:20:11 -0400
MIME-Version: 1.0
Content-Type: text/plain
i don't know of any formal classes that will teach you how to think your
way through every y2k issue in a variety of languages. here we have
to remediate a number of different languages; even the rather
experienced programmers did not anticipate the behavior that would
occur when certain pieces of code were operated with year 2000 dates
(01 jan 2000, etc). i think the best thing is to really be able to 'play
computer', to be able to visualize what will happen, and visualize
accurately what of the behavior will be important to that application,
and whether it should be fixed by interpreting the date and leaving
it as it is (y2k people call 'windowing', and at times there are quite
legitimate reasons for doing so), or changing the date to carry
century information. there is alot that has to be considered before
you decide how to handle many large, closely inter-related applications.
i can tell you alot more but the group as a whole might not really
think this is the right forum, so if you want to know more please
email me.
> ----------
> From: YEAR 2000 Means Writing Four Digits To Raise Consciousness
> of the MILLENIUM[SMTP:DPENKALA@qgraph.com]
> Reply To: Free-VMS@lp.se
> Sent: Tuesday, September 29, 1998 11:41 AM
> To: Free-VMS@lp.se
> Subject: VAX/DEC BASIC Learning ?
>
>
> Does anyone know where I could learn VAX or DEC BASIC from; meaning what
> book
> is available as the closest means of a learning exchange or what
> inexpensive
> classes would be preferrable to learning the language to the fullest.
> I'm going to be working Y2K date changes and conversion programs for a
> while
> and would like to become fluent in the language.
>
> Any help ?
>
> Dave Penkala
> Quad Graphics Corp.
>
================================================================================
Archive-Date: Tue, 29 Sep 1998 17:23:43 +0200
Message-ID: <199809291524.KAA31202@taarna.sector7.com>
From: jay@sector7.com
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
Subject: Re: what language?
Date: Tue, 29 Sep 1998 10:24:39 -0500
Sender: jay@sector7.com
[ Just so it is abundantly clear, I do not speak for my employer,
I am not my employer, and I most certainly am not Jon Power,
so don't EVEN start in on me. :) ]
I'm not going to argue about whether C is elegant. Such a discussion
seems irrelevant to me. I will, however, suggest that it is the
quickest path to a runnable system. I also submit that it is likely
that the first few "runnable" systems will include only code written
in one language, due to project bootstrapping concerns. This language
requires a compiler that cross-compiles and is available on as many
systems as possible to maximize use of available programming
resources.
I suspect this language will also be used to write the next few
compilers which will become available under the fledgeling FreeVMS
system. It is also likely that the only code which will run on the
first few iterations of the system will be written in this language.
It stands to reason, therefore, that if you want to increase the
chances of your code running on those systems, you will write it in
the chosen "first language."
It seems unlikely to me that anyone will expend the effort to build a
compiler which is as suited to this task as GNU C. The other GNU
compilers all require that you first port some portion (usually a
large portion) of GNU C or the GNU C libraries to the target system,
and so do not appear (to me) to be the best solution. However,
if someone wants to prove me wrong and build GNU BLISS from scratch,
please do. Many of us would like to see it! I'm just trying to be
realistic.
Jay
================================================================================
Archive-Date: Tue, 29 Sep 1998 17:39:53 +0200
Message-ID: <028001bdebbf$582a73e0$083693d0@hmiller>
Reply-To: Free-VMS@lp.se
From: "Henry W. Miller" <henrym@sacto.mp.usbr.gov>
To: <Free-VMS@lp.se>
Subject: Re: what language?
Date: Tue, 29 Sep 1998 08:39:25 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
-----Original Message-----
From: jay@sector7.com <jay@sector7.com>
To: Free-VMS@lp.se <Free-VMS@lp.se>
Date: Tuesday, September 29, 1998 8:27 AM
Subject: Re: what language?
>[ Just so it is abundantly clear, I do not speak for my employer,
> I am not my employer, and I most certainly am not Jon Power,
> so don't EVEN start in on me. :) ]
>
OK...
>I'm not going to argue about whether C is elegant. Such a discussion
>seems irrelevant to me. I will, however, suggest that it is the
>quickest path to a runnable system. I also submit that it is likely
>that the first few "runnable" systems will include only code written
>in one language, due to project bootstrapping concerns. This language
>requires a compiler that cross-compiles and is available on as many
>systems as possible to maximize use of available programming
>resources.
>
I would submit that you are probably correct, but I just have to
believe that there are, floating around out there, "public domain" Basic
and Pascal compilers. This may be a good thing or a bad thing.
>I suspect this language will also be used to write the next few
>compilers which will become available under the fledgeling FreeVMS
>system. It is also likely that the only code which will run on the
>first few iterations of the system will be written in this language.
>It stands to reason, therefore, that if you want to increase the
>chances of your code running on those systems, you will write it in
>the chosen "first language."
>
>It seems unlikely to me that anyone will expend the effort to build a
>compiler which is as suited to this task as GNU C. The other GNU
>compilers all require that you first port some portion (usually a
>large portion) of GNU C or the GNU C libraries to the target system,
>and so do not appear (to me) to be the best solution. However,
>if someone wants to prove me wrong and build GNU BLISS from scratch,
>please do. Many of us would like to see it! I'm just trying to be
>realistic.
>
Well, as was pointed out a number of months ago, the source code for
the PDP-11 BLISS compiler is available in source code. This is a good
thing.
However, the compiler is written in BLISS-10. This is not a good
thing, unless you want to spend a lot of time modifying the source code
to even get an initial bootstrap compiler, unless you happen to have
access to a running PDP-10. There are a few of them left, here and
there. Maybe we could get some support from XKL to help bootstrap the
compiler...
>Jay
>
>
>
-HWM
================================================================================
Archive-Date: Tue, 29 Sep 1998 18:05:17 +0200
Date: Tue, 29 Sep 1998 12:05:29 -0400
Message-ID: <199809291605.MAA27986@harvey.cyclic.com>
From: Jim Kingdon <kingdonc@cyclic.com>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
In-Reply-To: <199809291524.KAA31202@taarna.sector7.com> (jay@sector7.com)
Subject: Re: what language?
References: <199809291524.KAA31202@taarna.sector7.com>
> [ Just so it is abundantly clear, I do not speak for my employer,
> I am not my employer, and I most certainly am not Jon Power,
> so don't EVEN start in on me. :) ]
:-). Fair enough. Even if you were Jon Power, I'd advocate ignoring
you (or some such response) rather than starting in on you.
> I also submit that it is likely that the first few "runnable" systems
> will include only code written in one language, due to project
> bootstrapping concerns.
Maybe. If code can compile with (for example) p2c or f2c, the
bootstrapping concerns aren't too bad. I'll also notice that Red Hat
Linux (for example, the same might be true of *BSD, &c, as well) ships
with g77 and g++ and I suppose it probably will ship with GNU Pascal
if/when that is integrated into EGCS. Who knows, maybe even CHILL
(which is already part of EGCS) :-).
> This language requires a compiler that cross-compiles
If we were porting to brand-new hardware, cross-compiling would indeed
be the only approach. But if people are instead working on user-level
code which can run under your favorite existing free OS, then the
tools are more likely to "just be there and just work".
Someone else asked about PDP-10's. There is an emulator which runs on
the Alpha and is faster than the original hardware. I couldn't
immediately find a web page, but if people are having serious trouble
finding anything about it, I probably could dig something up.
================================================================================
Archive-Date: Tue, 29 Sep 1998 18:07:22 +0200
MIME-Version: 1.0
Date: Tue, 29 Sep 1998 11:09:53 -0500
Message-ID: <0000A023.eval@advocatehealth.com>
From: David.Dachtera@advocatehealth.com (David Dachtera)
Reply-To: Free-VMS@lp.se
Subject: Re: VAX/DEC BASIC Learning ?
To: Free-VMS@lp.se, YEAR 2000 Means Writing Four Digits To Raise Consciousness of the MILLENIUM <DPENKALA@qgraph.com>
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Description: cc:Mail note part
Dave,
I couldn't find (on short notice) anything on-line except
the following:
http://www.openvms.digital.com/commercial/basic/basic_tra
nslator_manual.pdf
This is actually documentation for a DEC BASIC to MS VB
translator. It's some 308 pages, as I recall (I printed it
out here at work). It doesn't explain the language elements,
but it does list the great majority of them. Otherwise,
there should be (barely) enough on-line help to carry you
through.
If you need more info., try writing to my personal address,
djesys@earthlink.net.
David J. Dachtera
dba DJE Systems
================================================================================
Archive-Date: Tue, 29 Sep 1998 18:22:17 +0200
Date: Tue, 29 Sep 1998 12:26:58 -0400
From: "Brian Schenkenberger, VAXman-" <system@TMESIS.COM>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
Message-ID: <009CCF2C.EAADC3F6.1@TMESIS.COM>
Subject: Re: Ramble on
David.Dachtera@advocatehealth.com (David Dachtera) sez:
> I agree C is an attrocity - so is Neophyte Trash as is
^^^^^^^^^^^^^^
It's great to see that the term I coined for the true meaning of NT
is being used by others.
> most Micro$haft shoveware - but, it's what the "world"
> has embraced to a large degree. Such is life, I guess.
I've long describe this phenomena with the following metaphorical:
50 billion flies eat shit -- It still doesn't make me want to rush
out for a plate full to satiate my palate.
This belief that all the world is Micro$loth is all the more reason
to create a FreeVMS. Linux is moving beyond the realm of a curio --
why not do the same for a VMS likeness.
At the same time, I would have to believe that if VMS were available
on a VAX or Alpha workstation, most of the subscribers to this list
would be satisfied. I will continue, as should others, to persuade
DEC/Compaq to make VMS available for the general masses at a reason-
able price. Others on this list will still want to effect some sort
of VMS clone for non-VAX and non-Alpha platforms. A wonderful goal,
although VMS to a FreeVMS is a far more complicated endeavour that a
unix to Linux. There's just much more to VMS and many things in VMS
(that most of us take for granted) would be sorely missed if not im-
plemented.
Whatever the final outcome -- affordable VMS from DEC/Compaq or the
realization of FreeVMS -- it's got to be a far better alternative to
schlockwarez from Emperor William III and the Evil Empire of Redmond.
--
VAXman- OpenVMS APE certification number: AAA-0001 VAXman@TMESIS.COM
================================================================================
Archive-Date: Tue, 29 Sep 1998 18:57:54 +0200
Date: Tue, 29 Sep 1998 12:57:33 -0400 (EDT)
From: Jerrold Leichter <leichter@smarts.com>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
Subject: RE: writing DCL
In-Reply-To: <000201bdebad$dc610fa0$412b9780@franklin.lle.rochester.edu>
Message-ID: <Pine.GSO.4.02A.9809291244020.15795-100000@just>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
| > Reading the published information about VMS internals is fine (with
| > emphasis on the word published), but keep in mind that the internals
| > of FreeVMS can and should be different.
|
| Personally, I find this particular restriction infernally frustrating.
| For me, the whole POINT of running something that claims to be "VMS," is
| that I can program in Macro and use the VMS internals I know-and-love.
| Without those two things, you don't have "FreeVMS;" you have "the VMS
| APIs on [pick your platform]." It is VMS' internals that make it VMS;
| the whole debate (e.g. with Jon Power of late) re whether-or-not QIOs
| and ASTs can be properly implemented atop Linux or NT or Mach or what-
| ever, becomes moot if ALL we're going to implement is the "look" (and,
| if we're lucky, SOME of the "feel") of a particular API. Sure, it'll
| then become convenient to port VMS-specific source code to FreeVMS (so-
| called) -- but what's the POINT if the VMS-specific-*LOOKING* features
| aren't really layered atop VMS-equivalent UNDERPINNINGS? The whole
| thing just loses about 90% of its worth-doing-ness, in my eyes, if it's
| only going to be allowed to go "so far, and no further."
Internal stuff has changed during the lifetime of VMS; some parts of it has
changed repeatedly.
Trying to maintain exactly the same internal structure, all the way down, is a
waste of time. For one thing, will you try to match the VAX or the Alpha
structure? Some of the differences are fundamental, and will only grow. The
*reason* they exist, and will grow, is that VAXes and Alphas are very
different machines, and what's appropriate on one may not be appropriate on
the other. Other architectures are neither VAXes nor Alphas, and *neither*
design may be appropriate. Further, in a portable implementation, *any*
design that's optimized for one particular architecture may be inappropriate.
I should think what we want from a FreeVMS, in order of decreasing importance,
is:
1. Compatibility with as many published/supported VMS API's as
possible;
2. Quality of implementation (reliability etc.) as good as possible;
3. Compatibility with as many widely used internal API's as possible.
Putting the last of these before the second says you want to be able to hack,
but you don't care as much about practical use. Is that really what you want?
It's my guess that *most* internals-based VMS hacks will not work in any
reasonable VMS. Then again, many of them have been painful to move from VAX
to Alpha VMS. I don't see it as particularly worthwhile to be able to move
such programs back and forth between FreeVMS and Alpha - or is it VAX? - VMS.
In fact, one worthwhile goal would be to make the most popular such hacks
unnecessary! For example, DCLCOMPLETE is very clever, but it's a hack because
the appropriate interfaces don't exist in VMS - and there are limits on what
it can do for the same reason. A FreeVMS DCL shouldn't need DCLCOMPLETE - it
should do completion itself. (For one thing, it can do so without some of
DCLCOMPLETE's oddball side-effects. For another, it can do a better job -
e.g., command name completion can be based on exactly the same algorithms that
will actually parse the command.)
Similarly, I'd expect FreeVMS to provide a standard, documented way to send
cross-process AST's. This is a fundamental VMS technique, used widely in many
"hacks", but for various reasons VMS has never supported the interface for any
use but for VMS itself.
-- Jerry
================================================================================
Archive-Date: Tue, 29 Sep 1998 19:12:15 +0200
Date: Tue, 29 Sep 1998 13:12:08 -0400 (EDT)
From: Jerrold Leichter <leichter@smarts.com>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
Subject: Re: what language?
In-Reply-To: <199809291524.KAA31202@taarna.sector7.com>
Message-ID: <Pine.GSO.4.02A.9809291259460.15795-100000@just>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
For better or worse, C is the language of choice for most OS development
today. If you don't use it, you'll still end up having to support its use so
that you can import all kinds of tools.
C++ is an alternative: It *can* be used much more safely than C, though the
sharp edges are all still there. The downside is that the damn language is so
large, everything is more complicated to get right, the compilers are much
slower ....
A free BLISS compiler, even if it existed, wouldn't be worth the effort.
BLISS is no safer a language than C; if anything, it's less safe. Its big
advantages were that it (a) was at an ever lower level than C, with ways to
specify things like exact register usage, essential when talking to MACRO
code; (b) had really aggressive optimizing compilers. These days, machines
are so fast that it's just not worth it to use a very low level, super-
optimizing compiler at the cost of safety, supportability, etc.
An interesting alternative is Modula-3. It's a safe language in the Pascal
tradition. A free compiler (which uses the Gnu back end) is available for it.
It's sufficiently powerful to build a real operating system - the SPIN OS is
written almost entirely in Modula-3. (The non-Modula-3 pieces aren't where
you might think: Everything actually written *for* SPIN is in M3. However,
to save some effort, I think SPIN uses some drivers borrowed from NetBSD.)
See http://www.research.digital.com/SRC/modula-3/html/home.html for general
information about the language. See
http://mason.gmu.edu/~sallain/html/m3projects.html for links to work being
done in Modula-3, including SPIN.
The biggest disadvantage of Modula-3 is that so few people know it. On the
other hand, it's a tiny language by contemporary standards. (Many of the good
things in Java are clearly inherited from Modula-3; those familiar with Java
will come up to speed pretty quickly. However, because it had different goals,
Java left out some of the things that make it practical to do very low-level
stuff in Modula-3. That's why Java is not a reasonable choice for an OS
implementation language, at least at present.)
-- Jerry
================================================================================
Archive-Date: Tue, 29 Sep 1998 20:17:21 +0200
Date: Tue, 29 Sep 1998 19:13:35 +0100 (BST)
From: Chris Hedley <cbh@teabag.demon.co.uk>
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
Subject: Re: Ramble on
In-Reply-To: <009CCF2C.EAADC3F6.1@TMESIS.COM>
Message-ID: <Pine.SGI.3.96.980929191050.1082C-100000@artichoke.teabag.demon.co.uk>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
On Tue, 29 Sep 1998, Brian Schenkenberger, VAXman- wrote:
> David.Dachtera@advocatehealth.com (David Dachtera) sez:
> > I agree C is an attrocity - so is Neophyte Trash as is
> ^^^^^^^^^^^^^^
> It's great to see that the term I coined for the true meaning of NT
> is being used by others.
I'm quite content with the term "No Technology," personally!
> At the same time, I would have to believe that if VMS were available
> on a VAX or Alpha workstation, most of the subscribers to this list
> would be satisfied. I will continue, as should others, to persuade
I'm not sure; often what people are looking for is a fully functional OS
complete with source so they can tinker/learn/just generally understand
how things are done. Well, I say "often" and "people," perhaps it's just
me! Anyway, that's the reason I prefer using Linux to, say, IRIX.
Chris.
================================================================================
Archive-Date: Tue, 29 Sep 1998 20:34:55 +0200
Message-ID: <030401bdebd7$bc8264c0$083693d0@hmiller>
Reply-To: Free-VMS@lp.se
From: "Henry W. Miller" <henrym@sacto.mp.usbr.gov>
To: <Free-VMS@lp.se>
Subject: Re: DEC in the home
Date: Tue, 29 Sep 1998 11:33:58 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Bill,
-----Original Message-----
From: Bill/Carolyn Pechter <pechter@shell.monmouth.com>
To: Free-VMS@lp.se <Free-VMS@lp.se>
Date: Tuesday, September 29, 1998 6:42 AM
Subject: DEC in the home
[SNIP]
>Hmm... is that where the "the DEC PC's must be installable with no
tools
>other than a standard screwdriver... and we'll include the screwdriver"
>rule came from back in the Rainbow/Pro/DecmateII days...
>
Actually, I think that it predates those systems by at least 3-4
years.
>I thought the test of having a DEC department manager install a machine
>with no training was a good way of going through the idiot user
>acceptance test.
>
At DEC, it was a widely held belief that a manager lost all of their
technical ability within three weeks of promotion. Entropy. At that
point, I'm not sure that some managers could even be trusted with a
screwdriver, or any sharp, pointed object for that matter.
>Boy was KO ever right on this one.
>
He was an engineer's engineer, without a doubt.
>Bill
>
>
>+----------------------------------------------------------------------
-----+
>| Bill and/or Carolyn Pechter | pechter@shell.monmouth.com
|
>| Bill Gates is a Persian cat and a monocle away from being a villain
in |
>| a James Bond movie -- Dennis Miller
|
>+----------------------------------------------------------------------
-----+
>
-HWM
================================================================================
Archive-Date: Tue, 29 Sep 1998 21:56:31 +0200
Message-ID: <36114970.88E95A01@comset.net>
Date: Tue, 29 Sep 1998 23:56:16 +0300
From: Anton Konashenok <kona@comset.net>
Reply-To: Free-VMS@lp.se
MIME-Version: 1.0
To: Free-VMS@lp.se
Subject: Re: what language?
References: <Pine.GSO.4.02A.9809291259460.15795-100000@just>
Content-Type: text/plain; charset=koi8-r
Content-Transfer-Encoding: 7bit
Jerry Leichter writes:
> These days, machines are so fast that it's just not worth it to use a
> very low level, super-optimizing compiler at the cost of safety,
> supportability, etc.
How about uVAX II ? I thought we would want to have the system run on
VAXen at least to validate API compatibility. Also, this "machines are
so fast" approach is very habit-forming. Look at Windoze's growth from
version to version. Personally, I'd be very glad to dust off that old
386 for Free-VMS.
Anton Konashenok
================================================================================
Archive-Date: Tue, 29 Sep 1998 22:30:59 +0200
Date: Tue, 29 Sep 1998 14:30:21 -0600 (MDT)
From: David Richardson <DRICHARDSON@EDMA.AVC.CALGARY.AB.CA>
Reply-To: Free-VMS@lp.se
Subject: Re: Writing DCL and Ramble
To: Free-VMS@lp.se, dave.richardson@avc.ab.ca, _drichardson@EDMA.AVC.CALGARY.AB.CA
Message-ID: <01J2DR82A1UA000NKA@AVC.CALGARY.AB.CA>
MIME-Version: 1.0
Content-Transfer-Encoding: 7BIT
Hi.
I'd like to get involved, perhaps with writing a DCL clone, but I don't own any
equipment to do the development on, and my disposable income has been disposed
of for the next little while.
Re Languages:
I'm of the opinion that we should use some of the Literate Programming tools,
like FWeb. (If you are not familiar with the term I'd recommend reading
"Literate Programming" by Donald E. Knuth. The ISBN is 0-937073-80-6. There
is FWeb documentation available at http://w3.pppl.gov/~krommes/fweb.html#SEC3).
Both referenced materials make much better (and substantially longer)
explanations than the 2 paragraphs that follow.
In a nutshell, you write your program in WEB: Web is a metalanguage that is a
combination of the language(s) you are using and TeX (Latex for FWEB). You
then run tangle to preprocess your code into something the language compilers
will accept, compile, and run. (Program.W and Program.CH --Tangle-> Program.c.
-CC->Program.obj --Link-> Program.exe.) You can alternately run Weave, which
produces a version of your code ready for the typesetter (TeX or LaTeX) to
produce a version which contains your documentation and explains the program in
the best order for your explanations, not the required order for the compiler.
The process allows change files as well: change files contain the OS/Hardware
specific code to allow local tailoring for your programming. While I may be a
bit naive about coding, I would expect that the calling standards and the
requirement for having the number of arguements available could initially be
implemented manually using change files. For Example a C function could be
implemented in your program, and C(x,y,z) could be replaced with C(3,x,y,z) in
a macro in the change file. Similarly C(a,b) could be replaced by C(2,a,b) or
C(2,A,B,<default 3rd param>) as necessary. This could be discontinued on
OpenVMS where the compilers implement the calling standards.
FWeb supports (if I recall correctly) Fortran 77, Fortran 90, Ratfor, C, and
C++, and TEX but uses LaTex as its typesetting program. It also has a language
independant mode for supporting any other language.
Re DCL Cloning:
I see the requirements for the DCL equivalent FCL (Free-VMS Command Language)
as (and this is still muddy first thoughts):
FCL should accept any DCL procedure and produce the same results, EXCEPT where a
feature of DCL has explicitly been removed or changed (Executable comments, for
example).
FCL should implement a command language that is a superset of the DCL SET
COMMAND. I see at least two areas where there should be some changes:
First it should be possible or required to use the set command command to
enter even builtin funtions to the FCL command tables for every command
(except perhaps for the set command command). I think (but could be
convinced otherwise) that having all the commands and lexical functions in
a single table, and mandating the lexical functions be described there as
well will make it easer (possible) to add user written lexical functions.
(I expect that means that someone will have to some up with a lexical
function API for free-vms? I'd expect that the lexical functions will need
access to symbols and the symbol data structures in FCL at a minimum.)
Doing this may allow other utilities to be able to use the bare FCL
executable with a different set of set commands, and not have to write
their own command interpreters. [See quote of Steve Lindsey below]
The second area that should be expanded on is having the ability to rewrite
the command line as it is seen by called programs. This feature of being
able to lie to the programs about what the user entered would allow FCL to
be used to add DCL like qualifiers to UNIX style programs on Free-VMS and
Open-VMS by rewriting the command line to be what the program expects. You
would define qualifiers in the DCL style, with rewrite rules that told FCL
what to replace the DCL qualifiers with on the command line to get the
equivalent results.
FCL should function on OpenVMS as an alternate command interpreter. This will
allow it to be fully tested before the kernel for Free-vms exists.
FCL should function on Linux, and other U*Xes as a (decidedly non-unix like)
shell. It should implement the CLI$ functions that are needed by *-VMS
originating programs to get their parameters to aid porting VMS software to the
other operating systems. With the requirement that it meet certain shell
requirments, like wildcard expansion and the additions to the Command Language
Syntax to allow re-writing, FCL could become quite a powerfull porting tool
between *-VMS and U*X worlds. Both ways, hopefully!
Whether or not FCL should function on Windows/dos/w95/nt systems allowing its
exploration there as well, possibly as an alternate dos command window for
95/nt, possibly as a replacement command.com on older systems or not is not
something I'm clear on. Perhaps it should be limited to the W95/98/NT group of
systems as a replacement for the COMMAND PROMPT option.
FCL should implement Pipes, either as a verb, or using unix style redirects or
both. This should be done in a general enough way that commands can have a
/output= qualifier grafted onto them using the piping facility, even if they
don't have it internally.
Steve Lindsey <stevelindsey@hotmail.com> wrote:
>I have written a number of utilities that have their own cli, a la ncp,
>authorize etc, it strikes me that dcl is just the daddy utility. it
>would be a good idea to write a bare bones utility that supports
>spawn
>show version
>help
>define key
>recall /all/load/save
>exit
>
What would be the minimum features or commands? Would SET COMMAND, HELP and
EXIT be enough, leaving the others to be added through the SET COMMAND by the
utility implementor?
Christopher Chiesa <cfchiesa@servtech.com> wrote:
>
>Jim Kingdon (kingdonc@cyclic.com) writes:
[snip]
>> (if having the
>> DCL-clone invoke unix programs seems important, could think about
>> various ways to make expansion optional, but this is a secondary goal.
>> If I were writing the code I'd probably play around with various ways
>> to do it and not break DCL compatibility too badly, but since I'm not
>> I should shut up now).
>
>IMHO this should be the responsibility of the C RTL, and should be an
>OPTIONAL behavior selectable at compile or link time. Specifically,
>the behind-the-scenes routine (is it VACX$MAIN? That doesn't look quite
The C library's access to the command line is by the parameters to the Main() ,
(IIRC, it's been years [like 10] since I did any C programming). Have the
parameters be the upper cased line ala VMS unless the command language
specified the rewrite rules, and if the line is re-written, pass the specified
re-write(s).
One DCL feature that I think should be removed is the executable comment. For
example, given the DCL script Demo_verify.com:
$ write sys$output f$verify()
$ overify = f$verify(1)
$ write sys$output f$verify()
$! 'F$Verify(0)
$ write sys$output f$verify()
$ set noverify
DCL produces the following output:
$ @demo_verify
0
$ write sys$output f$verify()
1
0
Where FCL should produce:
$ @demo_verify
0
$ write sys$output f$verify()
1
$! 'F$Verify(0)
$ write sys$output f$verify()
1
$ set noverify
David Richardson
OpenVMS Analyst
Alberta Vocational College - Edmonton
DRichardson@EDMA.AVC.Calgary.AB.CA (Open VMS, I'll get it)
Dave.Richardson@AVC.AB.CA (Exchange Server, Maybe ;-)
"OpenVMS is today what Microsoft wants Windows NT v8.0 to be!"
- 22-Sep-1998 at www.openvms.digital.com
================================================================================
Archive-Date: Tue, 29 Sep 1998 22:59:44 +0200
Reply-To: Free-VMS@lp.se
From: "Christopher Chiesa" <cfchiesa@servtech.com>
To: <Free-VMS@lp.se>
Subject: RE: writing DCL
Date: Tue, 29 Sep 1998 16:54:57 -0400
Message-ID: <000a01bdebeb$6a736030$412b9780@franklin.lle.rochester.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
In-Reply-To: <Pine.GSO.4.02A.9809291244020.15795-100000@just>
Jerrold Leichter (leichter@smarts.com) writes:
> Internal stuff has changed during the lifetime of VMS; some parts
> of it has changed repeatedly.
True. But (aside from the drastic changes necessitated under
Alpha) when was the last time it changed DRASTICALLY? I've prob-
ably been spoiled, spending most of my ten-plus-year VMS career
entirely between V4.4 and V5.4-3... there weren't TOO many chan-
ges, and those that there WERE, didn't affect me because most of
them occurred before I knew enough about VMS to NOTICE or UNDER-
STAND or CARE ABOUT them. :-)
> Trying to maintain exactly the same internal structure, all the
> way down, is a waste of time.
That's probably true if taken too far. However, I would still
like to see "as much as possible" of the internal structure pre-
served, as far down as is practical. I would also like to see
MORE structures preserved at LOWER and LOWER levels, as FreeVMS
evolves. Certainly I don't expect miracles out of V1.0.
(I may not even really be talking about internals in the same
sense as the rest of you. I'm talking about things like, "all
the complex and useful input-editing capabilities of the VMS
terminal driver, should also reside in the 'terminal driver'
under FreeVMS -- as opposed to their being moved to some OTHER
place in the architecture." Can THAT sort of thing, at least,
be preserved? I can't even imagine how one would IMPLEMENT
some parts of, say, unsolicited-input behavior, under any vari-
ety of *NIX... But that says more about my lack of detailed
*NIX knowledge, than about the impossibility of doing the deed...)
> For one thing, will you try to match the VAX or the Alpha
> structure? Some of the differences are fundamental, and will
> only grow. The
> *reason* they exist, and will grow, is that VAXes and Alphas are very
> different machines, and what's appropriate on one may not be
> appropriate on
> the other.
I would prefer to see the VAX version. VMS was designed hand-
in-glove with it, and is best suited to it. By comparison, Alpha
has to be put through contortions (see the Macro Porting Guide,
Alpha Architecture Handbook, and associated materials) to run
VMS; I'm STILL amazed it works as well as it does, though I ac-
cept the "fait accompli."
> Other architectures are neither VAXes nor Alphas, and
> *neither* design may be appropriate.
Could this explain why it has taken several YEARS to get to the
point of concensus on so much as the mere "underpinnings" of
FreeVMS? Can it be that the diehards who, at the beginning,
were adamant that FreeVMS could NOT be done on "just ANY" plat-
form, have simply taken these several years to become weary
of fighting, and to bow to the inevitable and accept the pro-
posals we now hear (and which horrify me no end) of implement-
ing FreeVMS "on top of Linux (which, itself, runs nine times
out of ten atop Intel hardware)?"
> Further, in a portable implementation, *any*
> design that's optimized for one particular architecture may be
> inappropriate.
So, what you're saying is that, no matter WHAT one does in creating
FreeVMS, the bottom line is that it won't be running on a VAX (most
likely), so that it will be CATEGORICALLY IMPOSSIBLE for FreeVMS to
ever truly "measure up" to "real" VMS. Could the gut-level realiza-
tion of this sad fact, be responsible for the all-too-obvious lack
of forward progress on the project? I've said it before, and I'll
say it again: if it doesn't share the essential PERFORMANCE and OPER-
ATIONAL properties of VMS -- i.e., guaranteed response-time to in-
terrupts, truly asynchronous I/O, truly efficient process-scheduling,
resource-distribution, etc. -- it almost ISN'T WORTH DOING. If I
wanted "NEVER QUITE as good," I would just give in altogether and
use NT and/or *NIX...
> I should think what we want from a FreeVMS, in order of
> decreasing importance,
> is:
>
> 1. Compatibility with as many published/supported VMS API's as
> possible;
> 2. Quality of implementation (reliability etc.) as good as
> possible;
> 3. Compatibility with as many widely used internal API's
> as possible.
And here I've been expecting "binary compatibility!" Load-and-go.
Damn.
> Putting the last of these before the second says you want to be
> able to hack,
> but you don't care as much about practical use. Is that really
> what you want?
Eh. That's hard to say. What constitutes "practical use?" Tell
me what uses the other subscribers of this list have in mind for
THEIR future FreeVMS systems, and I'll tell you whether I had a
similar use in mind. I *will* say, however, that in the ABSENCE
of that feedback, yes, your statement strikes me as a fair descrip-
tion of my position -- and that I literally CAN'T IMAGINE what
ELSE anyone might propose as an "end use" for a FreeVMS system!
Enlighten me! I can hardly wait to hear this!
> It's my guess that *most* internals-based VMS hacks will not work
> in any reasonable VMS.
Except of course under the "genuine article." Which is what *I*
had long been expecting/hoping to see reproduced as FreeVMS. It
only relatively recently became clear to me that it was "not to
be."
> Then again, many of them have been painful to move from VAX
> to Alpha VMS.
Agreed; see above. I consider this a shortcoming of ALPHA, and
will obviously consider it a shortcoming of any OTHER platform
which makes similar alterations necessary.
Still, you make a good point: I used OpenVMS Alpha and was actually
fairly happy with it once I got over being suspicious of it...
But ANYWAY...
> I don't see it as particularly worthwhile to be able to move
> such programs back and forth between FreeVMS and Alpha - or is it
> VAX? - VMS.
VAX. Definitely VAX. I don't particularly care about moving
programs back and forth, personally; 90% of what I use on VMS, is
stuff I wrote myself. If-and-when I go to port something, and a
particular RTL routine isn't available, I implement THAT ROUTINE
and move on...
> In fact, one worthwhile goal would be to make the most popular
> such hacks unnecessary!
Ooo... I wish you hadn't said "necessary!" My immediate reaction
to THAT word, is to ask, "since when does a 'hack' have to BE 'neces-
sary?'" The way I understand "hacking," a "hack" is "necessary"
only for the emotional satisfaction of the hacker. And if you
REALLY want to get down to brass tacks, in the last analysis NO
software, or hardware, or computing technology of ANY kind, is
really "necessary;" people got along just fine WITHOUT it for many
millennia. Even today, what you "hear on the street" about "com-
puters," seems sharply divided into two categories: those who use
computers primarily for entertainment, and LOVE them ("this new
game is SO K00L!"), and those who use them primarily for work, and
HATE them ("the damned thing is DOWN again!"). I don't know ANY-
BODY who WORKS with computers and doesn't have a serious "love/hate"
relationship with them.
> For example, DCLCOMPLETE is very clever, but it's a hack because
> the appropriate interfaces don't exist in VMS - and there are
> limits on what it can do for the same reason. A FreeVMS DCL
> shouldn't need DCLCOMPLETE - it should do completion itself.
> (For one thing, it can do so without some of DCLCOMPLETE's oddball
> side-effects. For another, it can do a better job - e.g., command
> name completion can be based on exactly the same algorithms that
> will actually parse the command.)
That's a matter of opinion, and one which, ultimately, I am not
qualified to debate: I have never seen nor used DCLCOMPLETE.
> Similarly, I'd expect FreeVMS to provide a standard, documented
> way to send cross-process AST's. This is a fundamental VMS tech-
> nique, used widely in many "hacks", but for various reasons VMS
> has never supported the interface for any use but for VMS itself.
I see nothing wrong with providing additional new "extension" APIs
to FreeVMS, above and beyond what "realVMS" ;-) offers. What I ob-
ject to is the all-too-apparent willingness to REPLACE chunks of
VMS with something that "merely" looks-and-feels "a lot like" VMS.
Another two cents from...
Chris Chiesa
cfchiesa@servtech.com
================================================================================
Archive-Date: Wed, 30 Sep 1998 00:20:41 +0200
Message-ID: <4FD6422BE942D111908D00805F3158DF0A75752E@RED-MSG-52>
From: Chris Kaler <ckaler@microsoft.com>
Reply-To: Free-VMS@lp.se
To: "'Free-VMS@lp.se'" <Free-VMS@lp.se>
Subject: RE: writing DCL
Date: Tue, 29 Sep 1998 15:20:13 -0700
> The only thing that will run VAX Macro is a Vax... Are you proposing
> writing a new macro assembler that looks like DEC's Macro32 for
> other processors -- or do you want to run FreeVMS on a real Vax.
So that statement isn't entirely true. You can compile VAX Macro
on an Alpha and generate Alpha instructions. Also, you'd be
amazed at how easily one can take VAX Macro and compile it on
some Unix systems...
================================================================================
Archive-Date: Wed, 30 Sep 1998 00:21:07 +0200
Sender: slover@nextwork.rose-hulman.edu
Message-ID: <36115D8E.E54D6FB4@Rose-Hulman.Edu>
Date: Tue, 29 Sep 1998 22:22:06 +0000
From: "Robert J. Slover" <robert.slover@Rose-Hulman.Edu>
Reply-To: Free-VMS@lp.se
MIME-Version: 1.0
To: Free-VMS@lp.se
CC: dave.richardson@avc.ab.ca
Subject: Re: Writing DCL and Ramble
References: <01J2DR82A1UA000NKA@AVC.CALGARY.AB.CA>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
David Richardson wrote:
> The second area that should be expanded on is having the ability to rewrite
> the command line as it is seen by called programs. This feature of being
> able to lie to the programs about what the user entered would allow FCL to
> be used to add DCL like qualifiers to UNIX style programs on Free-VMS and
> Open-VMS by rewriting the command line to be what the program expects. You
> would define qualifiers in the DCL style, with rewrite rules that told FCL
> what to replace the DCL qualifiers with on the command line to get the
> equivalent results.
Hmmm. Many (most?) of the GNU utilities use the 'getopts' (Get Options)
library to parse a command line and set flags and variables within the utility.
The method for defining the allowed set of options and what variables their
values get marshalled into is not exactly elegant, but it is straightforward.
I had at one time toyed with the idea of allowing VMS-style qualifiers by
giving getopts the ability to read a CLD-like file and accept a new set of
options with the / as a prefix character instead of - or --. I thought that
this might have the useful side effect of allowing a smart shell (tcsh) to
be modified to include tab-option completion, since it would have access
to the allowable set of options/arguments at any point in the command line.
This might also allow scripts to extract enough information to build simple
GUI wrappers on the fly for the hunt-and-point style of user. In other
words, Unix (Linux) could use this (IMHO).
All of these musings were going on when I was adding some command-line
options to a local GNU version of cpio that I had made customizations to. I
never got any farther than thinking about it.
A FreeVMS version of this library could be useful in a number of ways. It
could, by default, do things the VMS way and take care of the option
processing up-front in DCL and skip the parsing logic at the beginning of
ported utility 'main' functions. As an option, unix-like command syntax
could still be supported by handing a command line that didn't pass DCL
scrutiny directly to the utility. This would allow the same code to be
used in a FreeVMS 'POSIX' environment.
Just a thought. Maybe nuts, but thought I'd toss it out there.
--Robert
--
Robert J. Slover
Administrative Systems Manager
Rose-Hulman Institute of Technology
================================================================================
Archive-Date: Wed, 30 Sep 1998 00:23:17 +0200
Message-ID: <4FD6422BE942D111908D00805F3158DF0A75752F@RED-MSG-52>
From: Chris Kaler <ckaler@microsoft.com>
Reply-To: Free-VMS@lp.se
To: "'Free-VMS@lp.se'" <Free-VMS@lp.se>
Subject: RE: writing DCL
Date: Tue, 29 Sep 1998 15:22:37 -0700
SYS$CLI is not an issue. It is only a means by which DCL
makes interactions with the process space work. One example,
is to support LIB$GET_FOREIGN. Get some DCL-clone working
and then worry about how to tie into the application. More
than likely all you need to do is establish some shared
memory between the command processor and the application
space.
Chris
-----Original Message-----
From: Jim Kingdon [mailto:kingdonc@cyclic.com]
Sent: Tuesday, September 29, 1998 7:46 AM
To: Free-VMS@lp.se
Subject: Re: writing DCL
> The one that worries ME is SYS$CLI -- the underpinning of all the DCL-
> callback services and RTL routines. More generally, the THINGS that
> worry me are the "undocumented" parts of VMS
Shrug. If you want to write something which attempts to support
undocumented interfaces, programs which SYS$CMKRNL and then poke
around in the OS's address space, &c, no one is stopping you. I mean,
that's what I've been saying about Bazaars and such - if you want it,
write it and the rest of us shouldn't stop you. Of course it would be
specific to the VAX (such programs, generally, don't even run on Alpha
VMS).
However, I second the suggestion that you look into the Hobbyist's
License. Trying to implement the documented interfaces for VMS is a
_huge_ task; trying to implement the undocumented ones is that much
harder. Running real VMS on a VAX (or VAX emulator) strikes me as
much more realistic (just my opinion, of course).
================================================================================
Archive-Date: Wed, 30 Sep 1998 00:27:25 +0200
Message-ID: <4FD6422BE942D111908D00805F3158DF0A757530@RED-MSG-52>
From: Chris Kaler <ckaler@microsoft.com>
Reply-To: Free-VMS@lp.se
To: "'Free-VMS@lp.se'" <Free-VMS@lp.se>
Subject: RE: Writing DCL and Ramble
Date: Tue, 29 Sep 1998 15:26:53 -0700
> FCL should function on OpenVMS as an alternate command interpreter. This
will
> allow it to be fully tested before the kernel for Free-vms exists.
Don't go there - it serves no practical purpose. In order to be
an alternative command interpreter for VMS you need to do a bunch
of icky things and you must run in supervisor mode. That will limit
what you can call. Build something that works like csh and forks
applications. That will get the hard stuff done and you can use
it on any platform. Then figure out how to pass command line info,
synbols, etc. (which is all pretty easy and can be done in a platform
independent way). Stick to user mode.
Chris
================================================================================
Archive-Date: Wed, 30 Sep 1998 07:52:50 +0200
Date: Wed, 30 Sep 98 01:27:52 EDT
Message-ID: <009CCF9A0191ADE0.00000B29@ais.com>
From: bruce@ais.com
Reply-To: Free-VMS@lp.se
Subject: RE: writing DCL
To: Free-VMS@lp.se
Content-Type: text
Chris Chiesa (cfchiesa@servtech.com) writes:
>Jerrold Leichter (leichter@smarts.com) writes:
>
>> Other architectures are neither VAXes nor Alphas, and
>> *neither* design may be appropriate.
>
>Could this explain why it has taken several YEARS to get to the
>point of concensus on so much as the mere "underpinnings" of
>FreeVMS? Can it be that the diehards who, at the beginning,
>were adamant that FreeVMS could NOT be done on "just ANY" plat-
>form, have simply taken these several years to become weary
>of fighting, and to bow to the inevitable and accept the pro-
>posals we now hear (and which horrify me no end) of implement-
>ing FreeVMS "on top of Linux (which, itself, runs nine times
>out of ten atop Intel hardware)?"
>
>> Further, in a portable implementation, *any*
>> design that's optimized for one particular architecture may be
>> inappropriate.
>
>So, what you're saying is that, no matter WHAT one does in creating
>FreeVMS, the bottom line is that it won't be running on a VAX (most
>likely), so that it will be CATEGORICALLY IMPOSSIBLE for FreeVMS to
>ever truly "measure up" to "real" VMS. Could the gut-level realiza-
>tion of this sad fact, be responsible for the all-too-obvious lack
>of forward progress on the project? I've said it before, and I'll
>say it again: if it doesn't share the essential PERFORMANCE and OPER-
>ATIONAL properties of VMS -- i.e., guaranteed response-time to in-
>terrupts, truly asynchronous I/O, truly efficient process-scheduling,
>resource-distribution, etc. -- it almost ISN'T WORTH DOING. If I
>wanted "NEVER QUITE as good," I would just give in altogether and
>use NT and/or *NIX...
>
>> I should think what we want from a FreeVMS, in order of
>> decreasing importance,
>> is:
>>
>> 1. Compatibility with as many published/supported VMS API's as
>> possible;
>> 2. Quality of implementation (reliability etc.) as good as
>> possible;
>> 3. Compatibility with as many widely used internal API's
>> as possible.
>
>And here I've been expecting "binary compatibility!" Load-and-go.
>Damn.
I really DON'T understand this point of view. If all you want is VMS on
a VAX, and all you want to do is hack on it for fun, what's the point of
Free-VMS at all? You could just get a REAL VMS hobby license, buy a used
VAX (there are quite a lot of VAXstations available cheap), and hack away
to your heart's content. Often such used VAXen are available complete with
a VMS license that can be transferred cheaply, and then you don't even have
to bother with the hobby license.
Just a few reasons why people might want a Free-VMS to be available on
non-VAX platforms:
1) VAXen are getting rather long in the tooth. Although there are
still quite a few of them out there, they will have a finite
lifespan, especially since there aren't new models coming out
like there are for the Alpha, for example. It's entirely
reasonable to want to get something a bit more modern that has
more MIPs etc.
2) New VAXen are rather expensive compared to other machines such
as Alpha and 80x86 for the number of MIPs they have. This relates
back to #1.
3) Ability to move VMS programs to other platforms, possibly even
as a VMS-like environment running on another operating system.
This should not be dismissed lightly; many businesses and even
individuals have a lot of software that would be quite expensive
to port. (Many of my customers have been in this situation ...).
4) Desire for a more user-friendly command environment than, for
example, that available in many Unix shells or <shudder> the
MS-Windows COMMAND.COM process. This desire makes sense regardless
of the underlying OS. You really need more underpinnings than
just a Unix shell with DCL syntax, but I don't see