Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Following up on Rexx and Metal C
74 views
Skip to first unread message
Charles Mills
May 10, 2013, 2:41:55 PM5/10/13
to
I posted some questions here in late March about linking Rexx and Metal C. I
got some good input, and I promised to keep this list informed of what
worked out.
Well, the project is on indefinite hold, so I won't have any solid feedback
for the foreseeable future.
You may recall the problem I was trying to solve was "how can I write a
common function for both a C++ program and a compiled Rexx program? I want
to write it in a higher level language, and call it repeatedly from the Rexx
program without going through LE initialization every time."
FWIW, here is what I decided to do (before the project got quashed). I
decided to write the function in C conformant with the rules for Metal C. I
was then going to compile it in standard C and link it with the C++ code,
and compile it with Metal C to link with the Rexx code. (Yes, I am aware of
various linkage issues -- and their solutions.) If I wanted to for debug
purposes I could stick printf()'s and the like into the code under standard
C and either delete them or bracket them with #ifdef's before compiling
under Metal C.
(Anyone who wants to offer thoughts is welcome to do so but please note I am
not really looking for suggestions at this time as the project is on
indefinite hold.)
Charles
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to list...@listserv.ua.edu with the message: INFO IBM-MAIN
got some good input, and I promised to keep this list informed of what
worked out.
Well, the project is on indefinite hold, so I won't have any solid feedback
for the foreseeable future.
You may recall the problem I was trying to solve was "how can I write a
common function for both a C++ program and a compiled Rexx program? I want
to write it in a higher level language, and call it repeatedly from the Rexx
program without going through LE initialization every time."
FWIW, here is what I decided to do (before the project got quashed). I
decided to write the function in C conformant with the rules for Metal C. I
was then going to compile it in standard C and link it with the C++ code,
and compile it with Metal C to link with the Rexx code. (Yes, I am aware of
various linkage issues -- and their solutions.) If I wanted to for debug
purposes I could stick printf()'s and the like into the code under standard
C and either delete them or bracket them with #ifdef's before compiling
under Metal C.
(Anyone who wants to offer thoughts is welcome to do so but please note I am
not really looking for suggestions at this time as the project is on
indefinite hold.)
Charles
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to list...@listserv.ua.edu with the message: INFO IBM-MAIN
Anthony Rudd
May 11, 2013, 3:41:40 PM5/11/13
to
You mentioned in passing the problem with debugging Metal C by adding printf's. Unfortunately, Metal C does not offer I/O functions. I solved this problem by writing an extension library with basic I/O support.
Tony
Tony
Scott Ford
May 12, 2013, 10:14:04 AM5/12/13
to
Tony,
Read several of your C programs in Xephon, thanks for your contributions, it helped me a newbie in C
Scott ford
www.identityforge.com
from my IPAD
'Infinite wisdom through infinite means'
Read several of your C programs in Xephon, thanks for your contributions, it helped me a newbie in C
Scott ford
www.identityforge.com
from my IPAD
'Infinite wisdom through infinite means'
Charles Mills
May 12, 2013, 9:31:07 PM5/12/13
to
I thought it was an interesting work-around to have "dual-mode" code that ran either Metal or normal, possibly with #ifdef's around certain debug code such as printf()'s.
Charles
-----Original Message-----
From: IBM Mainframe Discussion List [mailto:IBM-...@LISTSERV.UA.EDU] On Behalf Of Anthony Rudd
Sent: Saturday, May 11, 2013 3:42 PM
To: IBM-...@LISTSERV.UA.EDU
Subject: Re: Following up on Rexx and Metal C
You mentioned in passing the problem with debugging Metal C by adding printf's. Unfortunately, Metal C does not offer I/O functions. I solved this problem by writing an extension library with basic I/O support.
Charles
-----Original Message-----
From: IBM Mainframe Discussion List [mailto:IBM-...@LISTSERV.UA.EDU] On Behalf Of Anthony Rudd
Sent: Saturday, May 11, 2013 3:42 PM
To: IBM-...@LISTSERV.UA.EDU
Subject: Re: Following up on Rexx and Metal C
You mentioned in passing the problem with debugging Metal C by adding printf's. Unfortunately, Metal C does not offer I/O functions. I solved this problem by writing an extension library with basic I/O support.
Anthony Rudd
May 14, 2013, 12:37:06 PM5/14/13
to
I am pleased to have helped in the past. I miss the Xephon publications, I also found them to be a very useful source of application code.
Scott Ford
May 14, 2013, 1:33:01 PM5/14/13
to
Tony,
I totally agree, good to see other techniques and new ideas ....
I totally agree, good to see other techniques and new ideas ....
Ed Finnell
May 14, 2013, 1:48:07 PM5/14/13
to
Sam has the XEPHON stuff aggregated into the _www.cbttape.org_
(http://www.cbttape.org) on the main page or /xephon/ directory /xephonm /$$index has
an index to pubs and topics in .pdf format. Velly nice.
In a message dated 5/14/2013 12:33:08 P.M. Central Daylight Time,
(http://www.cbttape.org) on the main page or /xephon/ directory /xephonm /$$index has
an index to pubs and topics in .pdf format. Velly nice.
In a message dated 5/14/2013 12:33:08 P.M. Central Daylight Time,
Scott Ford
May 14, 2013, 4:10:18 PM5/14/13
to
Ed,
Those pubs are a good resource . A lot of design knowledge over the years has been lost.
Those pubs are a good resource . A lot of design knowledge over the years has been lost.
0 new messages