Ginni Rometty is a former long-time CEO, president, and chairman of IBM.
An interesting interview with Podcaster Lex Fridman. I have just been
introduced to his podcasts,
and they are always good.
I learned why some interesting decisions were made at IBM which I really
couldn't understand.
I, OTOH, was appalled by C ever since it came out, much preferring PL/I.
As for assembler, there are a lot of things that I can do in a single statement
that are awkward and verbose in C. Further, 95% of my experience has been with
macro assemblers, and C's macro language looks like a joke.
--
On 3/4/23 18:11, Seymour J Metz wrote:
I, OTOH, was appalled by C ever since it came out, much preferring PL/I.
Certainly, the reasons for PL/I's failure are subject to debate, but
there are several factors that may have contributed:
1.
IBM was the primary stakeholder and developer of PL
Why switch to C++ rather than to a more modern language?
--
Shmuel (Seymour J.) Metz
http://mason.gmu.edu/~smetz3
From: IBM Mainframe Discussion List [IBM-MAIN@LISTSERV.UA.EDU] on behalf of
David Crayford [dcrayf...@gmail.com]
Sent: Monday, April 3, 2023
On 3/4/23 19:48, Seymour J Metz wrote:
Why switch to C++ rather than to a more modern language?
It's widely acknowledged that C++20 is one of the most contemporary
programming languages being used today [1]. With its compiled approach
and low-level memory access, it's difficult to find a more
If by "calling SVC 51" you mean "issuing SNAP(X)", we can continue. SVC 51 is
used for multiple purposes.
Use the DCB attributes that are documented or don't bother playing.
I am trying to dump a dataspace (that is owned by an other address space)
SNAPX doc says "The system dumps storage from
I know
I wrote a generic dump routine
If first flag byte +1 of the parm list bit0 is 0 its snap(x) NOT sdumpx
I am trying to dump a dataspace created in a SRB in another address space
the doc says I should be able to do it with snapx
Here is the doc
Use the DSPSTOR parameter on the SNAPX m
FWIW,
PL/X has a simple clear default: everything is passed by reference (and is
treated as input/output - readable and writeable). That is the historical
linkage standard. PLX will only do what is asked for.
Thus if you want something passed by-value, you "ask" for it by your definition
of th
Contemporary and modern mean very different things, and [[C++]] mentions
neither. Stroustrup admitted that the design of C++was compromised by
compatibility concerns.
From: IBM Mainframe Discussion List on behalf of
David Crayford
Sent: Monday, April 3
OK, I’ll rephrase. Modern C++ is one of the most modern programming languages
in use today. C compatibility is in the past. C++20 introduced modules and
reflection is already available in some compilers.
> On 3 Apr 2023, at 9:46 pm, Seymour J Metz wrote:
>
> Contemporary and modern mean very
"You keep using that word. I do not think it means what you think it means."
Modernity is determined by the original design, not by the latest tweak C++
still has defects inherited from C; there are later languages that were
designed ab initio, with no inherited defects.
_
> On 3 Apr 2023, at 10:03 pm, Seymour J Metz wrote:
>
> "You keep using that word. I do not think it means what you think it means."
>
> Modernity is determined by the original design, not by the latest tweak C++
> still has defects inherited from C; there are later languages that were
> des
ObDieJungfrauvonOrleans Believe that if you want' it wont change the reality.
From: IBM Mainframe Discussion List on behalf of
David Crayford
Sent: Monday, April 3, 2023 10:38 AM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: Fascinating Interview with Steve
On Mon, 3 Apr 2023 13:21:41 +, Peter Relson wrote:
>...
>Thus if you want something passed by-value, you "ask" for it by your
>definition of the parameters in an "entry declare" that can be used both by
>the caller and callee.
>
That feels like a function prototype.
>If you want somethi
On Mon, 3 Apr 2023 04:57:02 +, Frank Swarbrick wrote:
>
>(reply) Call by content is enforced by the caller. Call by value is enforced
>by the callee.
>
Ah. In my jargon I'd use "declared". So the callee is unaware of the
distinction between reference and content; it just sees an address i
I got the snap(x) to work dumping the data space owned by another address space
For my snap(x) dataset I was able to write messages
In my recovery where the program abended
And registers
Just wondering if I can do this with a sdump(x)
I know the dcb is RECFM=FB and lrecl=4160
> On Apr 3, 20
Yes, the callee is unaware of if the caller used pass by reference or pass by
content. That's a call site feature only. The callee uses (the default) call
by reference regardless.
A "by reference" parameter can be specified for any call, and need not be "the
same" for additional calls to the s
It depends on the language. In general, the caller must provide by the calling
conventions that the callee expects and the callee must abide by the
restriction in the parameter declarations.
From: IBM Mainframe Discussion List on behalf of
Frank Swarbri
Have you considered writing a message in a transaction dump?
From: IBM Mainframe Discussion List on behalf of
Joseph Reichman
Sent: Monday, April 3, 2023 11:37 AM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: Calling SVC 51
I got the snap(x) to work dumping
Transaction dump requires IPCS as well it’s unformatted seems restricted to one
address space
> On Apr 3, 2023, at 1:28 PM, Seymour J Metz wrote:
>
> Have you considered writing a message in a transaction dump?
>
>
> From: IBM Mainframe Discussion Lis
IPCS is much more convenient than scanning a printed dump.
From: IBM Mainframe Discussion List on behalf of
Joseph Reichman
Sent: Monday, April 3, 2023 1:41 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: Calling SVC 51
Transaction dump requires IPCS as w
I have used a few times but mainly for simple things like storage browse
If I want to dump data from another address space that’s the only way to go
Thanks
On Mon, Apr 3, 2023 at 1:50 PM Seymour J Metz wrote:
> IPCS is much more convenient than scanning a printed dump.
>
> ___
On Apr 3, 2023, at 12:03 AM, Tom Brennan
mailto:t...@tombrennansoftware.com>> wrote:
I actually always liked C, maybe because its original simplicity reminded me of
Assembler. I mean, what other language can you goof up a length value or
pointer and overwrite a bunch of other data areas by mis
ITYM PDP-7.
--
Shmuel (Seymour J.) Metz
http://mason.gmu.edu/~smetz3
From: IBM Mainframe Discussion List [IBM-MAIN@LISTSERV.UA.EDU] on behalf of
Pew, Curtis G [curtis@austin.utexas.edu]
Sent: Monday, April 3, 2023 2:37 PM
To: IBM-MAIN@LISTSERV.UA.EDU
On Mon, 3 Apr 2023 at 13:41, Joseph Reichman wrote:
>
> Transaction dump requires IPCS as well it’s unformatted seems restricted to
> one address space
I'm confused. The result of SDUMP is also unformatted, hence that 4160
FBS (don't forget the "S").
(I shouldn't call it "unformatted"; it's not
"C is a language that combines all the elegance and power of assembly language
with all the readability and maintainability of assembly language."
-Original Message-
From: IBM Mainframe Discussion List On Behalf Of
Pew, Curtis G
Sent: Monday, April 3, 2023 1:37 PM
To: IBM-MAIN@LISTSERV.
PROBDESCAD=probdescad
--
Shmuel (Seymour J.) Metz
http://mason.gmu.edu/~smetz3
From: IBM Mainframe Discussion List [IBM-MAIN@LISTSERV.UA.EDU] on behalf of
Tony Harminc [t...@harminc.net]
Sent: Monday, April 3, 2023 3:14 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Su
Thanks
> On Apr 3, 2023, at 3:32 PM, Seymour J Metz wrote:
>
> PROBDESCAD=probdescad
>
>
> --
> Shmuel (Seymour J.) Metz
> http://mason.gmu.edu/~smetz3
>
>
> From: IBM Mainframe Discussion List [IBM-MAIN@LISTSERV.UA.EDU] on behalf of
> Tony Harminc
I spend a lot time in the code determining the abend and formatting all the
register I would like to save the info in a file
> On Apr 3, 2023, at 3:32 PM, Seymour J Metz wrote:
>
> PROBDESCAD=probdescad
>
>
> --
> Shmuel (Seymour J.) Metz
> http://mason.gmu.edu/~smetz3
>
> ___
It's a simple matter to do that in IPCS.
--
Shmuel (Seymour J.) Metz
http://mason.gmu.edu/~smetz3
From: IBM Mainframe Discussion List [IBM-MAIN@LISTSERV.UA.EDU] on behalf of
Joseph Reichman [reichman...@gmail.com]
Sent: Monday, April 3, 2023 3:38 PM
To:
On 4/4/23 02:48, Seymour J Metz wrote:
ITYM PDP-7.
Believe that if you want, it doesn't make it reality.
From the ACM:
"The C programming language was originally developed by Dennis Ritchie
at Bell Labs in the early 1970s, primarily for use on the PDP-11
minicomputer. The PDP-11 was a popul
On 4/4/23 03:14, Schmitt, Michael wrote:
"C is a language that combines all the elegance and power of assembly language with
all the readability and maintainability of assembly language."
I like that quote! The major difference is that C is portable, which is
why it's so successful.
B was designed for PDP-7 then ported to the PDP-11 and used to create C..
https://en.wikipedia.org/wiki/B_(programming_language)
On Mon, Apr 3, 2023 at 5:13 PM David Crayford wrote:
>
> On 4/4/23 02:48, Seymour J Metz wrote:
> > ITYM PDP-7.
>
> Believe that if you want, it doesn't make it reality
On 3/4/23 22:57, Seymour J Metz wrote:
ObDieJungfrauvonOrleans Believe that if you want' it wont change the reality.
I know it's reality as I'm an experienced C++ programmer. Modern C++ is
a thing. It's a term used to describe the C++ language from C++11 and
newer, when the language was reinv
In the museum at IBM Hursley, they have a deck of punch cards containing
the original PL/I compiler. With typical British wit somebody has
written "the UNCOBOL" on the side. That was probably what they thought
at the time, that PL/I was a COBOL killer. Funny how things turn out.
https://photos
35 matches
Mail list logo
