On Sun, 29 Mar 2020 23:45:21 +0000, Seymour J Metz wrote:

>No,
>
??? "No" to which of my questions/conjectures.

We were novices except for our mentor who was heavily committed
otherwise, and the intellectual overhead of mastering NOTE/POINT
and remembering the offset in the block outweighed the
computational overhead.

I suppose we could kept and reused open DCBs to the highwater
mark of nested includes (usually just 1), but we didn't do even
that.

>and if there's only on DD it's more overhead to do multiple OPEN compared to 
>multiple FINDs.

________________________________________
From: Paul Gilmartin 
Sent: Sunday, March 29, 2020 5:45 PM

On Sun, 29 Mar 2020 21:01:51 +0000, Seymour J Metz wrote:

>Does SAS/C actually open as BPAM and use point, or does it just stuff the 
>member name in the JFCB and use BSAM/QSAM?
>
Does it work to "just stuff the member name in the JFCB and use BSAM/QSAM"
if the DDNAME refers to a concatenation?

Is the advantage of the "stuff" technique that QSAM handles blocking, BPAM
does not?

I suspect that many languages support BPAM at translation time; fewer at
execution time.

In a FOSS Pascal processor I was associated with, we used BPAM, at both
compilation and execution time.  We eschewed NOTE/POINT for nested
INCLUDE members and simply opened additional DCBs on the same DDNAME.
(This wasn't your 32K 360/20).  Worked fine on MVS; failed miserably on CMS.

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to