On Thu, 5 Jan 2017 13:50:25 -0500, Carmen Vitullo wrote:
>for my own sake I checked some sysout of my own and if you scroll over to the
>right in the jes2 dataset view from ? you see jes2 dataset attributes, I see
>RECFM V and U and VA and UA, no VS for any jes2 datasets
>
Ah, but did you allocate one? When I do that BPXWDYN(), I see:
SDSF JOB DATA SET DISPLAY - JOB LONGOUT5 (JOB05373)
COMMAND INPUT ===>
NP DDNAME StepName RecFM LRecL
JESMSGLG JES2 UA 133
JESJCL JES2 V 132
JESYSMSG JES2 VA 133
SYSTSPRT SETUP VBA 133
TESTO1 SETUP VBS 32748
Experimenting further, JCL seems to impose a limit of LRECL=32761;
ISPF and BPXWDYN a limit of LRECL=32760. I see no rationale for either of
those numbers -- the only one that would make sense is 32767, the max
signed halfword value. Allocating with ISPF assigns halftrack BLKSIZE;
no information for JCL because I never wrote to it.
And when BPXWDYN successfully allocates with LRECL=32750, ISPF
Data Set Utility reports LRECL=32756.
>----- Original Message -----
>From: "Jesse 1 Robinson"
>Sent: Thursday, January 5, 2017 12:40:16 PM
>Subject: Re: IEC141I 013-A8: how to read VS data sets?
>
>Like Walt, I'm skeptical of the assertion that a spool data set is VS, where
>'S' means 'spanned'. Spanned records are extremely rare in MVS. The only use
>for them I've ever come across is MANx SMF, which I believe predates dirt.
>Most standard utilities fail when opened to V(B)S records. Only IFASMFDP can
>read SMF data, and the output from that is ordinary VB.
>
I've been told that VBS was invented to support FORTRAN, so that FORTRAN
programs could move from tape to DASD.
And, for whatever reason, Rexx EXECIO has very lately been granted capability
for RECFM=VBS (and for RECFM=U).
--gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
