[Info-vax] BACKUP to the null device

[jbriggs444]

On Feb 9, 12:11 pm, koeh... at eisner.nospam.encompasserve.org (Bob
Koehler) wrote:
> In article <alpine.LFD.2.00.1102090949490.6... at libra.gmcl.internal>, Rob Brown <mylastn... at gmcl.com> writes:
>
>
>
> > I mean that this behaviour of NL: in BACKUP is undocumented.
>
> > Do you disagree?
>
>    I think BACKUP's behaviour is fully documented and not specific to
>    the NL: device.  Try it on the TT: device, sometime.  Try it on
>    an MBA:.  The behaviour of /save_set and when it is or is not on
>    by default is fully documented.  What BACKUP is doing when it is or
>    is not making a save set is fully documented.
>
>    It is not necessary to document each and every device if the
>    documentation is already sufficient to predict how it will handle
>    the device.  In the cited cases, it is.

I'm with Bob.  It's not a tape, so /SAVE_SET is not implied, so it's a
BACKUP copy operation (or a /PHYSICAL or /IMAGE media to
media backup).  BACKUP copy operations (or /PHYSICAL or
/IMAGE media to media backups) to the null device aren't
supported.

[It's not clear whether BACKUP checks output device type first
or for the /PHYSICAL and /IMAGE qualifiers first]

I suppose the argument is about whether a BACKUP copy
operation to the null device could or should be supported.

For me, the answer is clear:

Could: yes.
Should: what for?  Use /save_set [and /group=0 and /nocrc]
and get on with life.