F$LOGICAL

[Ewa Skotnicka]

Does anybody know when the lexical function F$LOGICAL was replaced by F$TRNLNM and if F$LOGICAL is still supported (but undocumented)? Or if someone could advice me where can I access old VMS documentation, especially release notes or DCL to check it myself.
Thanks.

[Jan van den Ende · Accepted solution]

Ewa,

The $LOGICAL functionality (including the DCL lexical) was superceeded by $TRNLNM with the introduction of VMS 4.0 (cluster support, and a reorganised SYSTEM DISK).

Of course, being VMS, they did NOT break the $LOGICAL functionality. It is STILL there, but discouraged. AFAIK (anyone with definitive info please jump in) the $LOGICAL functionality is now implemented as a jacketed call to $TRNLNM.
We still have some applications (without source code) with embedded use of $LOGICAL. Can be rather clumsy to work with.

On the documentation part I only have my own memory, which is not really suited to posting here :-) although for specific details it usually is sufficient.


hth,

Proost.

Have one on me.

jpe

[Wim Van den Wyngaert]

But they still use it in their own doc.

OpenVMS Guide to System Security
... f$logical("SYS$INPUT") then deassign sys$input $ read/end=next_command/prompt="$ "
sys$command ... This enables these users to log in at the OpenVMS login prompt ...

OpenVMS System Manager's Manual
... f$logical("SYS$INPUT") then deassign sys$input $ read/end=next_command/prompt="$ "
sys$command ... SYSMAN> ALF ADD VMS:.ZKO.VMSORG.SYSMAN.CLIENT1::SYSTEM FOOBAR. ...

[Uwe Zessin]

Yes, VAX/VMS V4.0 - the logical name subsystem was changed as well, which I'd say was the main reason for the change. I don't know if it is still present in today's documentation, but some system services for logical name handling had to be reworked as well.

[Wim Van den Wyngaert]

It's even actively used in current version of VMS. Found 156 matches in *.com of 7.3.

Wim

[Jan van den Ende]

Wim,

just curious.

How many of those are in TCPIP and other "recently" ported U8X-ware?
I too have occasionally stumbled over F$LOGICAL in places where I expected F$TRNLNM, but not that I recall in true "native" VMS stuff.
I would wish that those that do the porting would (before they begin!) get a somewhat deeper training in VMS, like the logical name system structure and the system disk cluster environment!
I still have nightmarish memories about the (already retired) SAN agent!

Proost.

Have one on me.

jpe

[Ewa Skotnicka]

Thank everybody, have (a better) VMS we all !
Ewa

[Wim Van den Wyngaert]

Jan,

Not in the tcp stuff.

Not acceptable are :

autogen.com, shutdown.com and all decw stiff.

Wim

[Robert Gezelter]

Ewa,

I don't have the documentation or the client sources to confirm it, however my recollection is that F$LOGICAL did not support the use of the LNM$FILE_DEV (in table LNM$PROCESS_DIRECTORY; defaulted in LNM$SYSTEM_DIRECTORY).

LNM$FILE_DEV can be used to insert logical name tables into the search sequence. It is used by some products (e.g., Oracle), and is fully documented. I did a 2004 OVMS Technical Journal article which demonstrated what can be done leveraging LNM$FILE_DEV (the Journal article can be found at http://www.rlgsc.com/publications/vmstechjournal/inheritance.html).

My recollection is that current "superseded" version of the LOGICAL interface ignores the sequencing provided by LNM$FILE_DEV and is hardwired to search the logical name tables in the default order.

I do not have time this morning to check my recollection experimentally. If my recollection is erroneous, my apologies.

- Bob Gezelter, http://www.rlgsc.com

[Jan van den Ende]

Bob,

no apologies required, your description fits what I know about it!

Proost.

Have one on me.

jpe