On 2024-09-14 14:49, Dan Cross wrote:
So, that
was all before I started work on it. So the current VMS
Finger is probably a later version of the one maintained at PSU.
I'm not sure about this part; as I understood it, Bryan wrote Penn
State finger from scratch. I find a version of it online:
https://www.digiater.com/openvms/freeware/v40/finger/psfinger/
But it seems that it's getting hard to come by. On my Alpha, I found
version 1.5, but I can't find that online (at least not in my cursory
search). I'm happy to upload that somewhere if anyone would like.
The aaareadme.txt file there says:
| DECUS FINGER DIFFERENCES
| ========================
| This is a list of things I can think of which differ.
|
| /HELP to the server is the only documentation.
| This is a munged copy of the DECUS Finger help file.
| I haven't given documentation any serious consideration yet.
|
| This is a clusterwide finger (/CLUSTER is default).
|
| New mail message headers are never listed (I haven't yet figured
| out how to match addresses so it only displays messages from the
| person doing the Finger).
|
| There is no Finger database. Names are read from the SYSUAF
| owner field. The entire SYSUAF is read at startup, and updated
| as needed for new accounts. There is not presently an easy way
| to disable printing the owner name.
|
| Added fields for account and nodename.
|
| No support for /SYSTEM, /IDLETIME, /TTTYPE, or /SWAPPED.
|
| /SORT only supports account, CPU time, Login time, and Username
| (Username is the default).
|
| Differs a bit in what an interactive process is.
|
| Only supports one filename for finger plan (default FINGER.PLN).
| Also prints a project line (FINGER.PROJECT).
|
| The format of the process listing (which fields can be displayed,
| which columns, and how wide the fields are) can be easily configured
| at assembly time.
|
| No support for Jnet, or TCP other then UCX/Multinet, at this time.
|
| The server will not forward Finger requests. Commands like
| Finger @watsun@cunixc are not valid.
So it seems that there was some contributions from one to the other
(be-
fore my maintainership).
The PSU one is in Macro-32. That's the sign of someone who REALLY
likes
Macro-32. DEC folks would probably have used Bliss-32. The Columbia one
(which turned into DECUS Finger) is written almost entirely in Fortran,
probably because Fortran was the only sensible cross-architecture
language
of the era (while I'm pretty sure one could write a Finger
implementation
in COBOL or RPG II, I don't envy the life of anyone who would try it
8-).
I'm not sure what the state of cluster support is in DECUS Finger - by
the time I released 5.1.35 I only had a single Alpha node. I'll have to
see how it is on a cluster and add that support if not present for
V52.2.