[135] in Zephyr_Comments
proposal for a new feature in zwgc
daemon@ATHENA.MIT.EDU (Mark Lillibridge)
Mon Nov 7 16:01:44 1988
Date: Mon, 7 Nov 88 16:00:40 EST
From: Mark Lillibridge <chariot@ATHENA.MIT.EDU>
To: zephyr-comments@ATHENA.MIT.EDU, athena-ws@ATHENA.MIT.EDU
Reply-To: chariot@ATHENA.MIT.EDU
I would like to propose that zwgc be extended to allow obtaining
the full name and nickname of the sender. (I will explain why the
nickname in a minute or two) That this is highly desired functionality
is clear from the large (>20) number of people using zfwrite (fake
zephyr) just so that they can send along their full names. I am sure
that if more people knew about zfwrite, they would also be using it to
do this.
That the functionally is needed is clear. There are, however,
several methods of providing it. One way which has been put forward
already is the so called "zwrite-signature" method. The idea here is
that each user can set the zephyr variable "zwrite-signature" to a name
and then whenever he does a zwrite, this name will be passed along as a
seperate field which the reciever can then decode and display. For
example, instead of getting "From: chariot" which probably means very
little to people who don't know me, they would get "From: chariot <Mark
Lillibridge>" which both allows them to reply and lets them know who
really sent the message.
Aside from the fact that the implementation of this was totally
screwed up in 6.0R (I'll be happy to provide horror stories on request),
this proposal has one major flaw. The flaw lies in the word "can". In
order for a user to have a reasonable signature, he must activately set
a "wizard-level" zephyr variable. How many people do you think will
stumble across this? To date, I believe I have seen a total of 5
non-athena people using this. (This in part explains why people are
using zfwrite for this purpose...) At the very least, this should be
changed so that the signature field is by default set to the user's full
real name. This allows him/her to still change his/her signature while
still insuring that novices will have reasonable signatures.
There is another drawback to the zwrite-signature proposal: it
requires every zwrite message to contain an extra 20-30 characters for
the person's name. My proposal eliminates this overhead. What I
propose is that whenever zwgc receives a packet, it calls hes_resolve to
obtain the passwd info for the sender (if any) and place the resulting
information (well, the full name and the nickname fields anyways) into
zwgc variables, say realname and nickname. Needless to say, this can be
implemented in more efficient ways, but you get the idea. This allows
the reciever to display if he wishes the senders real name and/or
his/her nickname. Note that there is no real net overhead here because
hesiod caches information.
The reason for providing the nickname field as well as the real
name field is so that users can use a shorter/different name for their
"signature" without having to give up on having their real name in the
sms database.
Well? What do people think?
- Mark
