Date: Mon, 11 Jul 94 21:37 EST
From: Raphael Freiwirth <RRF_+a_RCI_+lRaphael_Freiwirth+r%MHS+d_9006222E01F53FDC-CE7052220215880C%RCI_Incorporated@mcimail.com>
To: uri <uri@bunyip.com>
To: Sally Hambridge <sallyh@ludwig.intel.com>
Subject: Re: URC-Authors section
Message-Id: <93940712023739/0005242058NA5EM@mcimail.com>
MHS: Source date is: 11-Jul-94 22:57:09 EDT
Hi all (Especially Sally, Dirk, Dan and Steve)
ordinarily I am a very quiet listener on this particular mail list (there
seem to be enough people with opinions and ideas to progress ideas
without my help!)
However, my silence has ended...
My particular area of expertise is email, for a long time. I evolved
from a developer to an integrator to working for a company to try and
make it all work...
My biggest problem in email today is trying to make things work amongst
different email domains, and therefore different naming conventions.
Although X.400 came up with a schema that almost works (personal name),
the fields are not always long enough and still don't take into account
for certain needs of cultures (like what things really should look like).
The latest stab, using a field called "CommonName" really helps. This
is the field that you would use to give the correct representation of the
name. If you are doing something like a lookup for an author, this might
help. This CommonName MUST be accompanied by a surname. X.500 spells
this out as well, since in any culture, you really need both. The full
surname is the one that can be found in an insignificant amount of time
for a search, and keeps all the pieces of multiple pieced surnames
together. Where there are duplicates, the CommonName and Aliases can be
used.
The biggest problem is that you don't always want to keep "peoples"
names, the use of IBM was good (or ANY other company, branch,
application, etc). If you don't agree to some format (Kirk, how about as
a suggested implementor's guideline), you're screwed...
Trust me (and Steve). You'll never make a deterministic machine capable
of parsing any name field and coming up with the right answer everytime
(Dan, trying to come up with a way of resorting the name might make you
come up with a lot of wrong answers that a user or application will have
to figure out).
Lastly, I think Steve has got it right, you gotta do it here.
The suggestion of Sally's (and I believe Dan's) to go with lastname,
string is good.
The question is, what do you do with Generation Qualifiers? You can
build a scope on that (mostly deterministic as Steve pointed out, regards
to MR. I), so that it becomes
lastname GQ, string.
This field still needs an accompanying CommonName for further
clarification on names if possible. Some cultures may have a different
arrangement for their name components.
For companies, applications, etc, if no comma found, assume to be the
entire lastname and thus the searchstring. In this case CommonName and
the Name field should be identical.
Anyhow, my final question is why would you want to put the author's name
in the from or to fields of a message? Or maybe I just misinterpreted
that :.)
regards, ray
***********************************************************************
Raphael (Ray) Freiwirth
Currently Citicorp CGIN - Message Resource Center
My comments do not necessarily reflect opinions or positions of Citicorp
[x.400] /c=us/a=mci/pn=ray freiwirth phone: (703) 708 1153
[smtp] 5242391@mcimail.com
[MCI] 5242391