[sci.astro] comp.numeric deferred; try Stanford NA list

woods@hao.ucar.edu (Greg Woods) (01/28/88)

In article <40114@sun.uucp> dgh%dgh@Sun.COM (David Hough) writes:
>In particular the Numerical Analysis mailing list at Stanford
>was suggested as an alternative by many people, most of whom
>know it by the name "sci.math.num-analysis" by which it is
>available at some sites.

  I don't know about anywhere else, but while the newsgroup for
sci.math.num-analysis exists here (we are an Internet site and should
be getting the inet distribution) there are rarely any articles in it.
If the mailing list *is* being gatewayed into the newsgroup, it is broken
somewhere between here and the gateway point. Where is the gateway? I would
like to see this get fixed because there is a lot of interest in that topic
at our site.

>	na.problem@score.stanford.edu

  This is a hideous address to use. There should NEVER be periods in the local
part of an address. Many mailers will be confused by this. The user name
(the part to the left of the @-sign) should be a single token. Maybe THAT'S
why we never see any articles! :-)

--Greg

sow@cad.luth.se (Sven-Ove Westberg) (01/30/88)

In article <1116@hao.ucar.edu> woods@hao.UUCP (Greg Woods) writes:

>>	na.problem@score.stanford.edu

>  This is a hideous address to use. There should NEVER be periods in the local
>part of an address. Many mailers will be confused by this. The user name
>(the part to the left of the @-sign) should be a single token. Maybe THAT'S
>why we never see any articles! :-)

>--Greg

READ RFC822 Page 27. !!!!

Periods is allowed in the local-part in RFC822, period.

Sven-Ove Westberg, CAD, University of Lulea, S-951 87 Lulea, Sweden.
UUCP:    {uunet,mcvax}!enea!cad.luth.se!sow
Internet: sow@cad.luth.se

spaf@cs.purdue.EDU (Gene Spafford) (02/01/88)

In article <959@luth.luth.se> sow@cad.luth.se (Sven-Ove Westberg) writes:
 >In article <1116@hao.ucar.edu> woods@hao.UUCP (Greg Woods) writes:
 >
 >>>	na.problem@score.stanford.edu
 >
 >>This is a hideous address to use. There should NEVER be periods in the local
 >>part of an address. Many mailers will be confused by this. The user name
 >>(the part to the left of the @-sign) should be a single token. Maybe THAT'S
 >>why we never see any articles! :-)
 >
 >READ RFC822 Page 27. !!!!
 >Periods is allowed in the local-part in RFC822, period.

I'm sure Greg has read RFC822.  He didn't say that it was illegal to
use a period to the left of the @ sign -- simply that it should not be
done because some mailers get confused.  Unfortunately, not every
mailer strictly implements 822 (or 976) addressing, and some will do
the wrong thing when presented with an address having a "local part"
with a period in it.  Attempts should also be made to avoid using "%"
and "!" symbols in mailbox names for similar reasons.
-- 
Gene Spafford
Dept. of Computer Sciences, Purdue University, W. Lafayette IN 47907-2004
Internet:  spaf@cs.purdue.edu	uucp:	...!{decwrl,gatech,ucbvax}!purdue!spaf

woods@hao.ucar.edu (Greg Woods) (02/02/88)

In article <959@luth.luth.se> sow@cad.luth.se (Sven-Ove Westberg) writes:
>READ RFC822 Page 27. !!!!
>
>Periods is allowed in the local-part in RFC822, period.

   Sorry to repost on this, but I'm getting inundated with mail telling me
that periods are legal in the user part of the address according to the
standard (RFC-822). I *KNOW* that. I said the address was "hideous",
not "invalid". Sorry for the confusion. My point is that there are mailers
out there that will choke on a multiple-token user part of the address.
That these mailers are not standard-conforming is true but irrelevant.
What is important is 1) how many of them are there, and 2) what is the
probability of them getting fixed? The answers would seem to be 1) no one
really knows, and 2) next to nil. 
  All I can say about is is that if *I* were running a mailing list, I
wouldn't have multiple-token user parts of addresses. That's all. It
is up to the mailing list maintainer to decide if he wants to worry
about this or not.

--Greg