Re: Status for faces patch #14.

Ignatios Souvatzis (u502sou@mpirbn.mpifr-bonn.mpg.de)
Wed, 27 Mar 91 13:45:28 +0100

> From: John Mackin <john@syd.dit.csiro.au>
> From u502sou@mpirbn.mpifr-bonn.mpg.de (Ignatios Souvatzis)
> As far as I remember, RFC822 states that 'domains' and 'mailboxes'
> ignore case. I frequently get mail with domain or subdomain names
> uppercase or capitalized. 'gethostname' on the DECstations of the
> university's radio astronomy institute returns upper case names
> (probably set up this way to get it compatible to DECnet
> conventions, they talk DECnet to the VAXes...). [Patch included].
>
> distressingly ambiguous, and Mr. Souvatzis is absolutely wrong.
> Please, everyone, refer to RFC822, section 3.4.7, `Case Independence'.
> I won't quote it at length here. The point, in simple language
> and without using token names from the 822 grammar, is that
> everything on the right side of the `@' matches in either case,
> _but everything on the left of the `@' matches in exact case_,
> with the exception of the reserved, special name "postmaster"
> which must match in either case. (I am oversimplifying here
> by ignoring source routing; if I may use token names, <domain>s
> do indeed match in either case, but <mailbox>es include
> <local-part>s, which do _not_.)
>
> Please, people. RFC822 is all that stands between us and...
> and... (he shudders, and makes an odd Qabalistic sign in the
> air...) (dare I?) (YES!) X.400!!!!! Let's get it _right_.
>
> John.
>

Sorry, sorry. Mea maxima culpa.

As I wrote in my comment to my patch, I only REMEMBERED RFC822, I
didn't READ it.

a) I wrote 'mailbox' instead of 'local-part'.

b) 'local-part' would have been wrong, too, as John pointed out. (I
just loaded rfc822 from the tape --- from the end of a very big tape
--- and looked it up.

I only want to insist on the 'domain'. In my patch, I patched for the
'domain' comparison (and made it case-insensitive); I didn't make the
'local-part' comparison case-sensitive (one line above it in the
source). I sort of believed half of the program source and half of
rfc822... Maybe Rich B. can do that? Should he?

Btw, what should we do with X.400 addresses sneaking in into our
'real' net?

Ignatios