Bob, at 07:58 on Tue, Dec 20 1994, you wrote to Bill Grimsley ...

BG> Whereas proper software accepts anything at all. BTW, there are 
BG> quite a few points with 4 numerals in their address (apart from 
BG> Alex and Rachel, /nnn.1701 comes immediately to mind).

BL> I can't accept "anything at all" or it'll take *any* set of 
BL> numbers and send netmail there!

It'd be nice to make it fuckwit-proof, but isn't it better to allow any
combination on numbers (within reason of course), rather than restricting
it to a predetermined field of numerals?  Better to send the occasional
netmail to nul than have legitimate messages bounced IMO.

And what about netmail messages which are to be gated via UUCP, where the
current format is to send the message to UUCP{at}3:50/40 (entered in the To:
field of the message header), while the message body itself is addressed
much the same as QWK netmail is now, but with characters instead of
numerals (e.g To: diamondTS{at}aol.com)?

BL> Your comments have got me thinking again, and I think I'll change 
BL> to code to ignore the "To:" too... and just find the first address 
BL> (not in quotes) that fits the nnnnxnnnnxnnnn or nnnnxnnnnxnnnnxnnnn 
BL> pattern (n is an integer and x is anything else). It was a bit 
BL> fancy sending netmail from anyware, so I'll make it work *only* 
BL> from netmail.

In case it had escaped your attention, I'm no programmer, but for normal
QWK-type netmail, the above seems fine to me.  If I was you, I'd be seeking
a second opinion though. 

BL> What do you reckon?

UUCP stuff aside, it looks OK to me, but then what would I know? |-)

Regards, Bill

--- Msgedsq/2 2.3á