Subj : Re: Remember
To   : MATT BEDYNEK
From : FOXY FERGUSON
Date : Tue Aug 28 2001 11:51 pm

MB> *My* definition of compromise is anything that threatens to
MB> compromise a system.  A virus, a worm, or something that
MB> presents legal problems...ie. using a system to transmit
MB> illegal software.

I wasn't discussing "compromise", Matt...

MB> Should my software operate within specifications then I
MB> would accept no liability for loss of traffic due to other
MB> broken software. However I would try (within reason) to make
MB> allowances for it.

You missed the point, son. Let's try again... Scenario: there are
certain message specifications that are manditory and certain one
that are optional. Tosser "A" complies fully with all manditory
ones but NOT with all optional ones (implements, say, subset 1C
and 2B). Tosser "B" complies with all manditory ones but NOT with
all optional ones (implements subsets 1A, 1B, 3F, and 4C).

Now. If the optional specs recognized and allowed for by tosser
"A" are in *any* way different from those recognized and allowed
for by tosser "B", then you have two tossers that, in and of
themselves, fully comply with all manditory specifications BUT
can clash horribly when they try to interface with one another.

Neither tosser is "broke" but the use of the two \in combination\
on successive systems can give every appearance of something
being very wrong.

FF>> As another example, I have at least one mailer/tosser/bbs pkg
FF>> that I consider VERY broke because of the approach the author
FF>> took towards a certain critical issue, ..

MB> Heh.  What software would that be?

Platinum Xpress. I don't agree with the "dynamic" method Hector has
of handling netmail, since (in and of itself) it makes it impossible
for my site to route netmail.

--- Platinum Xpress/Wildcat! v1.4c
 * Origin: Pogo's (1:361/1)

.