Answering msg from andrew clarke to rowan crowe,
on Monday January 16 1995 at 20:50

 ac>>> It's certainly out-of-spec, that's for sure.  I don't think
 ac>>> it's a particularly serious issue though; I can't see most
 ac>>> current software failing because of this.

 >> I can. What about software which does "true" reply linking by
 >> ^aMSGID and ^aREPLY? The programmer may have assumed (quite
 >> reasonably, in this case) that the ^aREPLY string will be exactly
 >> the same as the original ^aMSGID string.

 ac> I would assume that most software that performs reply-linking on the
 ac> MSGID/REPLY fields would ignore the case of these fields.

    But that is an assumption on your part; where does it say in FTS/FSC that
programmers should ignore case? Why should I _have_ to ignore case when I
expect that the code on the other side will use the exact same ^aMSGID
string?

 >> MacWoof used to do other fun things, like use identical
 >> ^aMSGID serial numbers for different messages.

 ac> That's handy.

    MacKennel, sorry. They're all the bloody same. (-;

 >> Back to VFIDO...correct, it should be not be using the
 >> ^aMSGID line for any address information at all.

 ac> Indeed.

    But so many people still use it, unfortunately, as most of the time there
happens to be a valid ftn address present.

 >> And, since when has "FIDOnet" (sans quotes) been the correct
 >> capitalisation? (-:

 ac> Is the year 2000 a leap year?

    I'll wait until it comes around before deciding. See you at 23:59:59.99,
1999.

Best wishes, Ro.

---