PE> How do you implement an OPTIONAL kludge?

RS> Soorree, you havent grasped the difference between one like say the
RS> MSGID which is a single isolated optional kludge, and an optional
RS> kludge which gets changed at each step along the way. With THOSE,
RS> the optional concept falls down very badly when its FIRST used part
RS> way along the chain. THEN what you have in the kludge is fucked.

It doesn't fall down at all, it is simply the way it works.  How
would you have liked them to implement that OPTIONAL kludge?  It's
the most practical way there is, regardless of how you would have
liked them to implement it.

RS> Its just yet another example of when mindlessly attempting to
RS> proclaim what is and is not 'in spec' blows up in your face, JUST
RS> like it did on the embedded origin line between an SOT/EOT pair TOO.

No it didn't, Rod.

RS> Its nothing like as cut and dried as you are trying to claim.

It is.  You might like to state your exact position on the PATH line,
just who do you think is out of spec by not generating or updating
a PATH line when it is OPTIONAL?  I shudder to think who you would
claim is violating the spec on that.  BFN.  Paul.
@EOT:

---