Re: Injection-Date and reinjection

["Charles Lindsey" <chl@xxxxxxxxxxxxxxxx>]

In <45AD232E.7070408@xxxxxxxxxxxxxxx> "Forrest J. Cavalier III" <mibsoft@xxxxxxxxxxxxxxx> writes:

>There is no concise way to describe a "proper" re-injection.  There
>are too many special cases and exceptions.  The role of such software
>is not clear.  Is it a relay?  Posting agent?  Gateway?

>I find 3.3.2 poorly written and redundant.

I tend to agree, but am not sure your rewrite is correct either. I think
we really need to resolve the Injection-Date Issue (#1416), and then come
back to exactly what we say here.

>I consider the mention of "special exceptions" an attractive nuisance.
>An implementor looking for justification will always consider themselves
>"special."   That's just psychology.

Which is why our drafts sometimes needs to say "Don't try to do this at
home, kiddies", but of course the problem is to find a polite way of
saying that. Essentially, the draft needs to be written so that not only
does it specify the correct tecnical details, but also so that it gives the
correct _impression_ of how things are _intended_ to work

>I think mentioning reinjection will entice novices to try something that only 
>experts can be expected to do without causing trouble.

Indeed. Reinjection needs to happen in a few cases which are hard to
specify precisely in advance. But the impression to be given is definitely
that "you would recognize those cases when you see them, but it takes
great knowledge and skill to recognize them", though of course you can't
word it like that. But saying less rather than more is probably a move in
that direction.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133   Web: http://www.cs.man.ac.uk/~chl
Email: chl@xxxxxxxxxxxxxxxx      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5