From: Charles Lindsey (chl@clw.cs.man.ac.uk)
Date: Wed Mar 08 2000 - 04:20:54 CST
In <yl8zzu3bf8.fsf@windlord.stanford.edu> Russ Allbery <rra@stanford.edu> writes:
>Charles Lindsey <chl@clw.cs.man.ac.uk> writes:
>> Sure, if in-place updates are possible, the whole problem goes away.
>Agreed. That's another of the reasons why I want them.
OK, I think it is now established the the history file is not the
problem. Maybe I should update the pseudo comment text in the draft to
reflect our better understanding.
>> Fine! Do we actually need Xref header in articles anymore if the
>> information is in the voerview?
>Yeah, for slaving. But otherwise, not really for readers, if they can get
>it from overview.
Then surely the answer is for the master to pass it to the slave
out-of-band. I.e. as a separate parameter in some convenient notation, not
in the form of a header inside the article (which would then have to be
edited in situ by the slave).
>> Anyway, the missing mapping is the one from API to article numbers. That
>> is the one we need do the Replaces properly. It can be done by going to
>> the article and extracting the Xref, but there may be simpler ways.
>I don't know of a simpler way with INN's current design. The storage API
>token is an end-product sort of thing; it's not intended to be converted
>back into anything other than the actual article.
So it _will_ be necessary to consult the article, or can it be got from
the overview? But Replaces will be relatively rare.
BTW, if I ask for an article by Message-ID, how can I get the overview
entry for it? Or is it the case that I shouldn't ever need the overview in
that case?
-- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Email: chl@clw.cs.man.ac.uk Web: http://www.cs.man.ac.uk/~chl Voice/Fax: +44 161 437 4506 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