[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [leafnode-list] [1.9.*][2.0b*] Bug



Matthias Andree wrote:

> krasel@xxxxxxxxxxxxxxxxxxxxxxxxxxxx (Cornelius Krasel) writes:
> 
> > If #1 is (for some reason) not reliably connected to the network, it
> > will not propagate the article. In that case, posting over #2 (and
> > possibly #3) will ensure propagation of the article.
> > 
> > If the article is on #2 already, it can be safely unlinked since it
> > has been propagated over the network.
> 
> This is a play against the time of propagation between #1 and #2. If #1
> feeds #2 fast, that's alright. If it feeds #2 slow, you never know if it
> propagated properly. Of course, leafnode then posts on #2 and on #3 as
> well, but what's the use?

It increases the likelyhood that the article will propagate through the
whole network.

I use currently four servers, two and a half of which I am allowed to
post to. I got myself an account on news.cis.dfn.de because not all
news posted via news.informatik.uni-wuerzburg.de (esp. to various alt.*
groups) was propagated properly. After implementing multiple posting
in the current leafnode scheme, everything worked out fine. Currently,
most (if not all) news is propagated more quickly from news.informatik.
uni-wuerzburg.de to news.cis.dfn.de than fetchnews manages to post on
the second server.

BTW, since you seem to be aware of at least some internals of INN:
what's faster:

HEAD <msgid>

or

XOVER <msgid>

? ismsgidonserver() currently uses the first scheme, but I might switch
to the second, if it's advantageous. (It used to use STAT but it turned
out that nntpcache only returned nonsense in that case.)

--Cornelius.

-- 
/* Cornelius Krasel, U Wuerzburg, Dept. of Pharmacology, Versbacher Str. 9 */
/* D-97078 Wuerzburg, Germany   email: krasel@xxxxxxxxxxxxxxxxxxxxxxxxxxxx */
/* "Science is the game we play with God to find out what His rules are."  */

-- 
leafnode-list@xxxxxxxxxxxxxxxxxxxxxxxxxxxx -- mailing list for leafnode
To unsubscribe, send mail with "unsubscribe" in the subject to the list