davidh@utrc-2at.UUCP (David M Haynes) (11/10/85)
Gentlebeings:
Admittedly, I have not been reading net.news for very long (owing to having
access to a only a partial feed until very recently), but it seems to me
that an alternate solution to reducing the volume of the network could be
considered. [Again, this may have already been discussed, and I missed it]
I think it can be safely said that the back-bone sites have no problem
moving the majority of the news provided that the news being moved can be
shown to have some value.
The recent news cuts by henry@utzoo can be seen to be a direct result
of too much noise being distributed too far. A quick scan through net.sources
in the last little while will show a signal to noise ration of about 7 to 1
(7 "noise" articles to 1 "signal" article.) This can be shown in even
more clarity by looking at the recent messages from the European sites.
The vast majority of noise on the net is due to two things:
1. Inappropriate actions by a submitter. (submission to the wrong
group, submission rather than mail replies, etc.)
2. Submissions that are of no value outside the immediate area.
With respect to point one: Although it has always been the policy of the
network that "a submitter should police him/herself" this is clearly not
working effectively. Secondly, the solution of "peer pressure" by barraging
the submitter with mail messages has some very serious drawbacks. It may
serve to alienate the recipient who has made a honest error; it still may
cost the back-bone sites considerable money to process the mail messages.
With respect to point two: The network has a number of solutions for this
as well. (net.na, net.local, etc.) However, these safeguards are being
abused in the same manner as in point one. (I can attest to having no
interest in deals for automobiles available in Portland, but I still get
the messages about them in net.na.forsale)
So, clearly the net is unable to remain operating in the same manner as
it has in the past - witness the news cuts by henry@utzoo and the strong
message from Gene Spafford. However, the dropping of entire news groups
has its own problems - the signal is being thrown out with the noise!
The solution I would like to propose is actually based on a remark of Henry's
with respect to the creation of net.ont.sf-lovers to replace net.sf-lovers.
He said something to the affect that "I think this [the creation of localized
newsgroups] is going to become the way of the net." (Henry - sorry if that's
not exact, but I'm quoting from memory.)
The problem with a solely localized network is that the thinking tends to
become localized as well. I enjoy reading the views of others on the net,
especially those who are distributed geographically, who add a bit of their
culture to the net with every posting and make the net stronger.
What to do? I agree with Henry, the net will probably begin to form localized
groups for networking and general discussion, but I think that some careful
consideration should be given to a system of distilling the results from
these localized discussions and making them available to the rest of the
network, or something very valuable will be lost. Perhaps the system of
moderated newsgroups would be the best place for this, but, instead of having
one central moderator (who I am sure is overworked and under appreciated for
performing the task of sifting through all the submissions) there should
be several moderators (like the mod.uucp.map newsgroup). Then the vital issues
of each local newsgroup could be distributed over the larger network for
all and, since the noise is reduced locally, the volume of transmissions
would also be reduced.
-david-
David Haynes
..!utzoo!ecrhub!david
..!utzoo!yetti!utrc-2at!davidins_akaa@jhunix.UUCP (Kenneth Adam Arromdee) (11/14/85)
In article <164@utrc-2at.UUCP> davidh@utrc-2at.UUCP (David M Haynes) writes: >The solution I would like to propose is actually based on a remark of Henry's >with respect to the creation of net.ont.sf-lovers to replace net.sf-lovers. >He said something to the affect that "I think this [the creation of localized >newsgroups] is going to become the way of the net." (Henry - sorry if that's >not exact, but I'm quoting from memory.) Hmm, I didn't see anything about this in net.sf-lovers... would you please elaborate? -- If you know the alphabet up to 'k', you can teach it up to 'k'. Kenneth Arromdee BITNET: G46I4701 at JHUVM and INS_AKAA at JHUVMS CSNET: ins_akaa@jhunix.CSNET ARPA: ins_akaa%jhunix@hopkins.ARPA UUCP: ...{decvax,ihnp4,allegra}!seismo!umcp-cs!aplvax!aplcen!jhunix!ins_akaa ...allegra!hopkins!jhunix!ins_akaa
peter@graffiti.UUCP (Peter da Silva) (11/16/85)
Your suggestion would be equally well implemented by making the default
distribution as small as possible (state-wide, if you have that set up).
--
Name: Peter da Silva
Graphic: `-_-'
UUCP: ...!shell!{graffiti,baylor}!peter
IAEF: ...!kitty!baylor!peterron@brl-sem.ARPA (Ron Natalie <ron>) (11/18/85)
> Your suggestion would be equally well implemented by making the default > distribution as small as possible (state-wide, if you have that set up). > -- Of course, if you don't have a state or local distribution set up, the backbones could define a distribution called "regional" which means that the message would not get sent accross any backbone-to-backbone trunks. -Ron
peter@graffiti.UUCP (Peter da Silva) (11/21/85)
> Of course, if you don't have a state or local distribution set up, the > backbones could define a distribution called "regional" which means that > the message would not get sent accross any backbone-to-backbone trunks. That is a good idea. I like it... but then what do I know? I'm a bad guy. -- Name: Peter da Silva Graphic: `-_-' UUCP: ...!shell!{graffiti,baylor}!peter IAEF: ...!kitty!baylor!peter