[net.news] Sites running old news

sjb (11/04/82)

Mark Horton commented that maybe we should discuss this in
net.news, so here I am putting in my two cents.  No matter
how undesirable it is for us when some sites use old versions
of news, there is nothing we can do about it.  We could threaten
to not send any news at all to them, but that would make us
no better than the Moral Majority and would be totally
childish (that may be a bit redundant, but I won't go into
THAT here!).  What we should do is point out to those sites
the advantages of putting up newer versions of news, advantages
to them as well as us.  For example, there are fewer bugs
(yes, there are bugs, but I would wager that if you got a copy
of old gamma 1.3 news (ah, the ol' days!) and compared it to
2.9 B news, you would find many more bugs in gamma), and there are
many new features (unsubscribing within readnews probably
being the most useful).  However, if a site is set to run
old news, we cannot force them (nor should we try) to change.

Adam

mark (11/04/82)

Well, we can set standards, which sites are required to conform to.
An enforcement mechanism for standards would be to refuse to forward
news from sites that don't conform, or alternatively, refuse to
forward news articles that individually don't conform (if these can
be detected).  Alternatively, we can leave things alone.

We've got a change coming up to the header format anyway.  A soon to
go out version of news will generate an internet "Reply-To" line and
optionally use it instead of the From line.  After a few months, messages
will not be guaranteed to have useful From lines, but instead will have
Via lines that list the systems the message has gone through.  The
obvious conversion date is the internet date: Jan 1, 1983.  In practice,
however, I don't see how we can convert that quickly.  I think a more
realistic date would be sometime in February or March of 1983.

While we're at it we have the option of setting standards, possibly
enforceable, for certain newsgroups.  For example, we can require
that followups have titles beginning with "Re: " and have "References: "
header lines.  This makes them detectable by the readnews program,
so people can use the "-f" option to disallow followups.  We can
require that the followup command not post to net.general, and that
all user interfaces implement a followup, so that people don't manually
post to net.general.  Currently the software does both, but we have
the option of eliminating one of these.  We can also change the followup
command to refuse to post to net.general, prompting for another newsgroup
instead.

Of course, the standard netnews software ("B news") would implement the
standard.  Sites who wish to run their own software (which has always
been permissible and will continue to be) would be responsible for
bringing their software into compliance.

	Mark

sjb (11/05/82)

The followup command not allowing followups to net.general is so
easily bypassed that it's unbelievable.  All one has to do is submit
a followup (with a 'Re:' in the title) to net.general directly from
inews.  Well then, you might say, we'll just hard-wire inews to not
allow titles with 'Re:' in them to go into net.general.  Well then,
people will simply not put 'Re:' in the titles and the only result
will be that inews will be even slower than it is now.  There is just
no way to restrict something like that.  And, as I said before, punitive
measures like not forwarding to or from a site that does not conform
to the standards set is not only childish (in my opinion), but it also
hurts the entire net as well as the site, for the rest of the net will
not be able to communicate with people at that site by this medium.
The only result that I can foresee if such things come to pass is
that the net will be broken up into little subnets, each running
their own software, and none communicating with the other.  Why
change a working system?  Sure, we get angry and upset when people
make mistakes, and we try to help them overcome them (sometimes
rationally, sometimes irrationally), but the entire purpose of this
net is to provide an atmosphere in which conversations and discussions
that are helpful (and relaxing) to people and businesses can take
place.  Why ruin that atmosphere with strict little standards that
can only lead to its destruction?

Adam

CSvax:Physics:crl (11/06/82)

#R:alice:-105400:pur-phy:8500001:000:1687
pur-phy!crl    Nov  5 17:53:00 1982

We run version 2.9 of B news.  However, as you can probably tell,
we also run notesfiles.  I don't know how many sites run notes,
but there is a big problem with this entire net.general/net.followup
debate--notesfiles doesn't ship responses to net.followup, but
perhaps could be made to.  However, this entire problem would not
exist if there were some way for B news to group responses with
their parents like notesfiles does.  Then, there would only be
1 main article about Big Macs/Whoppers, with umpteen responses
from others.  A reader could then have the option of skipping to
the next "base" article (to borrow a term from notes).  This would
eliminate the need for net.followup, and, in general, unclutter
a LOT of newgroups (especially net.misc) that have the same problem.
Besides, I've never really seen the benefit of net.followup.  If you
read net.general at all, then you must be interested in some of the
responses, so you probably read net.followup.  If you don't, then
when you do see something interesting that you want to hear more 
about, you have to go searching sequentially through net.followup
which has many, many articles you've never seen.  I guess the title
searching would help however, but I've never used it.  Unfortunately,
I can't really think of a good way to implement the response idea
in B news.  Perhaps articles numbered 1.1 is the 1st response to
article 1?  Perhaps articles with responses have a subdir?  C'mon
folks, here's a chance for some REALLY constructive ideas to improve
the B news system.  (Ideas for improvements to notes should go to
net.notes, by the by).

			Charles LaBrec
			(not afraid to followup my name)
			pur-ee!physics:crl