sjb (12/07/82)
(To me) This is *VERY* unappeasing. The whole concept behind USENET is to provide a bulletin board system of exchange, in which authors (who are supposed to be responsible and self policing) can freely post queries, ideas, information, replies, followups, etc. and may feel unhindered in carrying on a conversation. Digestification is an obstacle to achieving this goal. It will impose delays of perhaps days from the time a letter is mailed to the moderator to the time when the majority of the net will see it. This alone is a strong argument against it. It places a double burden on the net in terms of traffic handling: Each article has to make it way first to the moderator and then to the rest of the net. It imposes a HEAVY job on the poor suckers who would become moderators; moderating a digest would be no easy task and would be very time consuming. What we need now to help ease the load being put on the net and its member machines is not digestification but common sense, responsibility, willingness to cooperate and lend CONSTRUCTIVE criticism, and overall good will. We must work together to keep the net pleasant and clean (I do not mean clean in a moral sense (not to say either that we should be totally immoral!), but rather in a physical sense; i.e. remove dead weight, make things fast, etc.) and enjoyable for people to read and contribute.
woods@sri-unix (12/07/82)
I agree with Adam (alice!sjb) that digestification is a hindrance to the
speed and convenience of the net. The newspaper as separate pieces in the
envelope analogy is not quite accurate to describe non-digest newsgroups. It's
more like your newspaper arriving in a filing cabinet with everything neatly
organized.
I would like to point out that we have already had a similar debate once
before on USENET regarding net.space vs. fa.space, and at that time there was
a very strong consensus that the individual artice format was preferred. I see
no reason to think that has changed. PLease lets find another way. I don't
want to see everything digestified.
And as for newsgroup removal, I am glad to see Adam has the guts to do
something about it. The flak he has taken makes it very clear that we need to
set some sort of policy regarding creation/deletion of newsgroups. So far
I have seen suggestions that only a small elite be allowed to create/delete
newsgroups, and the way it is now anyone can. I think it is clear we need
a happy medium of some kind, and it is also clear we need desparately to do
*something*. At the very least the software should be robust enough to prevent
a new group from being created by a typo (e.g. net.joke). I think expiring
newsgroups is an excellent idea, combined with at least some restiriction
on the creation of extremely specific, low volume groups (e.g. net.wobegon,
net.sctv).
GREG
ucbvax!{hplabs,menlo70}!hao!woods
harpo!seismo!hao!woods
decvax!brl-bmd!hao!woodsbstempleton (12/08/82)
I think I should clear up some misconceptions that people have shown in their replies to me on this subject. 1) Having a moderator does not double net load. First of all only a certain percentage of posters are moderated (say 50%). Secondly, the only extra traffic is the mail traffic from the sender to the moderator. On an optimized path to a central site, this is perhaps 3-4 sites. A message to the whole net involves around 400 sites. This means a 1% extra load, not a 100% extra load! In addition, the reduced traffic means a great reduction in load, actually. 2) Most people I have talked to have been much in favour of the delay as a way of calming down flames etc. Most comments are not super urgent, and anybody who wants to put things through urgently can by bypassing the moderator. If the moderator is fast, the delay will not be long on some paths anyway. In addition, there can be several moderators for certain groups like net.general. If there were enough people, you could have one moderator per site! 3) You don't moderate all the groups by a longshot 4) The moderators job is no worse than the job of anybody today who reads everything in a group. I envision a program that prints each article just like news, and where it says [ynq] you would be able to put in simple commands like "Post" or "Reject <filename>" where filename is a form letter to the poster giving the reason for rejection. (this would work 95% of the time, and simple one or two liners could handle the special cases) Thus the moderator (in this case just a filter) has only a tad more load than current news readers. If he wants to actually make digests like the arpanet, that's fine but will take more work and software. 5) Are we sure that what we want here is a b-board? Even if that is what we want, can we afford it with hundreds of sites? As time goes, thre will be thousands and thousands of unix computers, and people like Lauren have put news on CP/M systems. Do you know what that means? Without moderators, there is no chance we can link up a network that large.
mclure@sri-unix (12/08/82)
#R:alice:-126500:sri-unix:8200005:000:652 sri-unix!mclure Dec 7 21:48:00 1982 And yet you would prefer that fa.sf-lovers and net.sf-lovers remain separate. For my money, combining redundant groups with individual message traffic between them is the way to go. net.sf-lovers and sf-lovers@sri-csl will remain gatewayed such that individual messages will pass back and forth. If people want to read the digests on fa.sf-lovers, fine; the digests will contain both Arpanet *and* Usenet messages. Also, the criticism that sf-lovers individual message traffic will choke Usenet is silly. There are only 10-20k bytes worth of sf-lovers messages per day and whether we get one digest or 13 messages is a trivial difference. Stuart