Date: Wed, 5 Mar 1997 10:37:08 +0200 (IST) From: Eli Zaretskii To: Jesse Bennett cc: djgpp AT delorie DOT com, DJ Delorie Subject: Re: c.o.m.djgpp retro-moderated? In-Reply-To: <5fhksu$jc8$1@superb.csc.ti.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On 4 Mar 1997, Jesse Bennett wrote: > sword. It will reduce the email traffic but list subscribers who are > interested in these (or similar) topics will be excluded from the > discussions altogether. DJ, can you please explain what would retro-moderation mean for subscribers to the mailing list? Thanks. > I feel obligated to reiterate my concerns about the R-M approach that > has been proposed. This seems a very slippery slope, allowing > postings to be anonymously canceled with no accountability. There are > some technical issues here as well. What is to stop a disgruntled > individual from canceling legitimate posts as well (or worse, running > a CancelBot which targets his "enemies")? I think you are pushing this issue way too far. The authority to cancel messages is not given to everyone: the news group has to come up with a (hopefully small) number of individuals who are *trusted* by the group to exercise this authority only in those cases where the postings don't belong to the group. Thus the risk of *arbitrary* cancellation on *personal* or other irrelevant grounds does not exist; making that risk a real concern could IMHO dupe people into thinking that this is a moot issue. Trust is the basis for any DJGPP-related activity. (As a matter of fact, most human activities in a society involves certain degree of trust; you won't be able to make any significant progress in almost any field without trusting that certain ``rules of the game'' are mostly kept by others.) My real concern is how well *trustworthy* and *well-meaning* individuals can indeed classify the borderline postings in a way that doesn't prevent useful information from getting to people who might find it helpful. > Independently of any moderation disscussions I think it is worthwhile > to review the group charter periodically. What should be considered > on-topic for the group? For example, none of the following are djgpp > specific - should the group charter ban these discussions? > > * Discussions about gcc in general (language extensions, > optimizations, assembly language programming, etc.). > > * Discussions about C programming methods. Tips, tricks, etc. > > * Discussions about porting code using gcc but not djgpp specifically. > For example, I might be interested in porting an application using > Linux. Since the code should compile with little or no > modification under djgpp it seems likely that there might be some > interest in the djgpp community. These are all those ``borderline'' cases that worry me. My best answer to the above question is that discussions on these subjects are *somewhat* relevant to DJGPP, therefore they should *not* be banned. However, it is true that more often than not, these threads gradually become less and less relevant to DJGPP as they evolve. It would be wonderful if people had a degree of self-restraint to stop pursuing the issue (publicly; private email is OK) after a certain point. But it's naive to expect something like that in a real world. Even reviewing the group charter won't help here, I think (for example, I estimate that about 30-40% of the messages are covered by the FAQ, but people still post them, although the charter says they shouldn't). The *real* issue here is not whether a bunch of criminals will take control of this news group's traffic, the issue is this: how much are we annoyed by the noise that we get on an unmoderated group, and how much can we trust our trustees to let them cancel and/or re-route some of the messages. That is the issue that DJ was talking about; FWIW, I agree that it *should* be raised and discussed by everybody who cares to make their views public.