Xref: utzoo unix-pc.general:6624 comp.sys.att:10998 news.groups:25977 Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!mit-eddie!andante!alice!wilber From: wilber@alice.att.com (Bob Wilber) Newsgroups: unix-pc.general,comp.sys.att,news.groups Subject: Re: comp.sys.3b1.*? Summary: The Correct View on the Matter Message-ID: <11686@alice.att.com> Date: 30 Nov 90 02:15:09 GMT Reply-To: wilber@homxb.att.com Organization: AT&T, Bell Labs Lines: 95 In article <36335@cup.portal.com> thad@cup.portal.com (Thad P Floryan) writes: >... >2) separation of UNIXPC-related source from the general discussions seems a > good thing, because it's my observation that any newsgroup containing the > letters "*source*" enjoys both a longer expiry time and automatic archiving. Yes, and administrators provide the long expiration times and automatic archiving because they know that a *moderated* source group isn't full of junk that nobody's going to want to look at two weeks later. > ... >3) as for a moderated *.sources, my feeling is we propose whatever it takes > to assure a vote victory. Should be NO problem! The last posting to > comp.sources.misc was Oct 14, asking for a new moderator, hence the > incredibly increased traffic in alt.sources (which is NOT well-distributed). > > My preference is for NO moderation since the inclusion of a moderator in > the distribution chain increases the time-delay for what "may" often be a > critical need (much like my "extra drains in the pipe" example :-) > > "Kwik 'n dirty" sources such as my recent {send;pass;recv} would continue > to be in the general discussion newsgroup because the sources were only > incidental. If a separate sources group is to be created at all, it should be moderated. First, that will increase the probability of it passing, since the Guardians of the Net (TM) are much more likely to go thumbs down on an unmoderated sources group. More important, it is the only way to keep out posts like "Where do I get patch?" and "The Evils of Copyleft". Also, bug reports do not belong in sources groups. If you find a bug, send E-mail to the author. If you *must* let 1,000 people know of your troubles, that's what comp.sys. is for. The rule of thumb is: If it's not worth archiving for a long time, it doesn't belong in a sources group. The moderator should only ensure that the article posted is of the right sort, he should not be expected to check the code for viruses, track down bugs, or even ensure that the stuff compiles. So the delays shouldn't be too bad. The current bottleneck in comp.sources.misc is an unusual situation -- until recently the posting delay to that newsgroup has been very reasonable. Without moderation sources groups simply become discussion groups with an occasional posting of source. >4) there should be only one "group" which combines all of the existing > unix-pc.general, unix-pc.bugs, unix-pc.uucp, and unix-pc.test > > It appears we all read ALL the stuff anyway, so a SINGLE, central location > appears to be a logical change which also reduces newsgroup "clutter". > Correct. >5) so now we focus our attention on what the in comp.sys. and > comp.sources. should be named. > > My belief is the choice for should be orthogonal in the sense that a > comp.sys.FOOBAR would have a corresponding comp.sources.FOOBAR. Yes, this > is not quite the existing scheme of things, but just because someone else > screwed up does not mean we should continue the SNAFU; let's show them we > can do it RIGHT. > ... > This leaves us with the other extant identifiers "UNIXPC" and "3B1". The > BYTE Magazine system review (May 1986) was headlined "The AT&T UNIXPC". > All documents of any consequence refer to the system as UNIXPC. All use > I've seen of "3B1" has been only in this newsgroup, though I often refer > to the system as "3B1/UNIXPC/PC7300" in other newsgroups. > ... > So, in summary, I'm proffering: > > comp.sys.unixpc, and > comp.sources.unixpc > > and also for the archive name at osu-cis to be changed from "att7300" to > be whatever finally gets voted and approved. If you think an alternative hierarchy called "unix-pc" gets a fair number of '386 BIOS posts, wait until you have a main hierarchy group called comp.sys.unixpc! I agree that "Unix PC" is the name AT&T used, and that "3b1" is a bit of last minute marketing mumblage that doesn't even appear in the documentation. But nowadays there are lots of "Unix PCs" having nothing to do with the 3b1 so making a group called unixpc is just asking for a never ending stream of inappropriate postings. We're outnumbered, guys. IBM won its marketing war, with the result that the very word "PC" now means "IBM 80x86 compatible". (Which is definately not politically correct!) AT&T lost its marketing war, so "Unix PC" now means whatever people want it to mean, and usually they want it to mean an 80x86 running Xenix. So ah'm stickin' to mah guns and shooting for comp.sys.3b1 and comp.sources.3b1. I don't object strongly to comp.sys.att.3b1 and comp.sources.att.3b1, although I don't see much of an advantage to the longer names either. If "3b1" seems a bit obscure, well, it keeps the riff raff out. (In reference to another post, perhaps it should be called comp.sys.cult-unix-box. :-) Bob Wilber wilber@homxb.att.com