goudreau@dg-rtp.dg.com (Bob Goudreau) (11/10/89)
In article <73774@tut.cis.ohio-state.edu> Loyde W Hales <lwh@cis.ohio-state.edu> writes: > >>From wfp@dasys1.UUCP (William Phillips) writes: >>Are you trying to suggest that since comp.sys.ibm.pc is full of garbage, some >>of the garbage should be moved into comp.lang.c? Look, just because there's >>something wrong with comp.sys.ibm.pc doesn't mean it should infest comp.lang.c. >>Quite the contrary, in fact. > >I don't think he's saying that at all. I think he's saying that *.ibm.pc are >broad-set newsgroups, covering every aspect of the machines. This isn't >efficient for questions dealing with C. Hmmm. By that logic, the following diametrically opposed statement is equally reasonable: "I think he's saying that comp.lang.* are broad-set newsgroups, covering every aspect of the languages. This isn't efficient for questions dealing with IBM PCs." Perhaps what we really need to do is to split the comp.lang.c newsgroup into comp.lang.c.definition and comp.lang.c.implementations :-). ------------------- Bob Goudreau +1 919 248 6231 Data General Corporation ...!mcnc!rti!xyzzy!goudreau 62 Alexander Drive goudreau@dg-rtp.dg.com Research Triangle Park, NC 27709, USA
rob@kaa.eng.ohio-state.edu (Rob Carriere) (11/10/89)
In article <165@xyzzy.UUCP> goudreau@rtp48.dg.com (Bob Goudreau) writes: >In article <73774@tut.cis.ohio-state.edu> Loyde W Hales <lwh@cis.ohio-state.edu> writes: >>I don't think he's saying that [*.ibm.pc are overfilled with garbage so we need to siphon some of it over to here] at all. I think he's saying that *.ibm.pc are >>broad-set newsgroups, covering every aspect of the machines. This isn't >>efficient for questions dealing with C. > >Hmmm. By that logic, the following diametrically opposed statement is >equally reasonable: > > "I think he's saying that comp.lang.* are broad-set newsgroups, > covering every aspect of the languages. This isn't efficient for > questions dealing with IBM PCs." > >Perhaps what we really need to do is to split the comp.lang.c newsgroup >into comp.lang.c.definition and comp.lang.c.implementations :-). Perhaps what we really need to do is to stand back and _think_ for a moment. Why is the news split up in groups? So as to cause religious wars? Generate more posts? Or, perhaps, to help us digest the stuff? Ah, then it is a heuristic. Not a sacred law, but a heuristic. Like any heuristic worth its salt, news group separation needs to applied intelligently, or things get silly. If somebody wants to start a discussion about Budhism, this is clearly not the place. If somebody wants to know whether the piece of code {....} is portable, it is. In between, things get more vague. The reason? Humans are notoriously bad at categorizing, they work with fuzzy sets rather than the normal ones. So there is a mismatch between us and the newsgroup structure. So it won't quite work perfectly; what my fuzzy membership evaluator says is still acceptable, yours denounces as heresy, and war follows soon after. Please, I already have to work with _machines_ that insist on splitting all hairs in sight, if the _humans_ are going to join in, I think I'm going to scream, look for a nice tall building and try out Arthur Dent's theory on flight :-) Also, consider this: if we hadn't started this dicussion, think of all the fun we could have had in that bandwidth by pointing at IBM segment registers and snickering :-) In other words, considering that the machine-specific posts do not by any reasonable strecth of any healthy imagination constitute a flood (or even wet feet :-), isn't the cure causing more problems than the supposed disease? (And don't tell me it _shouldn't_ have. I don't give a hoot about shoulds. I have seen several people trying to start this campaign several times now, always with the same result: we have a flame war for a while, and then things go back to the old ways. It may not be pretty, but them's the facts) SR "Lessee, watercooled underwear, check. Insulating overwear, check. Flame-resistant coveral, mask and hood, check. Independent airsupply, check. Shields up, check. Deflector screens on, check. /dev/flame_drain mounted, check. Dummy activated, check. Comfy hiding place prepared, check. OK, I think I can send this now."
wfp@dasys1.UUCP (William Phillips) (11/18/89)
I was merely disagreeing with the original posting in an ironic manner. It is often unclear (especially to a novice ;-) what the exact nature of a problem is. Besides, it can be valuable to have some insights into the limitations of various implementations, whether they are dictated by the hardware/os or not. That's all. Sorry I opened my file. bill -- William Phillips System Co-administrator BEC Public Excess Unix New York wfp@dasys1.UUCP ..{cmcl2,hombre,cucard}!dasys1!wfp Sorry, witty quip temporarily out of service.