Path: utzoo!attcan!uunet!munnari.oz.au!bruce!mmcg
From: mmcg@bruce.cc.monash.oz (Mike McGaughey)
Newsgroups: comp.theory
Subject: Re: Homework problems?
Message-ID: <1923@bruce.cc.monash.oz>
Date: 15 Mar 90 13:11:03 GMT
References: <1420@oravax.UUCP>
Organization: Monash Uni. Computer Centre, Australia
Lines: 37

harper@oravax.UUCP (Douglas Harper) [15 Mar 90 03:24:39 GMT]:
> What about coupling this with crying "foul" if you see that a homework
> problem has been posted?  What about sending mail to the individual's
> postmaster asking that the CS (or whatever) department be notified that
> such and such a problem was posted by that person?

Yes, but what if it isn't homework?

ex: tonight I became interested in fitting an exponential curve to
some points.  I am a 3rd yr CS PhD, so my maths is a little rusty :-).
After some fiddling, I had an equation to solve :

	0 = a x^n1 + bx^n2 + cx^n3

where the only unknown is x, and n1-n3 are large numbers.  This isn't
analytically soluble, but there may be some upper bound to the number
of solutions (so I could write a search program).  As it is, I
concluded that there is, as there are only 3 terms and each term can
only change sign once - hence max of 4 solns.  If it had been a little
later at night (if I was a little more stupid :-), I probably would have
asked - "obviously a homework problem".

I think it's better to send a little note to the person, so they
get a chance to explain, have a monthly posting warning
about this, and start quietly ignoring those who post too many
homework questions.

Cheers,

	Mike.

BTW: the maths was prompted by a table of populations in sci.sceptic.
--
Mike McGaughey			ACSNET:	mmcg@bruce.cs.monash.oz

"Ignorance is stupid, but education causes college students"
	- P.J. O'Rourke, Republican Party Reptile.