Path: utzoo!censor!becker!eric
From: eric@becker.UUCP (Eric Siegerman)
Newsgroups: can.uucp
Subject: Re: new uucp maps are out
Summary: Humans still imperfect
Message-ID: <423@becker.UUCP>
Date: 18 Apr 89 03:48:32 GMT
References: <89Apr10.101145edt.38137@neat.ai.toronto.edu>
Reply-To: eks@kneller.UUCP (Eric Siegerman)
Distribution: can
Organization: There ain't no organization here.
Lines: 36

In article <89Apr10.101145edt.38137@neat.ai.toronto.edu> rayan@ai.toronto.edu (Rayan Zachariassen) writes:
>I edited the submissions for the following hosts:
>[list of hosts, plus other stuff, deleted]
>
>Would you be interested in a scheme that allows you to update your
>map entry via a server of some kind? If so, how?

It makes no difference to me whether Rayan or a program collates
kneller's map entry into the lists.  Seems to me that Rayan's
time/energy is the issue...

>My major concern with
>that is that I've got no syntax-checking program to check entries
>automatically.  Are people disciplined enough that it isn't necessary?

All but the ones you named :-).  Seriously, there will always be
errors in submitted map entries.  I suspect it's not as much a
matter of discipline as of the simple fact that people make
mistakes.  Writing a map entry is as error-prone as writing a
program -- less intrinsically difficult, but most of us spend a
lot more of our time writing code than we do writing map entries,
which goes a long way toward cancelling out the fact that it's an
intrinsically easier task.  (With the proliferation of personal
Usenet nodes, an increasing number of submitted map entries will
be their authors' first attempt!)

I'm not crying doom here, or arguing that Rayan (or anyone)
*should* continue to manually process map entries.  None of this
argues against a syntax-checking program.  My point is simply
that utai!path will always receive a few erroneous submissions,
and that any scheme which assumes the contrary is sure to
encounter occasional problems.

-- 
Eric Siegerman, Toronto, Ont.
eks@kneller.UUCP,  eric@becker.UUCP,  ...!utzoo!mnetor!becker!kneller!eks