[comp.lang.c] apleaforrestraint

jss@hector.UUCP (Jerry Schwarz) (05/26/89)

Please, please, please double check your answers to questions.
Recently there have been a rash of incidents where a beginner posts a
question and answer is posted that appears to have been from another
beginner and is completely wrong.   As a consequence, instead of
getting a couple of answers to the original question we see a flood
of followups to the wrong answer. This is understandable, since many
people feel it is more important to correct a wrong assertion in this
group than to answer a question.

I am not accusing anybody of behaving badly, I am sure that all
posters have good motivations.  But I urge you to either verify your
answers by running a test, or looking up a reference. Rely on your
"understanding" only if you are really a world class C expert, and
even then only if you are not posting under conditions of sleep
deprivation.

Think of posting to comp.lang.c as getting up in the front of a
filled 5,000 seat auditorium.  Doesn't that make you want to
spend a little time preparing? 

Jerry Schwarz
AT&T Bell Labs, Murray Hill

tneff@bfmny0.UUCP (Tom Neff) (05/27/89)

In article <11601@ulysses.homer.nj.att.com> jss@hector.UUCP (Jerry Schwarz) writes:
>Please, please, please double check your answers to questions.
>Recently there have been a rash of incidents where a beginner posts a
>question and answer is posted that appears to have been from another
>beginner and is completely wrong....

Let me second Jerry's comments (most of which I have deleted, but they 
were mild mannered) and suggest a few guidelines for posting to groups 
like this which tout le monde subscribe to. 

 * READ A C BOOK before posting questions here! Anyone with Usenet 
access and an interest in C programming should study Kernighan & 
Ritchie and/or Harbison & Steele thoroughly before posting anything to 
the newsgroup (and costing the net hundreds or thousands of dollars). 
Many (not all) of the beginners' questions posted here betray deep 
ignorance of the language and could have been answered more quickly 
and cheaply by cracking a book or leaning one's head into the office 
next door. Nobody is perfect, but there are reasonable measures one 
can take. 

 * WHEN DUMB QUESTIONS ARE POSTED ANYWAY, don't answer them via 
Followup articles in the newsgroup! This just generates more noise and 
expense. *MAIL*, repeat *MAIL* your reply to the person who posted the 
question. If the answer is of potential use to others, suggest in your 
mailed reply that the poster *summarize* responses to Netnews after 
they are collected and he has learned his lesson. NOTE: I am aware 
that mail is not always 100% reliable from Netnews, but when it comes 
to answering questions about how many elements "char a[33];" has, YOUR 
reply may not be the crucially vital one that must get through at all 
costs. :-) 

 * WHEN INTERESTING QUESTIONS ARE POSTED, which *will* happen all odds 
to the contrary :-), READ ALL FOLLOWUPS before posting your own! 
Someone may have completely addressed the issue before you. When 
something looks particularly juicy it's best to wait a day and THEN 
followup if no one else has covered it. Some wag will always point out 
that if EVERYONE waited a day we'd be in the same boat, etc etc, but 
in practice that's not how it works. Doug or Chris will nail the 
tough-but-interesting queries with 95% accuracy and your followup may 
only be needed to smooth out the edges. 

 * WHEN YOU POST C CODE as an example or counterexample or whatever, 
COMPILE AND RUN IT to make sure it works as-is!! Nothing is more 
frustrating (or more likely to generate kilobytes of repetitive 
scolding followups) than a typo-filled C fragment someone "winged" on 
the fly in the news editor without actually testing. 

 * MAKE A LIST OF GURU NAMES and netmail addresses as you read the 
newsgroup. If you have a C question, you can *MAIL* it to three or 
four people and be 95% certain of getting a good reply WITHOUT costing 
Usenet kilobucks. If you use this method you don't even have to worry 
whether it's a DUMB question or not. 

-- 

Remember, Netnews is a precious resource which costs people bucks. The 
above guidelines are gentle suggestions on how to conserve and 
optimize the resource. Happy coding! 

-- 
Tom Neff				UUCP:     ...!uunet!bfmny0!tneff
    "Truisms aren't everything."	Internet: tneff@bfmny0.UU.NET

daveh@marob.masa.com (Dave Hammond) (05/28/89)

In article <14354@bfmny0.UUCP> tneff@bfmny0.UUCP (Tom Neff) writes:

> * MAKE A LIST OF GURU NAMES and netmail addresses as you read the 
>newsgroup. If you have a C question, you can *MAIL* it to three or 
>four people and be 95% certain of getting a good reply WITHOUT costing 
>Usenet kilobucks. If you use this method you don't even have to worry 
>whether it's a DUMB question or not. 

This is, IMHO, a minor flaw in an otherwise stellar set of posting guidelines.

I have always felt that unsolicited C-related mail directly to Chris,
or Doug, or Henry was side-stepping the net, and, in a small way perhaps,
an invasion of their privacy.

If these folks were overloaded with personal e-mail regarding C, they might
feel less inclined to council as thoroughly in c.l.c.

Comments?

--
Dave Hammond
daveh@marob.masa.com

gwyn@smoke.BRL.MIL (Doug Gwyn) (05/30/89)

In article <712@marob.masa.com> daveh@marob.masa.com (Dave Hammond) writes:
>I have always felt that unsolicited C-related mail directly to Chris,
>or Doug, or Henry was side-stepping the net, and, in a small way perhaps,
>an invasion of their privacy.
>If these folks were overloaded with personal e-mail regarding C, they might
>feel less inclined to council as thoroughly in c.l.c.

I can't speak for the others, but the main reason I post responses to C
questions is to spread information to as many C programmers as possible.
Obviously, private e-mail does not accomplish that.  Since my employer
doesn't think I'm being paid to help programmers at large, reducing the
leverage of the effort I put into these notes is not a viable option; I
can't justify answering a flood of novice questions via private e-mail.

I don't know what can be done about the low S/N ratio in this newsgroup.
It's one of the few Usenet newgroups I still read, as others long ago
became intolerably noisy.  Changing the newgroup to be moderated might
help, at the cost of increased delays between posting and reading.

tneff@bfmny0.UUCP (Tom Neff) (05/30/89)

Through several mailed replies and Doug's posting I have come to
realize that the people who answer C questions in this newsgroup
probably don't want them arriving in the mail instead.  So I withdraw
the part of my suggested guidelines that said "make a list of gurus
and mail them your question first," even though this isn't too far
from what Spaf suggests in n.a.nu [-nanu?].  I apologize for including
anything other than the patently unarguable in that posting, cause it
was intended to be legitimately useful for newbies.  I may post it again
in a few weeks minus the offending clause.
-- 
Tom Neff				UUCP:     ...!uunet!bfmny0!tneff
    "Truisms aren't everything."	Internet: tneff@bfmny0.UU.NET

bobmon@iuvax.cs.indiana.edu (RAMontante) (05/31/89)

tneff@bfmny0.UUCP (Tom Neff) <14363@bfmny0.UUCP> :
-Through several mailed replies and Doug's posting I have come to
-realize that the people who answer C questions in this newsgroup
-probably don't want them arriving in the mail instead.
-		[ ... ]
-was intended to be legitimately useful for newbies.

I might suggest that the questions be mailed to these gurus, but their
answers be posted.  This would disseminate the information without
encouraging the wrong answers.  It might still not please the gurus.

In any case, though, I think the biggest problem is simply that those
most in need of some guidelines are those least likely to be aware of
them, capable of applying them, or inclined to try to follow them.  It's
the old "I'm brand new to the net, so I'll make a name for myself by
proposing that NULL be redefined as (char *)0" syndrome.

How about this for a posting policy:
    Anybody can post, but only approved gurus are allowed to use
    whitespace, capitals, numerals, and the letter `c'.

diamond@diamond.csl.sony.junet (Norman Diamond) (06/01/89)

inartikle<twoonethreenineeight@iuvax.sees.indiana.edu>bobmon@iuvax.sees.indiana.edu(ramontante)writes:
>How about this for a posting policy:
>    Anybody can post, but only approved gurus are allowed to use
>    whitespace, capitals, numerals, and the letter `c'.
iagree,butwhoapprovedyouasaguru?
--
normandiamond,sonykomputerssienselab(diamond%seesl.sony.ko.jp@relay.sees.net)
theaboveopinionsaremyown.|whyareprogrammerskritisizedfor
ifthey'realsoyouropinions,|re-implementingthewheel,whenkar
you'reinfringingmykopyright.|manufakturersarepraisedforit?