Re: Ads, FAQ's, and Bandwidth.

From: sjoerd@cs.vu.nl (Sjoerd Schreuder)

Date: Tue, 29 Jul 1997 00:00:00 +0000

First of all: Please don't use any characters above 127, and please
insert a CR around the 70th character on each line. That makes your
posts better readable for people who use a simple ascii screen.
One line looked like: "( n-f\173r2) and \027tr" on my screen.
I can't even post this message without deleting it...

kmgraphiks@aol.com (KMGraphiks) writes:

>Greg Lindahl writes:

>>kmgraphiks@aol.com (KMGraphiks) writes:
>>
>>> You're a master at the art of doubletalk then. In our discussion about
>>> whether moderators should promote here you say the games are promoted on
>>> your list better because there's no hype, hype is not the best way to
>>> promote things you say. This infers that you think moderator promotions
>>> here will be hype which isn't necessarily so.
>>
>>No, it does not infer that. Some moderator postings are hype. Some are
>>not.

[partially snipped]

>The Facts:
>A. Our subject is whether ads on the ng are a good thing or not
>B. I say you think Usenet tradition is more important than the PBM industry.
>C. YOU introduce the topic of HYPE which you say is bad, information is good.
>D. You say the PBM List provides information, not HYPE.
>E. You say the PBM list is more important to selling the industry than
>   ads in the ng.

I agree with facts C-E; B doesn't matter here, while A gives us the domain
we're talking about.

>Thus using the definitions 2, 3 & 4 above a reasonable person may
>"surmise" or "conclude" that you are "hinting" or "implying" that ads
>in the ng are bad because they will be full of HYPE (which is bad) and
>not INFORMATION (like your PBM List).

Wrong: one may not conclude that. As Greg says: "some moderator postings
are hype. Some are not"  Compare this with: "the PBM List provides
information, not HYPE". As we can prove that "if it's an ad, then it's
good" doesn't hold, it's only a small step to prefer the PBM List
above ads _IN THIS RESPECT ONLY_ (I don't say I prefer the PBM List above
ads in all respects; I don't say I don't prefer the PBM List above
ads in all respects either)

If we agree on the following:
A. If it's on the PBM List, then it's information
B. If it's on the PBM List, then it's not hype
C. There exists a topic that's both in an ad and info
D. There exists a topic that's both in an ad and hype
E. If it's info, then it's good
F. If it's hype, then it's bad.
K. if it's not good, then it's bad
L. if it's good, then it's not bad

Then the following has to be false:
N. if it's in an ad, then it's good

If you think N is true, one of A-F, K-L has to false. Please indicate which
one in that case.

For those who know some logic: (and for who doesn't: I've tried to
make it as clear as possible)

==== START OF LOGIC ====

Notation:
FA = for all (usage: "FA x: form" == every x will make "form" true)
EX = there exists (usage: "EX x: form" == some x will make "form" true)
~x = negation (usage: "~form" == not "form")
/\ = and/conjunction (usage: "a /\ b" == both "a" and "b")
\/ = or/disjunction (usage: "a \/ b" == "a" or "b" or both)
-> = implication (usage: "a -> b" == if "a" then "b")

Definitions:
L(x): x is on the PBM List
A(x): x is an ad in the ng
G(x): x is good
B(x): x is bad
H(x): x is hype
I(x): x is information

Given:
A. FA x: L(x) -> I(x)    "if it's on the PBM List, then it's information"
B. FA x: L(x) -> ~H(x)   "if it's on the PBM List, then it's not hype"
C. EX x: A(x) /\ I(x)    "there exists a topic that's both in an ad and info"
D. EX x: A(x) /\ H(x)    "there exists a topic that's both in an ad and hype"
E. FA x: I(x) -> G(x)    "if it's info, then it's good"
F. FA x: H(x) -> B(x)    "if it's hype, then it's bad"

I've added C & D myself, I hope everyone agrees with them.

Derived from A-F:
G. FA x: L(x) -> G(x)    "if it's on the PBM List, then it's good"
H. FA x: L(x) -> ~B(x)   "if it's on the PBM List, then it's not bad"
I. EX x: A(x) /\ G(x)    "there exist a topic that's both in an ad and good"
J. EX x: A(x) /\ B(x)    "there exist a topic that's both in an ad and bad"

If we say that "not good is bad, and not bad is good", we can add:
K. FA x: ~G(x) -> B(x)   "if it's not good, it's bad"
L. FA x: G(x) -> ~B(x)   "if it's good, it's not bad"

If we want to see if we can prove "if it's in an ad, it's bad", that's
pretty hard. We might easily prove that we can prove it: if we add
the reverse and are able to derive something that has to be false,
we're sure we can prove the original statement.
(false is usually something like: G(x) /\ ~G(x) ==
                              "x is both good and not good")

So let's try:
M. ~ ( FA x: A(x) -> B(x) )

I can't derive anything false from A-M, can you? My software can't either.
So "if it's in an ad, it's bad" can't be inferred from A-F, K-L.

Something I can prove wrong is:
N. FA x: A(x) -> G(x)    "if it's in an ad, then it's good"

If I use N and F in D, and then use L in the results, I get:
O. EX x: ~B(x) /\ B(x).  "there exists a topic that's both not bad and bad"
In other words: One of the assumptions (A-F, K-L, N) is wrong. If you
agree with A-f and K-L, you have to agree that N is wrong.

==== END OF LOGIC ====

>				You deny the possibility of this
>inference and accuse me of putting words in your mouth (I couldn't....it's
>full of foot).

I indeed deny the possibility of this inference: I've made explicit
all my assumptions, please tell me where I'm wrong.

>			I guess I should be happy that you do acknowledge that
>some moderator postings are not hype. Maybe if you were a little more
>careful about what opinions you linked together there'd be less chance
>of INFERring something you don't mean and won't support.

No need to react to this flame.

Sjoerd Schreuder

PS: For those who know something about logic: I'm assuming classical logic
and first order predicates. I've omitted several details to keep it
understandable. Please correct me if I've made a mistake.

References

Re: Ads, FAQ's, and Bandwidth.
kmgraphiks@aol.com (KMGraphiks)
Mon, 28 Jul 1997 00:00:00 +0000
Re: Ads, FAQ's, and Bandwidth.
kmgraphiks@aol.com (KMGraphiks)
Mon, 28 Jul 1997 00:00:00 +0000

Referenced By

Re: Ads, FAQ's, and Bandwidth.
kmgraphiks@aol.com (KMGraphiks)
Mon, 28 Jul 1997 00:00:00 +0000
Re: Ads, FAQ's, and Bandwidth.
kmgraphiks@aol.com (KMGraphiks)
Mon, 28 Jul 1997 00:00:00 +0000