About trolls (the current post-election wave of madness)

[Andrew Hamm]

[damn, sorry for the repeat posting Icon group, I forgot to include all
our cross-posted friends - I hope this can be a friendly exchange of
programming fun. Please reply to the x-posted if you all feel cool about
meeting the neighbours]

That posting and so many others are indeed trolling. I've seen it on
several very disparate groups that I visit. I've started a new thread so I
don't "reward" the troll as Nils suggested - that's an interesting
statement about them taking scores.

Anyway, I'm intrigued by this:

>You are only saying this because you never met the shortest
>python expression passing the Türing test :
>
>filter(lambda W : W not in 'ILLITERATE','BULLSHIT')
>
>Is there any equivalent in hermes, icon, or haskell ?

If you could explain it, we can try. If my guess of the meaning is
correct, then I think Perl could achieve a one-liner. But Icon doesn't
have much in the way of Lambda calculus - unless we can exploit
coroutines. Frank will probably have thoughts about that...

[not [quite] more i squared]

Andrew Hamm wrote:

> Anyway, I'm intrigued by this:
>
>
>>You are only saying this because you never met the shortest
>>python expression passing the Türing test :
>>
>>filter(lambda W : W not in 'ILLITERATE','BULLSHIT')
>>
>>Is there any equivalent in hermes, icon, or haskell ?
>
>
> If you could explain it,

The operational explanation is that it evaluates to : 'BUSH'
by filtering away from 'BULLSHIT', characters found in 'ILLITERATE'

But let's leave these technical details to the experts,
it passes the Türing test by perfectly modelling the overt
behavior of many patriotic Americans !

[Andrew Hamm]

not [quite] more i squared wrote:
>
> The operational explanation is that it evaluates to : 'BUSH'
> by filtering away from 'BULLSHIT', characters found in 'ILLITERATE'
>
> But let's leave these technical details to the experts,
> it passes the Türing test by perfectly modelling the overt
> behavior of many patriotic Americans !

heh heh not only a computer pun, but one which uses the Turing test. A
class act. Well done.

[not [quite] more i squared]

Well, thanks.

But, I fear you shouldn't underestimate the talent of your current president.
While my feeling is that he is in the service of the forces of destruction, I
see time and again that he has an eg "supernatural" relationship to
(scientific american) English.

For instance, a way to approach the President's religion, would be to see as a
map of his proximity to Jesus, the relationship of the W boson to the photon -
well known to current physics.

In a sense, that's more or less what you need if you want a line respectful of
laicity that you can nevertheless oppose to fundamentalist christian Bush adorers.

"Bush and Jesus are alike the way of the W boson and the photon".

(btw please note the huge difference between scientific theology and
theological science)

[Andrew Hamm]

not [quite] more i squared wrote:
>
> But, I fear you shouldn't underestimate the talent of your current
> president.

Not my President! I'm from Australia. Still, my Prime Minister takes every
opportunity to kiss the presidents, so maybe he *is* my president whether
I like it or not...

> For instance, a way to approach the President's religion, would be to
> see as a map of his proximity to Jesus, the relationship of the W
> boson to the photon - well known to current physics.

Ok - you've definitely lost me there. I used to enjoy particle physics as
a schoolkid, but it's lost to me now.

Codd bless Google: "[W boson] are carrier particles that mediate the weak
noocular (am.sp) force, much like the photon is the carrier particle for
the electromagnetic force." I'm with you now. But don't forget, God Wanted
George To Be In Charge At This Time. He had an epithany y'see.

> In a sense, that's more or less what you need if you want a line
> respectful of laicity that you can nevertheless oppose to
> fundamentalist christian Bush adorers.
>
> "Bush and Jesus are alike the way of the W boson and the photon".
>
> (btw please note the huge difference between scientific theology and
> theological science)

Actually, I think Icon does have a shorter Turing-complete expression:

write('bullshit' -- 'illiterate')

which yields

'bhsu'

Note the extraordinary similarity to the clarity of Bush's usual
utterances.

[not [quite] more i squared]

Andrew Hamm wrote:
>
> Actually, I think Icon does have a shorter Turing-complete expression:
>
> write('bullshit' -- 'illiterate')

the syntax is a joy.

>
> which yields
>
> 'bhsu'
>
> Note the extraordinary similarity to the clarity of Bush's usual
> utterances.

Now that's interesting, why is the result garbled, how are its letters
collected ? Maybe we can supplement Carl Rove (sp ?) to help W deliver a clear
speech in Icon :)

[not [quite] more i squared]

>> Actually, I think Icon does have a shorter Turing-complete expression:
>>
>> write('bullshit' -- 'illiterate')
>>

>> which yields
>>
>> 'bhsu'

>
> Now that's interesting, why is the result garbled, how are its letters
> collected ?

Of course, I see it now, the end result is alphabetically sorted. What
facilities does Icon offer to act on the collating sequence ?

Another track to initiate us to Icon, would be the implementation of the most
specialized "alphabetic substraction" operator that's consistent with the example.

The matter is, the letters of "bush" occur in "bullshit" *in the right order*,
and the remaining letters, "llit", also occur in order in "bullshit", while
they occur *in the right order* and *contiguously* in "illiterate".

That's a lot of constraints simultaneously obeyed, to the point that not all
of them are exploited in the original python expression. My question : doesn't
Icon-the-language manifest a preference by allowing a particularly economic,
elegant or natural algorithm expression that will compute "BUSH" given the
inputs "BULLSHIT" and "ILLITERATE" (while freely taking profit of the
above-mentionned constraints) ?

Cheers, Morris Carré
--
666 ?? -- 666 ~ .666 ~ 2/3 ~ 1 - 1/3 ~ tertium non datur ~ the excluded middle
~ << either you are with US, or you are against US - W >>

[not [quite] more i squared]

Andrew Hamm wrote:
> Actually, I think Icon does have a shorter Turing-complete expression:
>
> write('bullshit' -- 'illiterate')
>
> which yields
>
> 'bhsu'
>
> Note the extraordinary similarity to the clarity of Bush's usual
> utterances.
>
>

This neatly relates to the way "ie-rate" results from substracting to
"illiterate", the letters of "bullshit" wherever they first occur
(a conjugate of the initial operation).

To be sure, aren't the -rate- and -quality- of usage of "ie" (as sentence
leader) directly correlated to the matter of clarity ?

Hum, where does Kerry fit on that scale ?

[Steve Wampler]

On Mon, 15 Nov 2004 12:51:24 +0100, not [quite] more i squared wrote:

>>> Actually, I think Icon does have a shorter Turing-complete
>>> expression:
>>>
>>> write('bullshit' -- 'illiterate')
>>>
>>> which yields
>>>
>>> 'bhsu'
>>
>> Now that's interesting, why is the result garbled, how are its
>> letters collected ?
>
> Of course, I see it now, the end result is alphabetically sorted.
> What facilities does Icon offer to act on the collating sequence ?

...
The sorting is an 'artifact' of the implementation of csets.
Csets are, by definition, unordered sets of characters, but
the conversion of a cset back to a string (for printing, or
for use in operations that expect a string) forces the order
into collating sequence, since you have to have *some* order
to the characters in a string.

Incidently, the orginal ordering of characters was lost the
moment each sequence was represented as a cset (i.e. using
the single quotes as in 'bullshit' instead of the double
quotes as in "bullshit"...). But since "--" is a cset operation,
you'd still lose the order in the result even it the above
had been:

write("bullshit"--"illiterate")

> That's a lot of constraints simultaneously obeyed, to the point

> that not all of them are exploited in the original python
> expression.
> My question : doesn't Icon-the-language manifest a preference by
> allowing a particularly economic, elegant or natural algorithm
> expression that will compute "BUSH" given the inputs "BULLSHIT"
> and "ILLITERATE" (while freely taking profit of the above-
> mentionned constraints) ?

I'm sure there's a cleaner way (in my dim, dark memory I think
I've seen it, perhaps in one of the Icon book editions...),
but here's one solution:

procedure elide(s1,s2)
s := ""; s2 := cset(s2)
s1 ? while s ||:= 1(tab(upto(s2)),tab(many(s2)))
return s
end

Strictly speaking, the s2 := cset(2) isn't needed, but is
an efficiency improvement.