[9fans] could it write code too?

[9fans] could it write code too?

[Ronald G. Minnich]

http://www.pdos.lcs.mit.edu/scigen/

good if you're hunting tenure.

ron

Re: [9fans] could it write code too?

[Kenji Okamoto]

> http://www.pdos.lcs.mit.edu/scigen/

I have problem in English, so I'd like to make confirm it.
IS IT A JOKE? ☺

Here, we have some such money service to translate Japanese
to English, I mean science paper!   It seems there is some market
place for such needs which I can't believe.
How the original author can make it confirm that the translation is
comfortable or getting a good level of translation!   Unfortunately,
it's not a joke...

Kenji

Re: [9fans] could it write code too?

[Russ Cox]

> > http://www.pdos.lcs.mit.edu/scigen/
> 
> I have problem in English, so I'd like to make confirm it.
> IS IT A JOKE? ☺

It's not a joke.  WMSCI is a money-making conference that
spams academics non-stop begging for submissions.  
After seeing David Mazieres's submission
(http://www.scs.cs.nyu.edu/~dm/remove.pdf),
the students involved (some of my officemates)
spent about a week, off and on, building a random grammar
for generating research papers and submitted two 
randomly-generated papers.  One got in, one did not.

It may not be immediately obvious to a non-native
speaker, but the papers are just gibberish.  They don't
make sense from one sentence to the next.  Even the most
cursory reading of a paragraph or two should have been
enough to figure out that the papers were fakes.  And yet
one still got in.

The hope is that enough egg will end up on WMSCI's face
that they'll stop spamming us.  Then again, I doubt it.
There was an interesting post on Slashdot from someone
who said that he had first heard of iiisci (the parent organization)
when he saw a phishing email that used them as a link, and
when he emailed them to warn them that they were being used
like that (thinking their site had been taken over by hackers),
he started getting the conference spams.

Russ

Re: [9fans] could it write code too?

[Andy Newman]

Russ Cox wrote:
> After seeing David Mazieres's submission
> (http://www.scs.cs.nyu.edu/~dm/remove.pdf),

That paper is a work of beauty.  The graph and
scatter plot are very nice touches.

Re: [9fans] could it write code too?

[Kenji Okamoto]

> That paper is a work of beauty.  The graph and
> scatter plot are very nice touches.

However, no reference point for referenced papers [1] and [2].

Kenji

Re: [9fans] could it write code too?

[Wes Kussmaul]

[-- Attachment #1: Type: text/html, Size: 846 bytes --]

Re: [9fans] could it write code too?

[Ronald G. Minnich]

On Fri, 15 Apr 2005, Kenji Okamoto wrote:

> > http://www.pdos.lcs.mit.edu/scigen/
> 
> I have problem in English, so I'd like to make confirm it.
> IS IT A JOKE? ☺

well, it is kind of a joke. The program wrote the paper. The paper was 
accepted. 

at the same time, I've read worse papers than that one that were not 
created by a computer.

ron

Re: [9fans] could it write code too?

[Ronald G. Minnich]

is the old paper from pixar about rendering jello online anywhere? Another 
classic ...

ron

[9fans] Pixar and Jell-O

[Karl Sackett]

On 2005 April 14 at 20:45:24 -0600, Ronald G. Minnich wrote:
> is the old paper from pixar about rendering jello online anywhere? Another 
> classic ...

Google is your friend.
'Ray Tracing Jell-O Brand Gelatin' by Paul S. Heckbert
  <http://www.netfunny.com/rhf/jokes/88q3/20424.10.html>

-- 
Karl Sackett                     K4KRS                    krs@hiwaay.net

   "When fascism comes to America, it will be wrapped in the flag and
    carrying the cross."                               Sinclair Lewis

Re: [9fans] could it write code too?

[geoff]

Did festoon ever get distributed outside the labs?  Maybe it was a
late Research Unix thing only?  It did something similar: generated
plausible-looking papers of random gibberish, complete with tables,
graphs and equations.

Google suggests that a binary and man page escaped into 2.9BSD.

A sample:

.TL
An Analogously Uniform Centralized Organization
.AU "C. C. Festoon" CCF Headquarters 1113538458
.AS
Usually,
there is not a major objective which may have been apomisoescing the apparentnesses.
Assumptions were insufficiently federesced by that investigation.
The interaction neoenniates a casually intelligent requirements definition.
Anytime that a not unactivating correspondence veriates at a process,
an invention had dispolyesced the substantial change.
In the same way as described above,
those criterias had ultimesced a system philosophy.
[...]

Re: [9fans] Pixar and Jell-O

[Paul Lalonde]

Why, that's the paper that taught me about Jabobians!
A joke, but one that works!

Paul

On 14-Apr-05, at 8:57 PM, Karl Sackett wrote:

> On 2005 April 14 at 20:45:24 -0600, Ronald G. Minnich wrote:
>> is the old paper from pixar about rendering jello online anywhere? 
>> Another
>> classic ...
>
> Google is your friend.
> 'Ray Tracing Jell-O Brand Gelatin' by Paul S. Heckbert
>   <http://www.netfunny.com/rhf/jokes/88q3/20424.10.html>
>
> -- 
> Karl Sackett                     K4KRS                    
> krs@hiwaay.net
>
>    "When fascism comes to America, it will be wrapped in the flag and
>     carrying the cross."                               Sinclair Lewis
>

Re: [9fans] could it write code too?

[Russ Cox]

Festoon was in 2e Plan 9.  I think the papers they're
generating are more readable than festoon's.  The
general sound of everything is more plausible.  For example,
the conclusion that was accepted:

              IV.  CONCLUSION

  Here we motivated Rooter, an analysis of rasterization.
  We leave out a more thorough discussion due to resource
  constraints.  Along these same lines, the characteristics of
  our heuristic, in relation to those of more little-known 
  algorithms, are clearly more unfortunate.  Next, our algorithm
  has set a precedent for Markov models, and we expect that
  theorists will harness Rooter for years to come.  Clearly, our
  vision for the future of programming languages certainly 
  includes our algorithm.

Russ

Re: [9fans] Pixar and Jell-O

[Russ Cox]

> 'Ray Tracing Jell-O Brand Gelatin' by Paul S. Heckbert
>   <http://www.netfunny.com/rhf/jokes/88q3/20424.10.html>

A Slashdot post mentioned a hand-written fake paper 
titled "Ray Tracing in the Absence of Light".

Russ

Re: [9fans] could it write code too?

[Steve Simon]

http://www.tribug.org/pub/tuhs/Applications/Festoon/

-Steve

Re: [9fans] could it write code too?

[Wes Kussmaul]

>>That paper is a work of beauty.  The graph and
>>scatter plot are very nice touches. 
>>
> I recall a similar paper written by one Jack Torrance in...

Jack Torrance was played by another Jack... caretaker for a resort hotel...

Re: [9fans] could it write code too?

[geoff]

Thanks for the pointer to TUHS.

Yeah, festoon is a little too keen to invent words and isn't very good
at it.  One can turn down the frequency of invented words but I had no
luck at preventing them completely.

Re: [9fans] could it write code too?

[boyd, rounin]

festoon(6) was in 9th Ed.

for the old-style (on another topic) this _might_ be interesting:

    http://minnie.tuhs.org/UnixTree/Ausam/sys/defines.elec.html
--
MGRS 31U DQ 52572 12604