Aho Corasick source code needed, please help

[Dan Pop]

Re: [OT] Re: Aho Corasick source code needed, please help

In <c0b1um$jue$1@c hessie.cirr.com > Christopher Benson-Manica <ataru@nospam.c yberspace.org> writes:
[color=blue]
>Dan Pop <Dan.Pop@cern.c h> spoke thus:
>[color=green][color=darkred]
>>>Hence, the secret formula of Coca Cola.[/color][/color]
>[color=green]
>> How could it be secret? And whatever patent Coca Cola might have had on
>> it, it has expired long ago...[/color]
>
>I believe it's classified as a "trade secret" or something to that
>effect, which is treated differently than a patent or copyright.[/color]

A trade secret is a secret as long as it can be kept secret. Even if a
chemical analysis of Coca Cola would be illegal, if someone started
producing a beverage with the very same formula, you'd have to prove that
they didn't reinvent it from scratch.

Not to mention that, if a chemical analysis is illegal, you can't
legally prove that the two beverages have the same formula ;-)

The modern patent laws have been invented with the very purpose of
removing the need of trade secrets: you publish the formula of your
beverage and, for a certain period of time, no one else is allowed to use
it. Since it is published, the competitors cannot claim that they have
reinvented it from scratch, even if they actually did!

So, if you have a bright idea, be sure to patent it before someone else
has the same idea and wins the time-to-patent race ;-)

Dan
--
Dan Pop
DESY Zeuthen, RZ group
Email: Dan.Pop@ifh.de

[Alan Balmer]

Re: [OT] Re: Aho Corasick source code needed, please help

On 10 Feb 2004 19:32:07 GMT, Dan.Pop@cern.ch (Dan Pop) wrote:
[color=blue]
>In <c0b1um$jue$1@c hessie.cirr.com > Christopher Benson-Manica <ataru@nospam.c yberspace.org> writes:
>[color=green]
>>Dan Pop <Dan.Pop@cern.c h> spoke thus:
>>[color=darkred]
>>>>Hence, the secret formula of Coca Cola.[/color]
>>[color=darkred]
>>> How could it be secret? And whatever patent Coca Cola might have had on
>>> it, it has expired long ago...[/color]
>>
>>I believe it's classified as a "trade secret" or something to that
>>effect, which is treated differently than a patent or copyright.[/color]
>
>A trade secret is a secret as long as it can be kept secret. Even if a
>chemical analysis of Coca Cola would be illegal, if someone started
>producing a beverage with the very same formula, you'd have to prove that
>they didn't reinvent it from scratch.[/color]

A chemical analysis won't tell you how to duplicate it.[color=blue]
>
>Not to mention that, if a chemical analysis is illegal, you can't
>legally prove that the two beverages have the same formula ;-)[/color]

There are a number of ways of legally proving such a thing. In fact,
proving the same chemical analysis would not prove that products are
identical, either.[color=blue]
>
>The modern patent laws have been invented with the very purpose of
>removing the need of trade secrets: you publish the formula of your
>beverage and, for a certain period of time, no one else is allowed to use
>it. Since it is published, the competitors cannot claim that they have
>reinvented it from scratch, even if they actually did![/color]

Patent law encourages inventors to publish their inventions. It does
not require them to do so. Inventors have a choice as to whether they
want to rely on patent protection for a defined length of time, or
trade secret status for an undefined length of time.[color=blue]
>
>So, if you have a bright idea, be sure to patent it before someone else
>has the same idea and wins the time-to-patent race ;-)
>
>Dan[/color]

--
Al Balmer
Balmer Consulting
removebalmercon sultingthis@att .net

[Ben Pfaff]

Re: [OT] Re: Aho Corasick source code needed, please help

Alan Balmer <albalmer@att.n et> writes:
[color=blue]
> Inventors have a choice as to whether they want to rely on
> patent protection for a defined length of time, or trade secret
> status for an undefined length of time.[/color]

IANAL but I was unaware that trade secrets caused nasal demons.

[Dan Pop]

Re: [OT] Re: Aho Corasick source code needed, please help

In <qtqi20d8396ui5 a4eib2336ksvoqf sqlh1@4ax.com> Alan Balmer <albalmer@att.n et> writes:
[color=blue]
>On 10 Feb 2004 19:32:07 GMT, Dan.Pop@cern.ch (Dan Pop) wrote:
>[color=green]
>>In <c0b1um$jue$1@c hessie.cirr.com > Christopher Benson-Manica <ataru@nospam.c yberspace.org> writes:
>>[color=darkred]
>>>Dan Pop <Dan.Pop@cern.c h> spoke thus:
>>>
>>>>>Hence, the secret formula of Coca Cola.
>>>
>>>> How could it be secret? And whatever patent Coca Cola might have had on
>>>> it, it has expired long ago...
>>>
>>>I believe it's classified as a "trade secret" or something to that
>>>effect, which is treated differently than a patent or copyright.[/color]
>>
>>A trade secret is a secret as long as it can be kept secret. Even if a
>>chemical analysis of Coca Cola would be illegal, if someone started
>>producing a beverage with the very same formula, you'd have to prove that
>>they didn't reinvent it from scratch.[/color]
>
>A chemical analysis won't tell you how to duplicate it.[/color]

Please elaborate. Chemical analysis is not the same thing as atomical
analysis, telling which chemical elements are present and in what
proportions.
[color=blue][color=green]
>>Not to mention that, if a chemical analysis is illegal, you can't
>>legally prove that the two beverages have the same formula ;-)[/color]
>
>There are a number of ways of legally proving such a thing. In fact,
>proving the same chemical analysis would not prove that products are
>identical, either.[/color]

Where would the differences come from, in the case of a coke-like drink?
[color=blue][color=green]
>>The modern patent laws have been invented with the very purpose of
>>removing the need of trade secrets: you publish the formula of your
>>beverage and, for a certain period of time, no one else is allowed to use
>>it. Since it is published, the competitors cannot claim that they have
>>reinvented it from scratch, even if they actually did![/color]
>
>Patent law encourages inventors to publish their inventions. It does
>not require them to do so. Inventors have a choice as to whether they
>want to rely on patent protection for a defined length of time, or
>trade secret status for an undefined length of time.[/color]

That's why I said "removing the *need* of trade secrets" and NOT "removing
the trade secrets". These days, the period of exclusive usage granted by
a patent is, in most cases, orders of magnitude higher than the time it
takes to figure out a trade secret.

Dan
--
Dan Pop
DESY Zeuthen, RZ group
Email: Dan.Pop@ifh.de

[Alan Balmer]

Re: [OT] Re: Aho Corasick source code needed, please help

On 11 Feb 2004 13:33:53 GMT, Dan.Pop@cern.ch (Dan Pop) wrote:
[color=blue]
>In <qtqi20d8396ui5 a4eib2336ksvoqf sqlh1@4ax.com> Alan Balmer <albalmer@att.n et> writes:
>[color=green]
>>On 10 Feb 2004 19:32:07 GMT, Dan.Pop@cern.ch (Dan Pop) wrote:
>>[color=darkred]
>>>In <c0b1um$jue$1@c hessie.cirr.com > Christopher Benson-Manica <ataru@nospam.c yberspace.org> writes:
>>>
>>>>Dan Pop <Dan.Pop@cern.c h> spoke thus:
>>>>
>>>>>>Hence, the secret formula of Coca Cola.
>>>>
>>>>> How could it be secret? And whatever patent Coca Cola might have had on
>>>>> it, it has expired long ago...
>>>>
>>>>I believe it's classified as a "trade secret" or something to that
>>>>effect, which is treated differently than a patent or copyright.
>>>
>>>A trade secret is a secret as long as it can be kept secret. Even if a
>>>chemical analysis of Coca Cola would be illegal, if someone started
>>>producing a beverage with the very same formula, you'd have to prove that
>>>they didn't reinvent it from scratch.[/color]
>>
>>A chemical analysis won't tell you how to duplicate it.[/color]
>
>Please elaborate. Chemical analysis is not the same thing as atomical
>analysis, telling which chemical elements are present and in what
>proportions.[/color]

Actually, what you call atomical analysis is a goodly part of chemical
analysis. However, chemical analysis is a broad term, covering many
other analytical techniques as well. It's also quite limited. In a
complex substance or mixture of substances, the description of the
sample obtained by chemical analysis won't usually correspond to a
formula for duplicating the sample. Sorry if you don't believe that,
but further elaboration would require a course in analytical
chemistry, which I'm not inclined or qualified to give.[color=blue]
>[color=green][color=darkred]
>>>Not to mention that, if a chemical analysis is illegal, you can't
>>>legally prove that the two beverages have the same formula ;-)[/color]
>>
>>There are a number of ways of legally proving such a thing. In fact,
>>proving the same chemical analysis would not prove that products are
>>identical, either.[/color]
>
>Where would the differences come from, in the case of a coke-like drink?[/color]

The proposition follows from the discussion above.

Actually, many laboratory analyses of Coca-Cola have been done, and
analysts have proclaimed that the formula must contain this or that
set of ingredients, but none has been able to give a formula for
duplicating the product. There was a publication of what was claimed
to be the original formula (from written notes in the Pemberton
estate, not analysis) but the formula has changed since then.

In general, we have only limited success in synthesizing natural
substances, or even identifying them with any certainty, especially if
they have undergone some type of processing. (In spite of the weekly
miracles performed by the CSI teams on TV.)
[color=blue][color=green][color=darkred]
>>>The modern patent laws have been invented with the very purpose of
>>>removing the need of trade secrets: you publish the formula of your
>>>beverage and, for a certain period of time, no one else is allowed to use
>>>it. Since it is published, the competitors cannot claim that they have
>>>reinvented it from scratch, even if they actually did![/color]
>>
>>Patent law encourages inventors to publish their inventions. It does
>>not require them to do so. Inventors have a choice as to whether they
>>want to rely on patent protection for a defined length of time, or
>>trade secret status for an undefined length of time.[/color]
>
>That's why I said "removing the *need* of trade secrets" and NOT "removing
>the trade secrets". These days, the period of exclusive usage granted by
>a patent is, in most cases, orders of magnitude higher than the time it
>takes to figure out a trade secret.
>[/color]
But not always. The inventor is gambling. Coca-Cola's formula, and the
variations on it, have been kept secret since 1886. Not that it
actually matters - even if the formula was published, it's too late to
stop that particular juggernaut.

It is part of the mystique, though. Coca-Cola corporate rules are that
two people have possession of the formula, their identities are not
disclosed, and they never travel together. In actuality, I suspect
some of the people on the syrup production floor have a pretty good
notion as to what's in it.

Kentucky Fried Chicken goes even further. Their "secret blend of 11
herbs and spices" is partly made in two different places and combined
at a third location. Again, it doesn't matter much - the brand name is
what sells the chicken.

I'm afraid this is really getting off-topic ;-) I'll stop wasting
people's time with it. Protection of inventions has sort of a nebulous
connection to programming, but fried chicken is way too far off base.


--
Al Balmer
Balmer Consulting
removebalmercon sultingthis@att .net

[Dan Pop]

Re: [OT] Re: Aho Corasick source code needed, please help

In <gsik20h18bkenu g3g6mik95nk7trp mlalv@4ax.com> Alan Balmer <albalmer@att.n et> writes:
[color=blue]
>On 11 Feb 2004 13:33:53 GMT, Dan.Pop@cern.ch (Dan Pop) wrote:
>[color=green]
>>In <qtqi20d8396ui5 a4eib2336ksvoqf sqlh1@4ax.com> Alan Balmer <albalmer@att.n et> writes:
>>[color=darkred]
>>>On 10 Feb 2004 19:32:07 GMT, Dan.Pop@cern.ch (Dan Pop) wrote:
>>>
>>>>In <c0b1um$jue$1@c hessie.cirr.com > Christopher Benson-Manica <ataru@nospam.c yberspace.org> writes:
>>>>
>>>>>Dan Pop <Dan.Pop@cern.c h> spoke thus:
>>>>>
>>>>>>>Hence, the secret formula of Coca Cola.
>>>>>
>>>>>> How could it be secret? And whatever patent Coca Cola might have had on
>>>>>> it, it has expired long ago...
>>>>>
>>>>>I believe it's classified as a "trade secret" or something to that
>>>>>effect, which is treated differently than a patent or copyright.
>>>>
>>>>A trade secret is a secret as long as it can be kept secret. Even if a
>>>>chemical analysis of Coca Cola would be illegal, if someone started
>>>>producing a beverage with the very same formula, you'd have to prove that
>>>>they didn't reinvent it from scratch.
>>>
>>>A chemical analysis won't tell you how to duplicate it.[/color]
>>
>>Please elaborate. Chemical analysis is not the same thing as atomical
>>analysis, telling which chemical elements are present and in what
>>proportions .[/color]
>
>Actually, what you call atomical analysis is a goodly part of chemical
>analysis.[/color]

It's a purely physical analysis.
[color=blue]
>However, chemical analysis is a broad term, covering many
>other analytical techniques as well. It's also quite limited. In a
>complex substance or mixture of substances, the description of the
>sample obtained by chemical analysis won't usually correspond to a
>formula for duplicating the sample.[/color]

If you know the exact composition of the sample, and this is possible,
figuring out a formula for duplicating the sample shouldn't be too
difficult. Especially when you don't need an exact duplication, merely
something that most people cannot organolepticall y discern from the
original.
[color=blue]
>In general, we have only limited success in synthesizing natural
>substances,[/color]

Coke hardly qualifies as a natural substance.
[color=blue]
>It is part of the mystique, though. Coca-Cola corporate rules are that
>two people have possession of the formula, their identities are not
>disclosed, and they never travel together.[/color]

Someone must actually "implement" the formula, on an industrial scale, in
each and every Coca-Cola factory around the world.
[color=blue]
>In actuality, I suspect
>some of the people on the syrup production floor have a pretty good
>notion as to what's in it.[/color]

The big secret is that there is no unique Coca-Cola formula. The thing
tastes differently in different parts of the world and the caffeine
contents is also locale-specific.
[color=blue]
>I'm afraid this is really getting off-topic ;-)[/color]

Otherwise, the [OT] tag of the subthread wouldn't be justified ;-)

Dan
--
Dan Pop
DESY Zeuthen, RZ group
Email: Dan.Pop@ifh.de

[Keith Thompson]

Re: [OT] Re: Aho Corasick source code needed, please help

Dan.Pop@cern.ch (Dan Pop) writes:
[...][color=blue]
> If you know the exact composition of the sample, and this is possible,
> figuring out a formula for duplicating the sample shouldn't be too
> difficult.[/color]

I'm skeptical of this claim. I'm not enough of a chemist to be able
to justify my skepticism. Are you?

--
Keith Thompson (The_Other_Keit h) kst-u@mib.org <http://www.ghoti.net/~kst>
San Diego Supercomputer Center <*> <http://www.sdsc.edu/~kst>
Schroedinger does Shakespeare: "To be *and* not to be"

[Michael Wojcik]

Re: [OT] Re: Aho Corasick source code needed, please help


In article <c05fk0$136c3i$ 1@ID-179017.news.uni-berlin.de>, "osmium" <r124c4u102@com cast.net> writes:[color=blue]
>
> And thus the fourth category of intellectual property, the trade secret. No
> legal protection, just a way of doing business, as in Colonel Sander's
> "secret blend of 11 herbs and spices". Or whatever. To me, considering
> modern chemistry, this doesn't seem plausible. A lot of mystique and
> dependence on the gullibility of the general public. How hard could it be
> to reverse engineer a bottle of Coca Cola?[/color]

Not hard, and formulations that are so close to the various Coca Colas
(there's more than one) as makes no difference are well-known. Coca
Cola's "secret formula" is a marketing gimick, not a piece of key IP.

William Poundstone has written an entertaining series of books titled
_Big Secrets_, _Bigger Secrets_, and so forth which expose numerous
trade secrets that aren't particularly secret. The KFC seasoning blend
is also among them. (Hint: There aren't 11 components, at least not in
significant amounts.) Of course it is possible that Poundstone's
research and reverse engineering are wrong in some cases, but I suspect
he's generally right.

These books are of course under copyright, which brings us neatly back
around.

--
Michael Wojcik michael.wojcik@ microfocus.com

The way things were, were the way things were, and they stayed that way
because they had always been that way. -- Jon Osborne

[Arthur J. O'Dwyer]

Re: [OT] Re: Aho Corasick source code needed, please help


On Wed, 11 Feb 2004, Keith Thompson wrote:[color=blue]
>
> Dan.Pop@cern.ch (Dan Pop) writes:
> [...][color=green]
> > If you know the exact composition of the sample, and this is possible,
> > figuring out a formula for duplicating the sample shouldn't be too
> > difficult.[/color]
>
> I'm skeptical of this claim. I'm not enough of a chemist to be able
> to justify my skepticism. Are you?[/color]

Given a chemical formula for X, you can produce X from scratch.
Of course, you may require lots of expensive equipment, possibly even
including an atom-smasher to get the rarer elements, ;-) and you may
not be clever enough to find a *convenient* *commercially-feasible*
method of mass production of X; but you can certainly duplicate X
given enough time and money! It's just a matter of sticking atoms
together!
Remember, the Standard makes no claims about efficiency...

-Arthur

[Alan Balmer]

Re: [OT] Re: Aho Corasick source code needed, please help

On Wed, 11 Feb 2004 21:48:21 -0500 (EST), "Arthur J. O'Dwyer"
<ajo@nospam.and rew.cmu.edu> wrote:
[color=blue]
>
>On Wed, 11 Feb 2004, Keith Thompson wrote:[color=green]
>>
>> Dan.Pop@cern.ch (Dan Pop) writes:
>> [...][color=darkred]
>> > If you know the exact composition of the sample, and this is possible,
>> > figuring out a formula for duplicating the sample shouldn't be too
>> > difficult.[/color]
>>
>> I'm skeptical of this claim. I'm not enough of a chemist to be able
>> to justify my skepticism. Are you?[/color]
>
> Given a chemical formula for X, you can produce X from scratch.
>Of course, you may require lots of expensive equipment, possibly even
>including an atom-smasher to get the rarer elements, ;-) and you may
>not be clever enough to find a *convenient* *commercially-feasible*
>method of mass production of X; but you can certainly duplicate X
>given enough time and money! It's just a matter of sticking atoms
>together![/color]

Too much Startrek <G>. The universal synthesizer won't be invented for
a while yet. That's where the time comes in - a century or so <g>.
[color=blue]
> Remember, the Standard makes no claims about efficiency...
>
>-Arthur[/color]

--
Al Balmer
Balmer Consulting
removebalmercon sultingthis@att .net

[David Resnick]

Re: [OT] Re: Aho Corasick source code needed, please help

"Arthur J. O'Dwyer" <ajo@nospam.and rew.cmu.edu> wrote in message news:<Pine.LNX. 4.58-035.04021121452 10.10604@unix40 .andrew.cmu.edu >...[color=blue]
> On Wed, 11 Feb 2004, Keith Thompson wrote:[color=green]
> >
> > Dan.Pop@cern.ch (Dan Pop) writes:
> > [...][color=darkred]
> > > If you know the exact composition of the sample, and this is possible,
> > > figuring out a formula for duplicating the sample shouldn't be too
> > > difficult.[/color]
> >
> > I'm skeptical of this claim. I'm not enough of a chemist to be able
> > to justify my skepticism. Are you?[/color]
>
> Given a chemical formula for X, you can produce X from scratch.
> Of course, you may require lots of expensive equipment, possibly even
> including an atom-smasher to get the rarer elements, ;-) and you may
> not be clever enough to find a *convenient* *commercially-feasible*
> method of mass production of X; but you can certainly duplicate X
> given enough time and money! It's just a matter of sticking atoms
> together!
> Remember, the Standard makes no claims about efficiency...
>
> -Arthur[/color]

This is nonsense. The "chemical formula" of an item does not describe
it
necessary. In terms of chemical composition, a diamond = a buckyball
= graphite, all are pure Carbon. Furthermore, even exact knowledge of
the chemical composition AND structure of a substance does not mean
that you can construct it. The PROCESS by which something is made can
not necessarily be deduced from its chemical composition, or even its
final form. As a concrete example of this in the biotech industry,
the exact composition AND structures of many large proteins are know.
However, producing those proteins via chemistry may be extraordinarly
difficult as the PROCESS by which the proteins are induced to fold
into their "correct" final form is, well, complex. Information about
intermediate steps is lost (consider trying to convert assembly code
back to the original C, to mention something on topic here, yeah, I
know this is marked OT, but...).

Mind you, I expect something relatively simple like Coke could be
reverse engineered. But there would be significant trial and error
involved to get a process for mixing the ingredients that resulted in
the final product being identical in not just "chemical composition"
but actual perceived taste...

-David

[pete]

Re: [OT] Re: Aho Corasick source code needed, please help

David Resnick wrote:
[color=blue]
> Mind you, I expect something relatively simple like Coke could be
> reverse engineered. But there would be significant trial and error
> involved to get a process for mixing the ingredients that resulted in
> the final product being identical in not just "chemical composition"
> but actual perceived taste...[/color]

Coke has both natural and artificial ingedients.

The Thing which prevents Coke from being duplicated exactly,
is the Law Of Diminishing Returns.
If you're trying to duplicate Coke,
the chances are that what you really want,
is a cheap cola that tastes as good or better
and once you've got either of those, you're done.

http://www.snopes.com/cokelore/formula.asp

--
pete