post-doomsday computing

Segher Boessenkool wrote:
>
> Walter Bushell wrote:
> >>>And even pictorial symbols may be tricky. For example a picture of a bee
> >>>may be a rhebus indicating use the sound for bee. However in an unknown
> >>>language used by people dead and buried we don't know how that sounded.
> >>
> >>I've always wondered: if you put the Macintosh "dogcow" on a road sign,
> >>how many different ways will drivers interpret that sign?
> >
> > Moof!
>
> /* from Traps.h */
> enum {
> _Unimplemented = 0xA89F,
> _InitDogCow = 0xA89F,
> _EnableDogCow = 0xA89F,
> _DisableDogCow = 0xA89F,
> _Moof = 0xA89F
> };
> /* Moof indeed! */
>
Quite obviously, "moof!" is the sound made by the dogcow.
That should be clear to the most casual observer (listener)...

--
+----------------------------------------------------------------+
| Charles and Francis Richmond richmond at plano dot net |
+----------------------------------------------------------------+
 >> Stay informed about: post-doomsday computing 

In article <aHiVa.2181$o%2.1274@sccrnsc02>
ljelmore.TakeThisOut@comcast.net "Larry Elmore" writes:

> Erik Max Francis wrote:
> >
> > We have deciphered a great number of dead languages from a hell of a lot
> > less than an encyclopedia.
>
> Name one where we didn't have bi- or tri-lingual copies of writings,
> with one version in a language and script that we already understood.

Linear B? Mind you, that was an inspired bit of guesswork by Ventris &
Chadwick.

Not that I'm in any way attempting to support the rather sweeping
statement that Erik made.

(Still waiting for Etruscan to be deciphered; as was Claudius, and it had
been dead for a far shorter period in his time.)

--
Brian {Hamilton Kelly} bhk.TakeThisOut@dsl.co.uk
"We can no longer stand apart from Europe if we would. Yet we are
untrained to mix with our neighbours, or even talk to them".
George Macaulay Trevelyan, 1919
 >> Stay informed about: post-doomsday computing 

Robert J. Kolker wrote:
> Stefan Ram wrote:
>
>>
>>
>> So what do you know for sure?
>
> Damned little. Life is mostly guesswork. If your guesses are good you
> live to reproduce. If not, you die.
>
> Bob Kolker

Think of it as evolution in action.

(Sorry, couldn't resist!)

Scott Knights
 >> Stay informed about: post-doomsday computing 

On Sun, 27 Jul 2003 23:40:08 BST in alt.folklore.computers,
Tennant Stuart <tennant.RemoveThis@argonet.co.uk> wrote:

>In article <f1n3ivsvopfq1hhmnrchv7jva95lcpgt9k.RemoveThis@4ax.com>, Brian Inglis
><Brian.Inglis.RemoveThis@SystematicSw.ab.ca> wrote:
>
>> On Fri, 25 Jul 2003 17:31:26 BST in alt.folklore.computers,
>> Tennant Stuart <tennant.RemoveThis@argonet.co.uk> wrote:
>
>>> One does have a crib, it's called physics & chemistry.
>
>> Read (a translation of) a medieval or earlier science book
>> written in Greek or Latin for a different cultural view on
>> explaining a common subject. Now imagine all you had was the
>> original text (no diagrams or equations) and there was no common
>> culture or language to help with the translation: would you even
>> be able to identify the subject of the book as science?
>
>I'm not talking about medieval or earlier civilisations.
>
>I'm talking about future generations looking back at us, or
>we examining some extinct alien culture of about our level.

You don't think we might look like archaic, quaint, and
long-winded writers to future generations?
You don't think an extinct alien culture might be quaint, weird,
and, well, alien, to the point where it might be difficult to
decide just what it's culture was based on?

Every branch of science and technology creates its own jargon to
communicate more efficiently. Each jargon depends on the
languages in which fundamental concepts and ideas are
communicated and on the different paths taken during development
of a corpus of knowledge in those areas in which each language is
used. As time and thinking progresses, information is compressed
into laws, theories, hypotheses, conjectures, etc. which are
associated with the names of their primary communicators in our
culture. Without a historical corpus in a subject, and ability to
understand that corpus, how meaningful would technical articles
be which refer to basic knowledge encoded as combinations of
people's names, and other information referred to using words
which have common meanings, possibly in a different language, but
a specialized meaning in one area of jargon?

Thanks. Take care, Brian Inglis Calgary, Alberta, Canada
--
Brian.Inglis.RemoveThis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca)
fake address use address above to reply
 >> Stay informed about: post-doomsday computing 

Brian {Hamilton Kelly} wrote:

> (Still waiting for Etruscan to be deciphered; as was Claudius, and it had
> been dead for a far shorter period in his time.)

Is that the same Claudius as in the t.v. series - I Clavdivs-?

Bob Kolker

>
 >> Stay informed about: post-doomsday computing 

Brian Inglis wrote:

> Every branch of science and technology creates its own jargon to
> communicate more efficiently. Each jargon depends on the
> languages in which fundamental concepts and ideas are
> communicated and on the different paths taken during development
> of a corpus of knowledge in those areas in which each language is
> used. As time and thinking progresses, information is compressed
> into laws, theories, hypotheses, conjectures, etc. which are
> associated with the names of their primary communicators in our
> culture. Without a historical corpus in a subject, and ability to
> understand that corpus, how meaningful would technical articles
> be which refer to basic knowledge encoded as combinations of
> people's names, and other information referred to using words
> which have common meanings, possibly in a different language, but
> a specialized meaning in one area of jargon?

In short, the written tradition rests squarely on the oral tradition,
aka cultural context. If the context of every written article were
unpacked and made explicit their length would be multiplied by at least
a thousand.

Bob Kolker
 >> Stay informed about: post-doomsday computing 

Brian Inglis <Brian.Inglis.TakeThisOut@SystematicSw.ab.ca> writes:

> Every branch of science and technology creates its own jargon to
> communicate more efficiently. Each jargon depends on the languages
> in which fundamental concepts and ideas are communicated and on the
> different paths taken during development of a corpus of knowledge in
> those areas in which each language is used. [...] Without a
> historical corpus in a subject, and ability to understand that
> corpus, how meaningful would technical articles be which refer to
> basic knowledge encoded as combinations of people's names, and other
> information referred to using words which have common meanings,
> possibly in a different language, but a specialized meaning in one
> area of jargon?

And even when we're talking about *our own species*, this is the case.

Take historical music theory, as a for-instance. Ignore for a moment
questions of taste and quality; the project is figuring out how to
recreate the sound from the notation. With the vast majority of music
written before the advent of recording, we're working with a *best
guess* -- because "everybody knew" how violins or singers were
supposed to sound, or the proper amount of tempo variation to use when
playing Bach or Chopin.

It gets even worse when you start talking about music before 1500;
with the advent of keyboard instruments, the frame of reference for
talking about music shifted from the voice to the keyboard. As a
result, notation *changed* -- composers in the 15th century could
write an F, and expect that the performers would alter it to an F#
because the context required it. The performers, likewise, would see
a written F and have no qualms about altering it to an F# based on the
context. And this would not be seen as incompetence on the part of
composer or publisher, or as hubris on the part of the performers; it
was so commonplace, as a matter of fact, that remarkably few people
saw fit to comment on it.

In short, this is an enormous corpus of information, with volumes of
material (scores) and metamaterial (treatises and commentaries)
available, and a continuous oral tradition that goes back over a
thousand years, and we can't say to a certainty what 500-year-old
music should sound like. I despair of any communication with people
in the 25th century, let alone the 50th.

Charlton
 >> Stay informed about: post-doomsday computing 

On Mon, 28 Jul 2003 23:17:10 GMT, Robert J. Kolker wrote:
>
>
> Erik Max Francis wrote:
>>
>> We have deciphered a great number of dead languages from a hell of a lot
>> less than an encyclopedia.
>
> Were they dead or did they have a living link?
>
> Can you enumerate some of these "dead" languages (skip Egyptian
> hieroglyph and hieratic which we know from the Rosetta Stone).
>
> Also keep in mind that Latin has been spoken and written from the time
> of the Roman Empire (thank the R.C. for that), Hebrew and Aramaic have
> been spoken at various places since the time of Joshuah so strictly
> speaking they always have been live languages.

A language very close to hebrew would have been used in present-day
Syria from 1500ce at least. [1] Aramaic would have been a meta-language
through most of its history, like Latin, used to communicate among
those who used a differing language for everyday use. I suppose
that Aramaic would be more like a HL computer language, the users
brains would have translated it into their usual language to use the
message. (That would be the non-Aramaic native speakers). Like Latin
in Christian Europe. Or Arabic in Pakistan and Indonesia and Iran.

[Rad-Shamra finds]

--
greymaus
downloading Captain Kirk. ..................
ERROR
NO CARRIER
 >> Stay informed about: post-doomsday computing 

In article <3F2481BC.154F30C8.TakeThisOut@alcyone.com>, Erik Max Francis
<max.TakeThisOut@alcyone.com> wrote:

> Tennant Stuart wrote:

>> In article <bftpg8$evm$3@bob.news.rcn.net>, jmfbahciv.TakeThisOut@aol.com wrote:

>>> You are making the assumption that physics and chemistry
>>> are the same everywhere. That's not quite true .

>> Indeed? Please prove that assertion.

> It's certainly not known without a doubt that it's true, but a great
> deal of modern astronomy and cosmology relies on it being true. The
> general name for the idea is the mediocrity principle.

Quite so, Eric.


Tennant

--
____ ____ _ _ _ _ __ _ _ ____
(_ _)( ___)( \( )( \( ) /__\ ( \( )(_ _) Greetings to family
)( )__) ) ( ) ( /(__)\ ) ( )( friends & neighbours
(__) (____)(_)\_)(_)\_)(__)(__)(_)\_) (__) @argonet.co.uk & MCR
 >> Stay informed about: post-doomsday computing 

Brian {Hamilton Kelly} wrote:
> In article <aHiVa.2181$o%2.1274@sccrnsc02>
> ljelmore.DeleteThis@comcast.net "Larry Elmore" writes:
>
>
>>Erik Max Francis wrote:
>>
>>>We have deciphered a great number of dead languages from a hell of a lot
>>>less than an encyclopedia.
>>
>>Name one where we didn't have bi- or tri-lingual copies of writings,
>>with one version in a language and script that we already understood.
>
> Linear B? Mind you, that was an inspired bit of guesswork by Ventris &
> Chadwick.

I worded that carelessly. Mycenaen Greek is a "dead" language, but it
wasn't exactly an unknown language since it was ancestral to Homeric
Greek. In time span, the difference is about the same as between Middle
English and Modern English. Figuring out that the script encoded (in a
highly abbreviated way) an archaic form of Greek was a brilliant piece
of work. Had it not been a language we know, however, we'd still be
scratching our heads.

> Not that I'm in any way attempting to support the rather sweeping
> statement that Erik made.
>
> (Still waiting for Etruscan to be deciphered; as was Claudius, and it had
> been dead for a far shorter period in his time.)

IIRC, Ventris (and others) had spent years trying to relate Linear B to
Etruscan before he gave up on that pet theory.

The problem we face with Etruscan is almost the opposite of the problem
posed by Linear B. With Etruscan, we know their alphabet, we know many
words (if mostly personal and place names), but we know next to nothing
about the language or its grammar, except that it's definitely not
Indo-European.

--Larry
 >> Stay informed about: post-doomsday computing 

Brian {Hamilton Kelly} wrote:
> In article <aHiVa.2181$o%2.1274@sccrnsc02>
> ljelmore DeleteThis @comcast.net "Larry Elmore" writes:
>
>
>>Erik Max Francis wrote:
>>
>>>We have deciphered a great number of dead languages from a hell of a lot
>>>less than an encyclopedia.
>>
>>Name one where we didn't have bi- or tri-lingual copies of writings,
>>with one version in a language and script that we already understood.
>
> Linear B? Mind you, that was an inspired bit of guesswork by Ventris &
> Chadwick.

I worded that carelessly. Mycenaen Greek is a "dead" language, but it
wasn't exactly an unknown language since it was ancestral to Homeric
Greek. In time span, the difference is about the same as between Middle
English and Modern English. Figuring out that the script encoded (in a
highly abbreviated way) an archaic form of Greek was a brilliant piece
of work. Had it not been a language we know, however, we'd still be
scratching our heads.

> Not that I'm in any way attempting to support the rather sweeping
> statement that Erik made.
>
> (Still waiting for Etruscan to be deciphered; as was Claudius, and it had
> been dead for a far shorter period in his time.)

IIRC, Ventris (and others) had spent years trying to relate Linear B to
Etruscan before he gave up on that pet theory.

The problem we face with Etruscan is almost the opposite of the problem
posed by Linear B. With Etruscan, we know their alphabet, we know many
words (if mostly personal and place names), but we know next to nothing
about the language or its grammar, except that it definitely not
Indo-European.

--Larry
 >> Stay informed about: post-doomsday computing 

In article <_R_Ua.162342$ye4.109272@sccrnsc01>, bobkolker.DeleteThis@comcast.net
says...
>
>
> John Schilling wrote:
>
> > Storing the OTP in a read-once medium is the way to deal with that threat.
> > Quantum cryptography is the only way to accomplish this that is guaranteed
> > by the laws of physics, but reasonably good practical approximations can be
> > engineered using classical techniques.
>
> Come up with a 100 percent assurance that the RO disk that was delivered
> was the one that was sent originally. Consider an attack which
> substitutes a bogus pair of disks for the real ones.
>
> OTP is overrated which is why it is not used much.
>

If you're suggesting that the receiver *and* the senders OTP key disk
could be switched, then that has to be tricky to arrange.

If you're after a man-in-th-middle attack, then that can be defeated in
this case by sending the OTP encrypted message across a public medium -
the man in the middle sees it, but the intended recipient also sees it,
but can't read it.

If the disks haven't been compromised, OTP is so secure there's no real
problem <hmm - knowing of the mere existence of a message could in some
circumstances be as useful as the contents? - maybe leave that till
later...>.

Result - at worst one message compromised, but the compromise of the
whole pad is known for certain as well.
 >> Stay informed about: post-doomsday computing 

On Tue, 29 Jul 2003, Paul wrote:

> If the disks haven't been compromised, OTP is so secure there's no real
> problem <hmm - knowing of the mere existence of a message could in some
> circumstances be as useful as the contents? - maybe leave that till
> later...>.

There's a few "real" problems with using a pure OTP. (besides the
obvious key-management ones)

One of the more well-known ones is that if you know the plaintext of a
transmission, and have opportunity to change the ciphertext, you can
change what the receiver reads, and remain undetected.

Typical scenario would be if you leaked a "confidential" message to a
embassy, in the hope that it would be transmitted unmodified with OTP to
the homeland.

If you could then intercept and change the ciphertext, you could make it
appear that the embassy sent any message of up to that length.

Method:

Known plaintext xor intercepted ciphertext = OTP.
Wanted plaintext xor OTP = valid-looking ciphertext.

To guard against this possibility, it is advisable to have some sort of
authenthication on a message before OTP-encrypting it. Some kind of
digital signature will do fine.

Never use OTP to send "Yes". An attacker could guess it, and change it
to, for example "No".


Sincerely,
Eivind Kjørstad
 >> Stay informed about: post-doomsday computing