Discussion:
Technical limits on Wormhole Jumps
(too old to reply)
John Fairhurst
2013-06-22 05:55:31 UTC
Permalink
Hi,

In the Honorverse books the transition between systems using
wormholes is done maually in so much as you still have someone on
the bridge doing the equivilent of flicking a switch in change
the configuration of the sails.

However, I can't see any reason why this can't be fully automated,
especially given the Star Empire's increasing shortage of trained
staff. I could also see an automated shuttle based on a courier
that could be used for transmitting high priority communications
through the wormhole.
--
John Fairhurst
http://www.johnsbooks.co.uk



----Android NewsGroup Reader----
http://www.piaohong.tk/newsgroup
Loren Pechtel
2013-06-22 22:07:08 UTC
Permalink
On Sat, 22 Jun 2013 06:55:31 +0100 (GMT+01:00), John Fairhurst
Post by John Fairhurst
However, I can't see any reason why this can't be fully automated,
especially given the Star Empire's increasing shortage of trained
staff. I could also see an automated shuttle based on a courier
that could be used for transmitting high priority communications
through the wormhole.
That's what courier boats are for.
John Fairhurst
2013-06-23 05:26:48 UTC
Permalink
Post by Loren Pechtel
On Sat, 22 Jun 2013 06:55:31 +0100 (GMT+01:00), John Fairhurst
Post by John Fairhurst
However, I can't see any reason why this can't be fully automated,
especially given the Star Empire's increasing shortage of trained
staff. I could also see an automated shuttle based on a courier
that could be used for transmitting high priority communications
through the wormhole.
That's what courier boats are for.
Well, yes and no. Courier boats are most useful for sending
messages through hyperspace in general. What my automated
shuttlers would be is a wormhole specific message relay link for
something like the Manticoran/Lynx nodes of the Mantore
Terminus.
Quite specialised tech and they'd not be much smaller than courier
boats so they'd only be economical/safe on a fairly limited

number of routes I guess

Regards
--
John Fairhurst
http://www.johnsbooks.co.uk



----Android NewsGroup Reader----
http://www.piaohong.tk/newsgroup
Michael R N Dolbear
2013-06-23 19:28:34 UTC
Permalink
Post by John Fairhurst
Post by Loren Pechtel
On Sat, 22 Jun 2013 06:55:31 +0100 (GMT+01:00), John Fairhurst
Post by John Fairhurst
However, I can't see any reason why this can't be fully automated,
especially given the Star Empire's increasing shortage of trained
staff. I could also see an automated shuttle based on a courier
that could be used for transmitting high priority communications
through the wormhole.
That's what courier boats are for.
Well, yes and no. Courier boats are most useful for sending
messages through hyperspace in general. What my automated
shuttlers would be is a wormhole specific message relay link for
something like the Manticoran/Lynx nodes of the Mantore
Terminus.
Quite specialised tech and they'd not be much smaller than courier
boats so they'd only be economical/safe on a fairly limited
number of routes I guess
The Mad Wizard says a Courier boat is the minimum, fusion reactor,
hyper generator, impeller drive plus hull. All your automated shuttlers
would save would be life support, so this is 196,000 tons rather than
200,000 ?
--
Mike D
pyotr filipivich
2013-06-24 20:27:06 UTC
Permalink
Post by Michael R N Dolbear
Post by John Fairhurst
Post by Loren Pechtel
On Sat, 22 Jun 2013 06:55:31 +0100 (GMT+01:00), John Fairhurst
Post by John Fairhurst
However, I can't see any reason why this can't be fully automated,
especially given the Star Empire's increasing shortage of trained
staff. I could also see an automated shuttle based on a courier
that could be used for transmitting high priority communications
through the wormhole.
That's what courier boats are for.
Well, yes and no. Courier boats are most useful for sending
messages through hyperspace in general. What my automated
shuttlers would be is a wormhole specific message relay link for
something like the Manticoran/Lynx nodes of the Mantore
Terminus.
Quite specialised tech and they'd not be much smaller than courier
boats so they'd only be economical/safe on a fairly limited
number of routes I guess
The Mad Wizard says a Courier boat is the minimum, fusion reactor,
hyper generator, impeller drive plus hull. All your automated shuttlers
would save would be life support, so this is 196,000 tons rather than
200,000 ?
Which tells us a bit about the size of the probes which get sent
through wormholes to survey them.

OTOH, a courier boat could go through a wormhole and just "pass
off" the messages. Thus you have a three part/stage "pony" express" -
planet side to near the wormhole, transfer to the "boat" which goes
through the wormhole, and on the other side, transfers it for local
delivery.

OT3H - this might only be useful in systems with high message
traffic, which would most likely have a high level of physical
traffic. In other words, the pony "boat" would have to take a number
and get in line, unless it were possible to "cut the line". One
method would be to have the merchies slow a bit to a less than Optimal
wait between transits, and eventually you will have enough time to
"pop" a tiny boat through and have the wormhole re stabilize before
the freighter arrives "on schedule".
--
pyotr filipivich.
Just about the time you finally see light at the end of the tunnel,
you find out it's a Government Project to build more tunnel.
Loren Pechtel
2013-06-24 23:03:32 UTC
Permalink
On Mon, 24 Jun 2013 13:27:06 -0700, pyotr filipivich
Post by pyotr filipivich
Post by Michael R N Dolbear
The Mad Wizard says a Courier boat is the minimum, fusion reactor,
hyper generator, impeller drive plus hull. All your automated shuttlers
would save would be life support, so this is 196,000 tons rather than
200,000 ?
Which tells us a bit about the size of the probes which get sent
through wormholes to survey them.
OTOH, a courier boat could go through a wormhole and just "pass
off" the messages. Thus you have a three part/stage "pony" express" -
planet side to near the wormhole, transfer to the "boat" which goes
through the wormhole, and on the other side, transfers it for local
delivery.
Actually it wouldn't. Wormhole probes might not be build to last.
Post by pyotr filipivich
OT3H - this might only be useful in systems with high message
traffic, which would most likely have a high level of physical
traffic. In other words, the pony "boat" would have to take a number
and get in line, unless it were possible to "cut the line". One
method would be to have the merchies slow a bit to a less than Optimal
wait between transits, and eventually you will have enough time to
"pop" a tiny boat through and have the wormhole re stabilize before
the freighter arrives "on schedule".
For high volume systems you don't even need that. Manticore could
simply require that everyone using a junction accept a file while
they're in the queue and deliver it at the other end. It could be
encrypted with a one-time pad and thus utterly unbreakable. In
reality every message is sent many times along with the identity of
the ships it's sent on and an acknowledgement of each message sent is
sent back the other way.

In fact I would think something like this would be standard for *ALL*
shipping. Any ship heading somewhere would accept a file for the
target system. Since the cost of actually doing this is very minimal
I would think every port would simply mandate participation in the
mail system.
Michael R N Dolbear
2013-06-25 19:35:25 UTC
Permalink
Post by Loren Pechtel
On Mon, 24 Jun 2013 13:27:06 -0700, pyotr filipivich
Post by pyotr filipivich
Post by Michael R N Dolbear
The Mad Wizard says a Courier boat is the minimum, fusion reactor,
hyper generator, impeller drive plus hull. All your automated shuttlers
would save would be life support, so this is 196,000 tons rather than
200,000 ?
Which tells us a bit about the size of the probes which get sent
through wormholes to survey them.
Actually it wouldn't. Wormhole probes might not be build to last.
The textev seems to be that wormhole probes aren't "sent though"
wormholes at all. They merely survey the "gravitational whirlpool" of a
WHJ from normal space, sending back telemetry before being destroyed.

There is no account of anything smaller than a Courier boat traversing
a WHJ or using a hyper generator.
--
Mike D
JohnFair
2013-08-06 11:50:12 UTC
Permalink
Post by Loren Pechtel
For high volume systems you don't even need that. Manticore could
simply require that everyone using a junction accept a file while
they're in the queue and deliver it at the other end. It could be
encrypted with a one-time pad and thus utterly unbreakable. In
reality every message is sent many times along with the identity of
the ships it's sent on and an acknowledgement of each message sent is
sent back the other way.
In fact I would think something like this would be standard for *ALL*
shipping. Any ship heading somewhere would accept a file for the
target system. Since the cost of actually doing this is very minimal
I would think every port would simply mandate participation in the
mail system.
I can see something like that being acceptable for fairly low grade traffic but you'd want control over the delivery vessel for high grade traffic. Even one time pads can be broken - it's just the multi millennial processing periods that don't make it worthwhile now. Given the substantial increases in computer power available to these people you might find it taking a few days if not hours running on high quality hard/software.

I'd still prefer an automated message boat - it's the crews I was thinking of more than any downsizing of the volume of the ship
Aahz Maruch
2013-08-06 12:59:45 UTC
Permalink
I can see something like that being acceptable for fairly low grade traffic=
but you'd want control over the delivery vessel for high grade traffic. Ev=
en one time pads can be broken - it's just the multi millennial processing =
periods that don't make it worthwhile now. Given the substantial increases =
in computer power available to these people you might find it taking a few =
days if not hours running on high quality hard/software.=20
You have a cite for breaking one-time pads?
--
Hugs and backrubs -- I break Rule 6 http://rule6.info/
<*> <*> <*>
Worried about terrorists? Of course I'm worried about terrorists. The
terrorists I'm worried about are in the White House.
Bob Casanova
2013-08-06 17:20:39 UTC
Permalink
On 6 Aug 2013 05:59:45 -0700, the following appeared in
Post by Aahz Maruch
I can see something like that being acceptable for fairly low grade traffic=
but you'd want control over the delivery vessel for high grade traffic. Ev=
en one time pads can be broken - it's just the multi millennial processing =
periods that don't make it worthwhile now. Given the substantial increases =
in computer power available to these people you might find it taking a few =
days if not hours running on high quality hard/software.=20
You have a cite for breaking one-time pads?
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a
published document, I suspect that, while a particular pad
might yield to an exhaustive analysis of *all* published
documents by a sufficiently advanced computer, there would
be a fairly large number of potential decryptions of a
message, with no way to tell which is correct.
--
Bob C.

"The most exciting phrase to hear in science,
the one that heralds new discoveries, is not
'Eureka!' but 'That's funny...'"

- Isaac Asimov
Loren Pechtel
2013-08-06 21:49:39 UTC
Permalink
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a
published document, I suspect that, while a particular pad
might yield to an exhaustive analysis of *all* published
documents by a sufficiently advanced computer, there would
be a fairly large number of potential decryptions of a
message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.

Thus something based on a published work is not a one time pad
although it uses some of the same concepts.
Aahz Maruch
2013-08-06 23:33:07 UTC
Permalink
Post by Loren Pechtel
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a published
document, I suspect that, while a particular pad might yield to an
exhaustive analysis of *all* published documents by a sufficiently
advanced computer, there would be a fairly large number of potential
decryptions of a message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.
Yeah, and I'll bet they have better RNGs in the future. ;-)
--
Hugs and backrubs -- I break Rule 6 http://rule6.info/
<*> <*> <*>
Worried about terrorists? Of course I'm worried about terrorists. The
terrorists I'm worried about are in the White House.
Scott Lurndal
2013-08-07 13:49:06 UTC
Permalink
Post by Aahz Maruch
Post by Loren Pechtel
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a published
document, I suspect that, while a particular pad might yield to an
exhaustive analysis of *all* published documents by a sufficiently
advanced computer, there would be a fairly large number of potential
decryptions of a message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.
Yeah, and I'll bet they have better RNGs in the future. ;-)
Better than this?

http://www.lavarnd.org/
Loren Pechtel
2013-08-07 21:16:25 UTC
Permalink
Post by Aahz Maruch
Post by Loren Pechtel
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a published
document, I suspect that, while a particular pad might yield to an
exhaustive analysis of *all* published documents by a sufficiently
advanced computer, there would be a fairly large number of potential
decryptions of a message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.
Yeah, and I'll bet they have better RNGs in the future. ;-)
A true one-time pad uses a *TRUE* random number generator--say, a
geiger counter next to something radioactive.
Aahz Maruch
2013-08-08 05:48:35 UTC
Permalink
Post by Loren Pechtel
Post by Aahz Maruch
Post by Loren Pechtel
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a published
document, I suspect that, while a particular pad might yield to an
exhaustive analysis of *all* published documents by a sufficiently
advanced computer, there would be a fairly large number of potential
decryptions of a message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.
Yeah, and I'll bet they have better RNGs in the future. ;-)
A true one-time pad uses a *TRUE* random number generator--say, a
geiger counter next to something radioactive.
First of all, note the smiley. Second, I was making a funny about
algorithms for generating random numbers.
--
Hugs and backrubs -- I break Rule 6 http://rule6.info/
<*> <*> <*>
"I don't object to sex before marriage, but two minutes before?!?"
Bob Casanova
2013-08-07 17:13:59 UTC
Permalink
On Tue, 06 Aug 2013 14:49:39 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a
published document, I suspect that, while a particular pad
might yield to an exhaustive analysis of *all* published
documents by a sufficiently advanced computer, there would
be a fairly large number of potential decryptions of a
message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.
Thus something based on a published work is not a one time pad
although it uses some of the same concepts.
Cite? While the best pads are based on a random key, any pad
in which the key is used only once qualifies (since it still
requires knowledge of the key to crack), at least so I've
been led to believe.
--
Bob C.

"The most exciting phrase to hear in science,
the one that heralds new discoveries, is not
'Eureka!' but 'That's funny...'"

- Isaac Asimov
Loren Pechtel
2013-08-07 21:16:25 UTC
Permalink
Post by Bob Casanova
On Tue, 06 Aug 2013 14:49:39 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a
published document, I suspect that, while a particular pad
might yield to an exhaustive analysis of *all* published
documents by a sufficiently advanced computer, there would
be a fairly large number of potential decryptions of a
message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.
Thus something based on a published work is not a one time pad
although it uses some of the same concepts.
Cite? While the best pads are based on a random key, any pad
in which the key is used only once qualifies (since it still
requires knowledge of the key to crack), at least so I've
been led to believe.
No. A one time pad is uses each byte in the key only once. That's
what the "one time" refers to. A given ciphertext can decrypt to
*ANY* plaintext of the right length depending on the key used.
Bob Casanova
2013-08-08 17:21:40 UTC
Permalink
On Wed, 07 Aug 2013 14:16:25 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
On Tue, 06 Aug 2013 14:49:39 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a
published document, I suspect that, while a particular pad
might yield to an exhaustive analysis of *all* published
documents by a sufficiently advanced computer, there would
be a fairly large number of potential decryptions of a
message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.
Thus something based on a published work is not a one time pad
although it uses some of the same concepts.
Cite? While the best pads are based on a random key, any pad
in which the key is used only once qualifies (since it still
requires knowledge of the key to crack), at least so I've
been led to believe.
No. A one time pad is uses each byte in the key only once. That's
what the "one time" refers to.
OK, thanks. I assume that for a printed key "each byte"
translates to "each letter"?

(As you can probably tell without much effort, cryptography
isn't my strong suit; I'm a retired EE with a mild interest
in the subject.)
Post by Loren Pechtel
A given ciphertext can decrypt to
*ANY* plaintext of the right length depending on the key used.
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
--
Bob C.

"The most exciting phrase to hear in science,
the one that heralds new discoveries, is not
'Eureka!' but 'That's funny...'"

- Isaac Asimov
pyotr filipivich
2013-08-09 05:56:37 UTC
Permalink
Post by Bob Casanova
On Wed, 07 Aug 2013 14:16:25 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
On Tue, 06 Aug 2013 14:49:39 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a
published document, I suspect that, while a particular pad
might yield to an exhaustive analysis of *all* published
documents by a sufficiently advanced computer, there would
be a fairly large number of potential decryptions of a
message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.
Thus something based on a published work is not a one time pad
although it uses some of the same concepts.
Cite? While the best pads are based on a random key, any pad
in which the key is used only once qualifies (since it still
requires knowledge of the key to crack), at least so I've
been led to believe.
No. A one time pad is uses each byte in the key only once. That's
what the "one time" refers to.
OK, thanks. I assume that for a printed key "each byte"
translates to "each letter"?
Probably.

The important thing is that the key (the one time pad) be long.
Best case scenario is that it is at least as long as the message; thus
each letter in the plain text has a "unique" correspondent in the key.
Post by Bob Casanova
(As you can probably tell without much effort, cryptography
isn't my strong suit; I'm a retired EE with a mild interest
in the subject.)
Post by Loren Pechtel
A given ciphertext can decrypt to
*ANY* plaintext of the right length depending on the key used.
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
In theory. In practice, not so likely.
--
pyotr filipivich.
Just about the time you finally see light at the end of the tunnel,
you find out it's a Government Project to build more tunnel.
Bob Casanova
2013-08-09 17:23:15 UTC
Permalink
On Thu, 08 Aug 2013 22:56:37 -0700, the following appeared
in alt.books.david-weber, posted by pyotr filipivich
Post by pyotr filipivich
Post by Bob Casanova
On Wed, 07 Aug 2013 14:16:25 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
On Tue, 06 Aug 2013 14:49:39 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
Yeah, good luck with that. Especially if the pad is based on
anything not published. And even if it's based on a
published document, I suspect that, while a particular pad
might yield to an exhaustive analysis of *all* published
documents by a sufficiently advanced computer, there would
be a fairly large number of potential decryptions of a
message, with no way to tell which is correct.
A true one-time pad is based on a truly random source of bytes.
Thus something based on a published work is not a one time pad
although it uses some of the same concepts.
Cite? While the best pads are based on a random key, any pad
in which the key is used only once qualifies (since it still
requires knowledge of the key to crack), at least so I've
been led to believe.
No. A one time pad is uses each byte in the key only once. That's
what the "one time" refers to.
OK, thanks. I assume that for a printed key "each byte"
translates to "each letter"?
Probably.
The important thing is that the key (the one time pad) be long.
Best case scenario is that it is at least as long as the message; thus
each letter in the plain text has a "unique" correspondent in the key.
Thanks; that also seems to me to be required by Loren's
comment, since a key shorter than the message would by
definition preclude "letter (or byte) uniqueness".
Post by pyotr filipivich
Post by Bob Casanova
(As you can probably tell without much effort, cryptography
isn't my strong suit; I'm a retired EE with a mild interest
in the subject.)
Post by Loren Pechtel
A given ciphertext can decrypt to
*ANY* plaintext of the right length depending on the key used.
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
In theory. In practice, not so likely.
Granted.
--
Bob C.

"The most exciting phrase to hear in science,
the one that heralds new discoveries, is not
'Eureka!' but 'That's funny...'"

- Isaac Asimov
Clive D. W. Feather
2013-08-15 21:35:55 UTC
Permalink
Post by pyotr filipivich
The important thing is that the key (the one time pad) be long.
Best case scenario is that it is at least as long as the message; thus
each letter in the plain text has a "unique" correspondent in the key.
If the key is not as long as the message, then it's not a one-time pad
and a standard Kerchoffs superimposition can be used to break it.
Post by pyotr filipivich
Post by Bob Casanova
Post by Loren Pechtel
A given ciphertext can decrypt to
*ANY* plaintext of the right length depending on the key used.
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
In theory. In practice, not so likely.
If it's a true one-time pad, then *EVERY* message of the requisite
length is equally likely to be the answer. There is no way to
distinguish them.
--
Clive D.W. Feather | Home: <***@davros.org>
Mobile: +44 7973 377646 | Web: <http://www.davros.org>
Please reply to the Reply-To address, which is: <***@davros.org>
Bob Casanova
2013-08-16 17:10:17 UTC
Permalink
On Thu, 15 Aug 2013 22:35:55 +0100, the following appeared
in alt.books.david-weber, posted by "Clive D. W. Feather"
Post by Clive D. W. Feather
Post by pyotr filipivich
The important thing is that the key (the one time pad) be long.
Best case scenario is that it is at least as long as the message; thus
each letter in the plain text has a "unique" correspondent in the key.
If the key is not as long as the message, then it's not a one-time pad
and a standard Kerchoffs superimposition can be used to break it.
That seems obvious. What wasn't previously obvious (to me)
was the "random key" requirement, but then, I'm no
decryption expert, not by a long shot.
Post by Clive D. W. Feather
Post by pyotr filipivich
Post by Bob Casanova
Post by Loren Pechtel
A given ciphertext can decrypt to
*ANY* plaintext of the right length depending on the key used.
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
In theory. In practice, not so likely.
If it's a true one-time pad, then *EVERY* message of the requisite
length is equally likely to be the answer. There is no way to
distinguish them.
That was also the conclusion presented elsethread. Thanks,
and thanks for your other response.
--
Bob C.

"The most exciting phrase to hear in science,
the one that heralds new discoveries, is not
'Eureka!' but 'That's funny...'"

- Isaac Asimov
Loren Pechtel
2013-08-10 03:06:48 UTC
Permalink
Post by Bob Casanova
Post by Loren Pechtel
No. A one time pad is uses each byte in the key only once. That's
what the "one time" refers to.
OK, thanks. I assume that for a printed key "each byte"
translates to "each letter"?
(As you can probably tell without much effort, cryptography
isn't my strong suit; I'm a retired EE with a mild interest
in the subject.)
If you are only encoding letters you can use letters as your key. If
you need to encode a wider range of data then your key must also cover
a wider range.

From a practical standpoint when the computer is doing the
cryptography each "letter" is in the range of 0 to 255 so you can
encode whatever you have.
Post by Bob Casanova
Post by Loren Pechtel
A given ciphertext can decrypt to
*ANY* plaintext of the right length depending on the key used.
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
"a fairly large number" contains an incorrect qualifier--it's a *VERY*
large number. Nothing can be ruled out.
Bob Casanova
2013-08-10 16:44:51 UTC
Permalink
On Fri, 09 Aug 2013 20:06:48 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
Post by Loren Pechtel
No. A one time pad is uses each byte in the key only once. That's
what the "one time" refers to.
OK, thanks. I assume that for a printed key "each byte"
translates to "each letter"?
(As you can probably tell without much effort, cryptography
isn't my strong suit; I'm a retired EE with a mild interest
in the subject.)
If you are only encoding letters you can use letters as your key. If
you need to encode a wider range of data then your key must also cover
a wider range.
From a practical standpoint when the computer is doing the
cryptography each "letter" is in the range of 0 to 255 so you can
encode whatever you have.
Post by Bob Casanova
Post by Loren Pechtel
A given ciphertext can decrypt to
*ANY* plaintext of the right length depending on the key used.
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
"a fairly large number" contains an incorrect qualifier--it's a *VERY*
large number. Nothing can be ruled out.
OK, although in this context I'd also call a google a
"fairly large number". ;-)

And ISTM the potential size of the number would be dependent
on the number of characters available as well as on the
possible number of permutations resulting in actual words.
--
Bob C.

"The most exciting phrase to hear in science,
the one that heralds new discoveries, is not
'Eureka!' but 'That's funny...'"

- Isaac Asimov
Loren Pechtel
2013-08-10 17:27:10 UTC
Permalink
Post by Bob Casanova
Post by Loren Pechtel
Post by Bob Casanova
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
"a fairly large number" contains an incorrect qualifier--it's a *VERY*
large number. Nothing can be ruled out.
OK, although in this context I'd also call a google a
"fairly large number". ;-)
I would interpret "fairly large" as saying it's large but not really
large.
Bob Casanova
2013-08-11 17:38:56 UTC
Permalink
On Sat, 10 Aug 2013 10:27:10 -0700, the following appeared
in alt.books.david-weber, posted by Loren Pechtel
Post by Loren Pechtel
Post by Bob Casanova
Post by Loren Pechtel
Post by Bob Casanova
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
"a fairly large number" contains an incorrect qualifier--it's a *VERY*
large number. Nothing can be ruled out.
OK, although in this context I'd also call a google a
"fairly large number". ;-)
I would interpret "fairly large" as saying it's large but not really
large.
Ah, of course... ;-)
--
Bob C.

"The most exciting phrase to hear in science,
the one that heralds new discoveries, is not
'Eureka!' but 'That's funny...'"

- Isaac Asimov
pyotr filipivich
2013-08-10 16:50:36 UTC
Permalink
Post by Loren Pechtel
Post by Bob Casanova
Post by Loren Pechtel
No. A one time pad is uses each byte in the key only once. That's
what the "one time" refers to.
OK, thanks. I assume that for a printed key "each byte"
translates to "each letter"?
(As you can probably tell without much effort, cryptography
isn't my strong suit; I'm a retired EE with a mild interest
in the subject.)
If you are only encoding letters you can use letters as your key. If
you need to encode a wider range of data then your key must also cover
a wider range.
From a practical standpoint when the computer is doing the
cryptography each "letter" is in the range of 0 to 255 so you can
encode whatever you have.
Post by Bob Casanova
Post by Loren Pechtel
A given ciphertext can decrypt to
*ANY* plaintext of the right length depending on the key used.
Not sure what you're saying here. Do you mean essentially
the same thing as my statement above, "...there would be a
fairly large number of potential decryptions of a message,
with no way to tell which is correct"?
"a fairly large number" contains an incorrect qualifier--it's a *VERY*
large number. Nothing can be ruled out.
This is why you must never compute all the digits of Pi. OR if
you do, never transform the numbers into strings of letters. Because
eventually, you will find all the Written material all humanity. While
it might be interesting to read some of the "lost literature" out
there, you'll also be getting all the classified material, trashy
romance novels, fan fiction, and freshman Writing 101 essays ever
produced. And there is far more of the later, than the former.
--
pyotr filipivich.
Just about the time you finally see light at the end of the tunnel,
you find out it's a Government Project to build more tunnel.
Loren Pechtel
2013-08-10 17:27:10 UTC
Permalink
On Sat, 10 Aug 2013 09:50:36 -0700, pyotr filipivich
Post by pyotr filipivich
Post by Loren Pechtel
"a fairly large number" contains an incorrect qualifier--it's a *VERY*
large number. Nothing can be ruled out.
This is why you must never compute all the digits of Pi. OR if
you do, never transform the numbers into strings of letters. Because
eventually, you will find all the Written material all humanity. While
it might be interesting to read some of the "lost literature" out
there, you'll also be getting all the classified material, trashy
romance novels, fan fiction, and freshman Writing 101 essays ever
produced. And there is far more of the later, than the former.
And the NSA breaks down your door at 3am.
pyotr filipivich
2013-08-11 06:54:32 UTC
Permalink
Post by Loren Pechtel
On Sat, 10 Aug 2013 09:50:36 -0700, pyotr filipivich
Post by pyotr filipivich
Post by Loren Pechtel
"a fairly large number" contains an incorrect qualifier--it's a *VERY*
large number. Nothing can be ruled out.
This is why you must never compute all the digits of Pi. OR if
you do, never transform the numbers into strings of letters. Because
eventually, you will find all the Written material all humanity. While
it might be interesting to read some of the "lost literature" out
there, you'll also be getting all the classified material, trashy
romance novels, fan fiction, and freshman Writing 101 essays ever
produced. And there is far more of the later, than the former.
And the NSA breaks down your door at 3am.
You might want to run a Regex search for the warrant, orders to do
so. Or the arrest record ...

tschus
pyotr
--
pyotr filipivich.
Just about the time you finally see light at the end of the tunnel,
you find out it's a Government Project to build more tunnel.
Clive D. W. Feather
2013-08-15 21:34:20 UTC
Permalink
Post by Bob Casanova
Cite? While the best pads are based on a random key, any pad
in which the key is used only once qualifies (since it still
requires knowledge of the key to crack), at least so I've
been led to believe.
This is false. If the "pad" is a piece of English text then, even though
it's only used once, you break it by solving the message and the "pad"
simultaneously.

Assume we're using the classic system of a key containing only letters
and where you add plaintext and key modulo 26 (A+A=A, A+B=B, A+C=C, ...,
B+A=B, B+B=C, B+C=D, ... B+Y=Z, B+Z=A, C+A=C, C+B=D, C+C=E, ... C+X=Z,
C+Y=A, C+Z=B, ...

Basically, for each of the 26 possible letters in the cyphertext, there
are 26 possible additions that reached it. But these 26 have very
different probabilities, so you only look at the few most likely. By
then examining surrounding letters, you can determine which is most
likely (if you've got T?E in one string and A?D in the other, and the
intermediate letters add to U, then H+N is a pretty likely answer).

Experts (of which I am not) can handle this pretty quickly. These days I
expect it would be simple to write a program to do it.

One-time pads are only safe if they are truly random and used exactly
once.
--
Clive D.W. Feather | Home: <***@davros.org>
Mobile: +44 7973 377646 | Web: <http://www.davros.org>
Please reply to the Reply-To address, which is: <***@davros.org>
Loren Pechtel
2013-08-06 21:49:39 UTC
Permalink
On Tue, 6 Aug 2013 04:50:12 -0700 (PDT), JohnFair
Post by JohnFair
I can see something like that being acceptable for fairly low grade traffic but you'd want control over the delivery vessel for high grade traffic. Even one time pads can be broken - it's just the multi millennial processing periods that don't make it worthwhile now. Given the substantial increases in computer power available to these people you might find it taking a few days if not hours running on high quality hard/software.
No, one time pads can't be broken, period, if they're done right.
pyotr filipivich
2013-06-24 20:27:06 UTC
Permalink
Post by John Fairhurst
Hi,
In the Honorverse books the transition between systems using
wormholes is done maually in so much as you still have someone on
the bridge doing the equivilent of flicking a switch in change
the configuration of the sails.
However, I can't see any reason why this can't be fully automated,
I can: "Needs of the Plot-line." I suspect a lot of what is going
on viz verbal orders spoken by captains is to inform the reader of
what is happening.
After all, what "happens" is that the captain says "execute the
maneuver" and the proper body pushes the correct button. But I can
see automating transitions and reconfigurations.
Post by John Fairhurst
especially given the Star Empire's increasing shortage of trained
staff.
I could also see an automated shuttle based on a courier
that could be used for transmitting high priority communications
through the wormhole.
They call them ... "Probes".
--
pyotr filipivich
You've read my sig file, that's enough Social Interaction for the Day.
Loading...