This Slashdot article, referring to the so called "World's First Formally-Proven OS Kernel",was brought to my attention by a colleague who is aware of my experience with formally proven OS' in the 1980's. What follows is my response to the claim of being first, and the value of proving the correctness of an OS:
I am aware of at least two instances of operating system kernels that were built in the late 1970's / early 1980's using formal proofs of correctness. I will talk about my experience with one of them.
One of them, produced in the late 1970's was a kernel designed for a specialized environment. This kernel/OS was a reasonably functional kernel complete with multiprocessing, time-sharing, file systems, etc. Unfortunately while the formal proof for this kernel was solid, the axiomatic set on which this formal proof was based did not perfectly match its operating environment. This mismatch proved to be fatal to the OS security.
This formally proven OS took years to create and prove its correctness. Those who developed and maintained the OS were very proud of their work. There were plans underway to create a commercial version of their work and to market it through a certain hardware vendor on which their OS ran.
When I was a student intern working for the organization where that developed this OS worked, I worked in their OS environment from time to time. I came in from the outside where my OS experience was with a wide variety OS' such as MVS, NOS, KRONOS, TOPS-10, RSTS/E, VMS, Multics, and Unix (5th/6th/7th edition). I had enough experience in jumping into new OS environments that I felt comfortable as a user in this one, even though it was unusual.
An important point to observe here is I was one of the first people who enter this OS environment from the outside. I was not steeped in the development world of the OS. I brought with me, ideas that differed from the OS developers. As a young student, I believed that the OS should work for me instead of getting in my way. To help come up to speed, I ported over my collection of OS-independent tools and soon began coding away on my assigned tasks.
Then one day, working within my OS-independent tools, something very odd happened. By mistake, I did something that produced an unusual result. I quickly realized that something was very wrong because the result was "impossible" according to the formal proof of OS correctness. Under the rules set down by my employer I immediately contacted the appropriate security officer and the next thing I knew, I was in a long sequence of meetings with people trying to figure out what in the hell happened.
In the first meeting after my mistake, I learned that I had been reassigned to a new team. I was assured that I was not being disciplined, far from it: I had made a number of people very happy and they moved paperwork mountains to move me into their team. This team was given a task of attempting to repeat my previous "mistake" as well as to discover if exploits that are more general were possible against this OS. We were assigned to work âoeundercoverâ as developers under test/QA installations using this OS. In the end, the team was successful in discovering a much more general form of the OS hole I accidentally found.
What went wrong with the OS formal proof? Well the mapping from the formal logic world to the real world of hardware, physics, people, and the environment was flawed. In other words, when you added in the "real-world", the proof was incomplete. Attempts by the OS developers to expand their proof to address the "real-world" only produced a set of inconsistent theorems. I believe the OS project was abandoned after the OS developers failed multiple times to expand their formal proof to deal with âoereal-worldâ environments.
During this experience I was introduced to two "Security Camps": One, "the absolutists" as they called themselves, included people who worked on this formally proven OS. The opposing camp called the
''If your environment does not permit the use of the following hardware components,
you must physically disable them...
Only an Apple Certified technician can physically disable these components without
voiding the warranty on your computer. A limited number of Apple Certified
technicians can remove preapproved components.
After an Apple Certified technician removes the component the technician logs a
special note with Apple Care, indicating that the computer has had a component
properly removed. Most components removed by Apple technicians can be reinstalled,
if needed.
Also, see your local Apple representative for more information.
Note: If you are in a government organization and need a letter of volatility for Apple
products, send your request to AppleFederal@apple.com.''
FYI: A similar action can be taken for hand held devices such as an Apple iPhone.
BTW: You can still use an external camera/microphone for services such as iChat on a MacBook where the built in devices have been removed. When permitted, plugging in an external camera/microphone will temporarily restore such capability. Moreover, by physically removing such external devices when they are not in use, you can better control them.:-)
So buy your MacBook, have a Apple Certified technician remove the offending components, and if needed get a letter of volatility. Q.E.D.
You said the laptop drive crashed. Perhaps the altitude combined with the "odd air" as you put it contributed to the drive crash? What was your "non-Windoz laptop" and did that suffer any problems?
I imagine the dry air made static discharge a problem. That must have made working with electronics a challenge.
BTW: Were you working on the lake drilling project or something related to ice-core research or what?
I agree your working conditions may not have been the worst, but from what you said it was one of the more odd coding conditions I've read about.
Regarding the engineering challenge: "Provide energy from fusion"
The Sun and other stars have been doing this for billion of years. On earth, H-Bombs did this decades ago. Heck, I've accelerated Deuterium ions into a target containing Tritium in a lab and calculated the energy that was released in the resulting fusion reaction.
I think they way to say something along the lines of: Produce power from commercially viable fusion reactors.
CookieCuller gives you the option to remove all non-protected cookies when firefox restarts. You choose which cookies, if any, that deserve protection.
I configure CookieCuller as follows:
Tools->Add-ons->CookieCuller->Preferences and check: [x] Delete Unprotected Cookies on Startup
View->Toolbars->Customize and Drag the CookieCuller icon (A chocolate chip cookie) to the toolbar panel
Click on the CookieCuller icon, select any cookies you which to protect (if any) and click Protect Cookie
Even if a site requires you to accept their long life cookies, on restart CookieCuller will toss your cookies!:-)
The article says: ''He managed to reach as high as 115 MPG...''
In a stock unmodified Honda Insight I was able to achieve a 114.9 MPH over round trip of 55 miles. The route involved a 1500 foot change in elevation over a pass and back down. Not a huge hill, but not exactly flat either. Time wise, I kept to the traffic flow which was mostly going at or over speed limit on the freeway, expressway and local streets.
''100+ MPG driving over some Colorado mountain passes''
I have managed 79 MPG on a round trip up and down an 8k change in elevation. Not 100 MPG, but I suspect it is some multiple of the mileage you achieved in that Oldsmobile Convertible.:-)
''Article also doesn't talk about this, but I bet for long trips, the savings in gas that is offset by the extended driving times ends up valueing their time at a pretty darn low $$$/hour.
Sure, I'm all for more fuel efficient cars and less fossil fuel burning, but there is a tradeoff.''
Careful driving can produce good mileage without requiring exotic car mods or time consuming driving.
On long trips I frequently go 800+ miles between fill ups, going at or above the speed limit, with an 80+ MPG.
FYI: My car's lifetime mileage (over 55k miles driving) is just under 70 MPG.
In the late 90's I was part of a team that was hacking Lava Lite (R) lamps. We built a system that converted SGI Indycam (digital camera) images of Lava Lites into cryptographic seeds of a pseudo-random number generator. And while classic lavarand has been replaced by LavaRnd (directly generating random numbers using lens capped webcams), it remains one of my favorite hacks / creations.
female mathematicians to add to the math geek list
on
Top Ten Geek Girls
·
· Score: 1
Sophie Germain and Emmy Noether were very significant mathematicians. Another significant mathematician to add to the long list of "female math geeks" is Emma Lehmer.
''... solar systems which contain planets like a 'Hot Jupiter' have a higher chance of also containing Earth-like planets.''
According to the new IAU definition of a Planet, an Earth-like planet must orbit our Sun.
Due the IAU's poor definition of a planet (one that restricts the term to only objects that orbit our Sun)
one cannot even refer to Earth-like planets around other stars because they cannot fit the IAU's new definition.
Finally, to improve the on topic-ness of this post, permit me to add this note about the
significance of idea, if it turns out to have merit:
Over the past 10 to 15 years, many astronomers assumed that one needed a Jupiter-like object
in a stable near-circular orbit, preferably beyond a "life forming zone",
before Earth-like objects could stay in the "life forming zone" long enough to form life.
This new idea suggests that such Earth-like objects can arise where there are Jupiter-like
objects orbiting very close to a star.
If their idea is correct, then this means that a number of the stars with known "things-formerly-known-as-extra-solar-planets"
containing a Hot Jupiter may also contain a stable Earth-like object in a "life forming zone".
BTW: This is not intended to be a complete history of the topic, I'm only pointing out a few highlights that others might find interesting.
A number of years ago, the question of a definition of a planet was raised as a result of discoveries of "planets" outside of our solar system as well as a growing number of Kuiper Belt Object (KBO) discoveries in our solar system. The IAU Division of Planetary Systems Sciences technical committee wisely chose to delay a decision on a definition until a more significant body of data was obtained.
In the mean time, a well meaning but widely misunderstood suggestion from an esteemed Astronomer suggested that the planet Pluto also be given a nice round minor planet number (i.e., reserve the next multiple of 10000). His intent was to recognize the special nature of Pluto as a large member of the KBO (Kuiper Belt Object) family. He never intended to demote Pluto from planet status. However, the press took the phrase "making Pluto a minor planet" and blew the controversy way out of proportion.
An executive committee recommendation on Planet definition was formed to draft a proposal for a definition of a planet. Minutes from the IAU executive committee indicated that they favored definitions that were based on measurable physical properties over arbitrary values. For example, they signaled that they were NOT inclined to look favorably on proposals such as "limit the number of 9 planets", or proposals that set an arbitrary minimum size of a Planet.
Last January at the AAS conference, an IAU liaison announced that the IAU executive committee was scheduled to produce a report on its recommendations just prior to the IAU 26th IAU General Assembly in Prague (Aug 14 to 25, 2006). The liaison recommended that any final comments and recommendations be submitted to the exectuive committee at least a month prior to the IAU general assembly.
I was part of a group that submitted a recommendation that the definition of a Planet encompass a requirement that "it must orbit a primary fuser with sufficient mass to deform it into an spheroidal / oblate spheroidal shape". We realized that our proposal could result in redefining several bodies as planets including the large asteroid Ceres. We proposed that a new sub-class of Planets could be defined (again based on measurable physical properties) to acuminate these new dwarf planets.
We were told that a number of other groups had submitted similar of very similar proposals. I have not examined the executive committee report in detail, however it appears that IAU executive committee agrees, in principle, with such proposals.
On Tuesday 2006 August 22, 12:45-13:45 (local Prague time), in Forum Hall, executive committee recommendation on Planet definition will be presented. Based on the unanimous recommendation of the executive committee, I am hopeful of a favorable outcome form the IAU General assembly.
Is it advised to switch GnuPG from SHA-1 for signatures to SHA-256
I recommended that:
Existing applications that use SHA1, where possible, should changed to use SHA256 before the end of 2008
So GnuPG should be modified to use SHA-256 before the end of 2008,
in my opinion. However I also said:
For interoperability with older applications and hardware, these applications may have to also support SHA1
Given the number of SHA1 PGP key signatures out there, I suspect that GnuPG
will have to support SHA1 for a long time to come.
Unless I'm mistaken, GnuPG and friends do not support SHA256 today.
Therefore even if you were to somehow sign your new PGP key using
the SHA256 hash, not many people would be able to process it.
So what needs to happen is that the developers / maintainers of
PGP / GnuPG code base need to extend the application and data formats
to allow for SHA256. As I stated:
Existing applications and protocols should be modified to be algorithm agile by the end of 2008, if not sooner
In the case of PGP / GnuPG and friends, the if not sooner is a must.
My paper talks about the general problem at a high level. It gives a summary of common opinions expressed at the NIST Cryptographic Hash conference. Moreover, it gives developers some specific cryptographic hash recommendations.
For the impatient, here is a summary for my recommendations for 2005-2006:
Avoid non-standard cryptographic hashes
Stop using MD5 now except for:
MD5 HMAC and MD5 hashed Passwords
Replace MD5 HMAC and MD5 hashed Passwords
with SHA256 or SHA1 before end of 2007
Existing applications that use SHA1, where possible,
should changed to use SHA256 before the end of 2008
For interoperability
with older applications and hardware, these applications may have to also support SHA1
If you must support both SHA256 and SHA1, take care so that
a "man in the middle" cannot inappropriately downgrade
Until a new Advanced Hash Standard (AHS)
is adopted, new applications and hardware should be
designed to use SHA256
For interoperability
with older applications and hardware, these applications
may have to also support SHA1
If you must support both SHA256 and SHA1, take care so that
a "man in the middle" cannot inappropriately downgrade
All new applications and protocols must be designed
to be algorithm agile
Existing
applications and protocols should be
modified to be algorithm agile by the end of
2008, if not sooner
SHA384 or SHA512 may be used in place of SHA256
in the above examples
Keep in mind that SHA384 and SHA512 are slower
and larger than SHA256 or SHA1
Because it is possible that SHA1 will become
unacceptably weak before 2008, and because
SHA256 may become vulnerable to attack before
Advanced Hash Standard (AHS) is adopted, a
defense in depth approach must be taken
FYI: a1d0c6e83f027327d8461063f4ac58a6 is the
ASCII hex MD5 hash of the ASCII string "42".
Therefore, if that string had been your hard drive,
then your hard drive would have been tracked
in near zero time.:-)
Back to the question: "How Long to Crack an
'Encrypted' HD?": it all depends on how
well it is done. It also depends on where the
disk key is stored. It is easier to crack a
drive if the key is kept on the drive or left
up to lazy humans to type in each time.
I'm not kidding about the last point. There
are hard drive encryption products where drive
is automatically mounted / accessed without
human intervention. These products derive
the decryption key from stored state on the
hard drive. Sure they pull tricks such as
storing the key material in a sector marked
as "bad", but if you reverse engineer their
process you can find the drive key and
begin cracking the drive in milliseconds.
There are hard drive encryption products where
a human must enter a password / pass-phrase
access the drive
decryption key. The time to crack the drive
depends on how easy to guess the unlocking
password / pass-phrase. This guessing can
be done in parallel starting with common /
poorly selected passwords / pass-phrases first.
Too many people don't want to type in
difficult / hard to type passwords.
A guessing attack would frequently be successful
against drives encrypted with products that require
a human to type something.
for a puzzle, try small or 1-line IOCCC winner
on
Brain Teasers for Coders?
·
· Score: 2, Interesting
If you are looking for puzzles (but NOT
coding style), try some of the International
Obfuscated C Code Contest winners:
Print out the
korn.c
source.
By just looking at it, ask them to tell
you what it does and why.
Note that this 1-liner has stumped some
people who have been coding in C for > 31
years.
So if they get stuck, help them out by asking
asking some questions such as:
In korn.c, what type of
symbol is unix?.
And:
Why does Korn subtract 0x60?
Hint: 'a' in ASCII is 0x61.
And:
On a non-Unix system, what will this program do?
p.s. We are in the middle of judging the 2005
entries. There are some good ones this year,
IMHO.
"Making the door/frame even a strong magnet is useless because the field strength falls off..."
You are correct.
That is why I believed that
if someone went to the effort to
construct such a door, then that
person probably lacked
access to someone who knew how to design
isolation data centers.
"I would just make my datacenter...
pulsing field"
Exactly.
A pulsed field is much
harder to shield against than a static field.
The very best (non-classified) shields fail
above 0.5T.
And they only approach shielding at 0.4T
under the most ideal (which includes
a constant field) conditions.
"Of course, I'd have to have encrypted fiber-optic Internet connections, etc..."
Yes which is why I suggested that there was
"... likely a number of other ways to get data into / out of your data center without going through that magnetic door.".
Their magnetic door seemed like it
was an example of a cute
paper door with a dead-bolt lock.
I'd worry about the digital equivalent of
paper walls long before I'd put a magnetic data
trap in place.
BTW: One "best practice" is to use
a magnetic pulse inside a man-trap
to detect the presence of
electronic equipment (drives, tapes,
anything electronics),
read-only storage (CDs, DVDs,
etc.), and shielded regions (mu-metal
boxes, etc.).
The intent is not to erase and destroy
but to detect, inspect and correct.
Mu-metal is a mixture containing two or more metallic elements or metallic and nonmetallic elements usually fused together or dissolving into each other when it is a molten-alloy. For example: 77% nickel, 15% iron and the rest usually copper and molybdenum. Mu-metal's main characteristic is that it is very magnetically soft.
Surrounding a room temperature region with a reasonable mu-metal can shield that region from magnetic fields up to, say, 2000 Oersteds (0.2 Tesla). An external magnetic field below that level cannot penetrate the mu-metal enclosed region very well.
Some mu-metal mixtures are better (> 4000 Oersteds = 0.4 Tesla effective shielding) and some are worse at shielding (< 1000 Oersteds = 0.1 tesla). There are non-mu-Metal magnetic field shielding methods as well.
I would not be surprised if a disk drive inside of a good mu-metal case can pass through your door without any problem.
On top of that counter-hack there is likely a number of other ways to get data into / out of your data center without going through that magnetic door.
If you are trying to build/run a high security isolation data center then I'd recommend that you consult with someone who specializes in van Eck / Tempest security. Not only will they me able to assist you with building a good trap, they will be able help create an environment that can protect your data against a wide variety of other threats as well. Isolation data centers are a speciality. Unless you have that skill, or access to someone who does, all that you wind up with is a
dead-bolt on a paper door
that only gives you the illusion of security.
As
rjh
also pointed out, the ANSI-X9.17 pseudo-random number
generator (a 3-DES based PRNG) is a
high
quality PRNG.
So if you lack a good hardware random source,
or if your hardware random source cannot deliver
quality values fast enough for your purposes,
then the ANSI-X9.17 might be your next
best choice.
>... digits from pi are indeed an acceptable source of randomness...
Pi in base Pi is not a good source of random digits.
In particular, Pi in base Pi is:
10.00000000000000000000000000000...
I suspect that random sequence quality level
sited in the paper would apply to digits of Pi in
any non-transcendental base: such as base 10.
Many transcendental bases, such as base e,
should exhibit the a similar quality level to
that sighted in the paper.
However I suspect that any base that was a
result of a polynomial function of Pi might not
do so well.
We started seeing this type of spam back in June of 2004. In our case the referrer spam was attempting to get webalizer to create links in the "top N referrer" table back to their pron sites.
Our initial attempt to solve this was to complain to the ISP of the referrer spammers. That did no good. The ISP was willing to listen, but not to act.
We did manage to actually track down the jerks who were doing the referrer spam. They told us that they were attempting to create links back to their sites for better search engine placement.
Our work-a-round was two fold.
For various reasons we wanted to keep these our
webalizer stats externally accessible.
So we requested bots (the ones
that follow the rules at least) to not index our
external stats and we modified webalizer to
not form links back to the referrers.
We edited our robots.txt file to exclude legit bots from our stats:
User-agent: *
Disallow:/stats
We also
patched
webalizer v2.01-10 to no longer form
URLs to referrers.
Now only a plain text line without the leading http:// shows up in the table.
The original referrer spammers gave up when
they lost off the the links back to their sites.
The bottom of the 0.basic.patch prevents
webalizer
from forming links back to referrers.
See
README-FIRST
for details on this patch set.
The Orbit
of 2001 DA42 (minor planet
number 25924) may be viewed online.
Just type in 2001 DA42 and click SEARCH to see the orbit parameters.
Then click Show Orbit (Java required).
The asteroid in inclined about 1.73 degrees
from Earth's orbit. It lines outside of the
orbit of Mars with a Semimajor axis of
about 2.41 AU.
As of 26 Jan 2006 12:30 UTC, it was located
approximately:
RA 12:16:18
Dec -3:20:19
(which in the constellation Virgo: about
3 degrees due south of the 4th magnitude
star Zaniah;
about 14 degrees away (and somewhat west) from Jupiter;
about 19 degrees away (and also somewhat west) from the 1st magnitude
star Spica)
2001 DA42 is currently a very dim object:
with a
apparent
visual magnitude of approximately 20.4.
That is about 360 times fainter than Pluto.
You will not be able to visually
see 2001 DA42
with your typical "bark yard telescope".
Around March 2005 DA42
will peak at about magnitude 19.
Around May 2009 it will peak at about 18.5.
Around June 2013 it will peak at about 18.0.
That is about as bright as 2001 DA42 typically gets.
In regards to the
parent
posting and Larry Wall:
The two awards given to
Larry
Wall and his code are available.
His
1986
grand prize is particularly
noteworthy.
Note that I am neither confirming nor
denying that the anonymous winner is Larry.
I just highlighting those two awards and
the 1986 entry in particular.
Slashdot Mirror
This Slashdot article, referring to the so called "World's First Formally-Proven OS Kernel",was brought to my attention by a colleague who is aware of my experience with formally proven OS' in the 1980's. What follows is my response to the claim of being first, and the value of proving the correctness of an OS:
I am aware of at least two instances of operating system kernels that were built in the late 1970's / early 1980's using formal proofs of correctness. I will talk about my experience with one of them.
One of them, produced in the late 1970's was a kernel designed for a specialized environment. This kernel/OS was a reasonably functional kernel complete with multiprocessing, time-sharing, file systems, etc. Unfortunately while the formal proof for this kernel was solid, the axiomatic set on which this formal proof was based did not perfectly match its operating environment. This mismatch proved to be fatal to the OS security.
This formally proven OS took years to create and prove its correctness. Those who developed and maintained the OS were very proud of their work. There were plans underway to create a commercial version of their work and to market it through a certain hardware vendor on which their OS ran.
When I was a student intern working for the organization where that developed this OS worked, I worked in their OS environment from time to time. I came in from the outside where my OS experience was with a wide variety OS' such as MVS, NOS, KRONOS, TOPS-10, RSTS/E, VMS, Multics, and Unix (5th/6th/7th edition). I had enough experience in jumping into new OS environments that I felt comfortable as a user in this one, even though it was unusual.
An important point to observe here is I was one of the first people who enter this OS environment from the outside. I was not steeped in the development world of the OS. I brought with me, ideas that differed from the OS developers. As a young student, I believed that the OS should work for me instead of getting in my way. To help come up to speed, I ported over my collection of OS-independent tools and soon began coding away on my assigned tasks.
Then one day, working within my OS-independent tools, something very odd happened. By mistake, I did something that produced an unusual result. I quickly realized that something was very wrong because the result was "impossible" according to the formal proof of OS correctness. Under the rules set down by my employer I immediately contacted the appropriate security officer and the next thing I knew, I was in a long sequence of meetings with people trying to figure out what in the hell happened.
In the first meeting after my mistake, I learned that I had been reassigned to a new team. I was assured that I was not being disciplined, far from it: I had made a number of people very happy and they moved paperwork mountains to move me into their team. This team was given a task of attempting to repeat my previous "mistake" as well as to discover if exploits that are more general were possible against this OS. We were assigned to work âoeundercoverâ as developers under test/QA installations using this OS. In the end, the team was successful in discovering a much more general form of the OS hole I accidentally found.
What went wrong with the OS formal proof? Well the mapping from the formal logic world to the real world of hardware, physics, people, and the environment was flawed. In other words, when you added in the "real-world", the proof was incomplete. Attempts by the OS developers to expand their proof to address the "real-world" only produced a set of inconsistent theorems. I believe the OS project was abandoned after the OS developers failed multiple times to expand their formal proof to deal with âoereal-worldâ environments.
During this experience I was introduced to two "Security Camps": One, "the absolutists" as they called themselves, included people who worked on this formally proven OS. The opposing camp called the
The dashed form of the English name is available at assist those who might actually want to read all or part of the +324 Megabyte name. :-)
FYI: A similar action can be taken for hand held devices such as an Apple iPhone.
BTW: You can still use an external camera/microphone for services such as iChat on a MacBook where the built in devices have been removed. When permitted, plugging in an external camera/microphone will temporarily restore such capability. Moreover, by physically removing such external devices when they are not in use, you can better control them. :-)
So buy your MacBook, have a Apple Certified technician remove the offending components, and if needed get a letter of volatility. Q.E.D.
You said the laptop drive crashed. Perhaps the altitude combined with the "odd air" as you put it contributed to the drive crash? What was your "non-Windoz laptop" and did that suffer any problems?
I imagine the dry air made static discharge a problem. That must have made working with electronics a challenge.
BTW: Were you working on the lake drilling project or something related to ice-core research or what?
I agree your working conditions may not have been the worst, but from what you said it was one of the more odd coding conditions I've read about.
Regarding the engineering challenge: "Provide energy from fusion"
The Sun and other stars have been doing this for billion of years. On earth, H-Bombs did this decades ago. Heck, I've accelerated Deuterium ions into a target containing Tritium in a lab and calculated the energy that was released in the resulting fusion reaction.
I think they way to say something along the lines of: Produce power from commercially viable fusion reactors.
- 4*MPAA^4 + 885876010*MPAA^3 + 221652134*MPAA^2 + 785088213*MPAA + 1217422679
- int(MPAA^4.07357869332201118231406118925401369491
7 4402)
- int(sinh(4.10661181245779322741624495053040769792
3 5073*ln(MPAA)))
- int(exp(64.49413380395982197152851466623071867219
5 108554589) * MPAA)
- MPAA * 573005406241 * 17835599295426313 + 11 * 73 * 1516093
- MPAA + 224682693016770468661336710171264397 * 59
- etc.
protected under the freedom of expression?- 2 * 2 * 2 * 2 * 2 * 2 * 5 * 19 * 12043 * 216493 * 836256503069278983442067
- int(exp(85.47753466631182170634469303376305403678
0 258841))
- int(tan(1.570796326794896619231321691639751442023
1 48749037327327204052898028418844448835))
- int(cosh(86.1706818468717670157619251552212306048
5 5759))
- 1657034860998682206377358237675213082^2 - 1657034860998682206377358237675213078^2
- 828517430499341103188679118837606544^2 - 828517430499341103188679118837606536^2
- 414258715249670551594339559418803278^2 - 414258715249670551594339559418803262^2
- 331406972199736441275471647535042626^2 - 331406972199736441275471647535042606^2
- 207129357624835275797169779709401651^2 - 207129357624835275797169779709401619^2
- 165703486099868220637735823767521328^2 - 165703486099868220637735823767521288^2
- 87212361105193800335650433561853358^2 - 87212361105193800335650433561853282^2
- 82851743049934110318867911883760694^2 - 82851743049934110318867911883760614^2
- 43606180552596900167825216780926736^2 - 43606180552596900167825216780926584^2
- 41425871524967055159433955941880407^2 - 41425871524967055159433955941880247^2
- 21803090276298450083912608390463482^2 - 21803090276298450083912608390463178^2
- 17442472221038760067130086712370854^2 - 17442472221038760067130086712370474^2
- 10901545138149225041956304195231969^2 - 10901545138149225041956304195231361^2
- 8721236110519380033565043356185712^2 - 8721236110519380033565043356184952^2
- 4360618055259690016782521678093426^2 - 4360618055259690016782521678091906^2
- 2180309027629845008391260839047853^2 - 2180309027629845008391260839044813^2
- 137593196130422835371365792407646^2 - 137593196130422835371365792359474^2
- 68796598065211417685682896239952^2 - 68796598065211417685682896143608^2
- 34398299032605708842841448192234^2 - 34398299032605708842841447999546^2
- 27518639226084567074273158597142^2 - 27518639226084567074273158356282^2
- 17199149516302854421420724240633^2 - 17199149516302854421420723855257^2
- 13759319613042283537136579479216^2 - 13759319613042283537136578997496^2
- 7653988170512128366170538222546^2 - 7653988170512128366170537356574^2
- 7241747164759096598492936898874^2 - 7241747164759096598492935983606^2
- 6879659806521141768568290100898^2 - 6879659806521141768568289137458^2
- 3826994085256064183085269760752^2 - 3826994085256064183085268028808^2
- 3620873582379548299246469135888^2 - 3620873582379548299246467305352^2
- 3439829903260570884284145773029^2 - 3439829903260570884284143846149^2
- 1913497042628032091542636179334^2 - 1913497042628032091542632715446^2
- 1810436791189774149623235940846^2 - 1810436791189774149623232279774^2
- 1530797634102425673234109722842^2 - 1530797634102425673234105392982^2
- 1448349432951819319698589576418^2 - 1448349432951819319698585000078^2
etc.I recommend the CookieCuller firefox plugin to control cookies:
CookieCuller gives you the option to remove all non-protected cookies when firefox restarts. You choose which cookies, if any, that deserve protection.
I configure CookieCuller as follows:
Even if a site requires you to accept their long life cookies, on restart CookieCuller will toss your cookies! :-)
In a stock unmodified Honda Insight I was able to achieve a 114.9 MPH over round trip of 55 miles. The route involved a 1500 foot change in elevation over a pass and back down. Not a huge hill, but not exactly flat either. Time wise, I kept to the traffic flow which was mostly going at or over speed limit on the freeway, expressway and local streets.
''100+ MPG driving over some Colorado mountain passes''
I have managed 79 MPG on a round trip up and down an 8k change in elevation. Not 100 MPG, but I suspect it is some multiple of the mileage you achieved in that Oldsmobile Convertible. :-)
''Article also doesn't talk about this, but I bet for long trips, the savings in gas that is offset by the extended driving times ends up valueing their time at a pretty darn low $$$/hour.
Sure, I'm all for more fuel efficient cars and less fossil fuel burning, but there is a tradeoff.''
Careful driving can produce good mileage without requiring exotic car mods or time consuming driving. On long trips I frequently go 800+ miles between fill ups, going at or above the speed limit, with an 80+ MPG.
FYI: My car's lifetime mileage (over 55k miles driving) is just under 70 MPG.
In the late 90's I was part of a team that was hacking Lava Lite (R) lamps. We built a system that converted SGI Indycam (digital camera) images of Lava Lites into cryptographic seeds of a pseudo-random number generator. And while classic lavarand has been replaced by LavaRnd (directly generating random numbers using lens capped webcams), it remains one of my favorite hacks / creations.
Sophie Germain and Emmy Noether were very significant mathematicians. Another significant mathematician to add to the long list of "female math geeks" is Emma Lehmer.
According to the new IAU definition of a Planet, an Earth-like planet must orbit our Sun. Due the IAU's poor definition of a planet (one that restricts the term to only objects that orbit our Sun) one cannot even refer to Earth-like planets around other stars because they cannot fit the IAU's new definition.
I wish this was a troll post, but it is not. It is am example of the problems with the recent IAU resolution. It is the reason why I joined a number of astronomers and signed the Petition Protesting the IAU Planet Definition.
Finally, to improve the on topic-ness of this post, permit me to add this note about the significance of idea, if it turns out to have merit:
A number of years ago, the question of a definition of a planet was raised as a result of discoveries of "planets" outside of our solar system as well as a growing number of Kuiper Belt Object (KBO) discoveries in our solar system. The IAU Division of Planetary Systems Sciences technical committee wisely chose to delay a decision on a definition until a more significant body of data was obtained.
In the mean time, a well meaning but widely misunderstood suggestion from an esteemed Astronomer suggested that the planet Pluto also be given a nice round minor planet number (i.e., reserve the next multiple of 10000). His intent was to recognize the special nature of Pluto as a large member of the KBO (Kuiper Belt Object) family. He never intended to demote Pluto from planet status. However, the press took the phrase "making Pluto a minor planet" and blew the controversy way out of proportion.
An executive committee recommendation on Planet definition was formed to draft a proposal for a definition of a planet. Minutes from the IAU executive committee indicated that they favored definitions that were based on measurable physical properties over arbitrary values. For example, they signaled that they were NOT inclined to look favorably on proposals such as "limit the number of 9 planets", or proposals that set an arbitrary minimum size of a Planet.
Last January at the AAS conference, an IAU liaison announced that the IAU executive committee was scheduled to produce a report on its recommendations just prior to the IAU 26th IAU General Assembly in Prague (Aug 14 to 25, 2006). The liaison recommended that any final comments and recommendations be submitted to the exectuive committee at least a month prior to the IAU general assembly.
I was part of a group that submitted a recommendation that the definition of a Planet encompass a requirement that "it must orbit a primary fuser with sufficient mass to deform it into an spheroidal / oblate spheroidal shape". We realized that our proposal could result in redefining several bodies as planets including the large asteroid Ceres. We proposed that a new sub-class of Planets could be defined (again based on measurable physical properties) to acuminate these new dwarf planets.
We were told that a number of other groups had submitted similar of very similar proposals. I have not examined the executive committee report in detail, however it appears that IAU executive committee agrees, in principle, with such proposals.
On Tuesday 2006 August 22, 12:45-13:45 (local Prague time), in Forum Hall, executive committee recommendation on Planet definition will be presented. Based on the unanimous recommendation of the executive committee, I am hopeful of a favorable outcome form the IAU General assembly.
I recommended that:
So GnuPG should be modified to use SHA-256 before the end of 2008, in my opinion. However I also said:
Given the number of SHA1 PGP key signatures out there, I suspect that GnuPG will have to support SHA1 for a long time to come.
Unless I'm mistaken, GnuPG and friends do not support SHA256 today. Therefore even if you were to somehow sign your new PGP key using the SHA256 hash, not many people would be able to process it. So what needs to happen is that the developers / maintainers of PGP / GnuPG code base need to extend the application and data formats to allow for SHA256. As I stated:
In the case of PGP / GnuPG and friends, the if not sooner is a must.
For the impatient, here is a summary for my recommendations for 2005-2006:
See the paper for mode details.
Back to the question: "How Long to Crack an 'Encrypted' HD?": it all depends on how well it is done. It also depends on where the disk key is stored. It is easier to crack a drive if the key is kept on the drive or left up to lazy humans to type in each time.
I'm not kidding about the last point. There are hard drive encryption products where drive is automatically mounted / accessed without human intervention. These products derive the decryption key from stored state on the hard drive. Sure they pull tricks such as storing the key material in a sector marked as "bad", but if you reverse engineer their process you can find the drive key and begin cracking the drive in milliseconds.
There are hard drive encryption products where a human must enter a password / pass-phrase access the drive decryption key. The time to crack the drive depends on how easy to guess the unlocking password / pass-phrase. This guessing can be done in parallel starting with common / poorly selected passwords / pass-phrases first. Too many people don't want to type in difficult / hard to type passwords. A guessing attack would frequently be successful against drives encrypted with products that require a human to type something.
I'd pick some of the short programs and 1-liner winners.
My favorite 1-line winner was submitted by David Korn (of ksh fame):
Print out the korn.c source. By just looking at it, ask them to tell you what it does and why.
Note that this 1-liner has stumped some people who have been coding in C for > 31 years. So if they get stuck, help them out by asking asking some questions such as:
And: And:p.s. We are in the middle of judging the 2005 entries. There are some good ones this year, IMHO.
You are correct. That is why I believed that if someone went to the effort to construct such a door, then that person probably lacked access to someone who knew how to design isolation data centers.
" I would just make my datacenter ...
pulsing field "
Exactly. A pulsed field is much harder to shield against than a static field. The very best (non-classified) shields fail above 0.5T. And they only approach shielding at 0.4T under the most ideal (which includes a constant field) conditions.
" Of course, I'd have to have encrypted fiber-optic Internet connections, etc... "
Yes which is why I suggested that there was "... likely a number of other ways to get data into / out of your data center without going through that magnetic door.". Their magnetic door seemed like it was an example of a cute paper door with a dead-bolt lock. I'd worry about the digital equivalent of paper walls long before I'd put a magnetic data trap in place.
BTW: One "best practice" is to use a magnetic pulse inside a man-trap to detect the presence of electronic equipment (drives, tapes, anything electronics), read-only storage (CDs, DVDs, etc.), and shielded regions (mu-metal boxes, etc.). The intent is not to erase and destroy but to detect, inspect and correct.
On top of that counter-hack there is likely a number of other ways to get data into / out of your data center without going through that magnetic door.
If you are trying to build/run a high security isolation data center then I'd recommend that you consult with someone who specializes in van Eck / Tempest security. Not only will they me able to assist you with building a good trap, they will be able help create an environment that can protect your data against a wide variety of other threats as well. Isolation data centers are a speciality. Unless you have that skill, or access to someone who does, all that you wind up with is a dead-bolt on a paper door that only gives you the illusion of security.
IAAICS (I am an isolation chamber specialist).
As rjh also pointed out, the ANSI-X9.17 pseudo-random number generator (a 3-DES based PRNG) is a high quality PRNG. So if you lack a good hardware random source, or if your hardware random source cannot deliver quality values fast enough for your purposes, then the ANSI-X9.17 might be your next best choice.
Pi in base Pi is not a good source of random digits. In particular, Pi in base Pi is:
I suspect that random sequence quality level sited in the paper would apply to digits of Pi in any non-transcendental base: such as base 10. Many transcendental bases, such as base e, should exhibit the a similar quality level to that sighted in the paper. However I suspect that any base that was a result of a polynomial function of Pi might not do so well.
Our initial attempt to solve this was to complain to the ISP of the referrer spammers. That did no good. The ISP was willing to listen, but not to act.
We did manage to actually track down the jerks who were doing the referrer spam. They told us that they were attempting to create links back to their sites for better search engine placement.
Our work-a-round was two fold. For various reasons we wanted to keep these our webalizer stats externally accessible. So we requested bots (the ones that follow the rules at least) to not index our external stats and we modified webalizer to not form links back to the referrers.
We edited our robots.txt file to exclude legit bots from our stats:
We also patched webalizer v2.01-10 to no longer form URLs to referrers. Now only a plain text line without the leading http:// shows up in the table. The original referrer spammers gave up when they lost off the the links back to their sites.
The bottom of the 0.basic.patch prevents webalizer from forming links back to referrers. See README-FIRST for details on this patch set.
The current distance from Earth (as of 25 Jan 2005) is about 1.734 AU (equals about 259 402 932 kilometers or 161 185 509 miles).
The asteroid in inclined about 1.73 degrees from Earth's orbit. It lines outside of the orbit of Mars with a Semimajor axis of about 2.41 AU. As of 26 Jan 2006 12:30 UTC, it was located approximately:
2001 DA42 is currently a very dim object: with a apparent visual magnitude of approximately 20.4. That is about 360 times fainter than Pluto. You will not be able to visually see 2001 DA42 with your typical "bark yard telescope".
Around March 2005 DA42 will peak at about magnitude 19. Around May 2009 it will peak at about 18.5. Around June 2013 it will peak at about 18.0. That is about as bright as 2001 DA42 typically gets.
Note that I am neither confirming nor denying that the anonymous winner is Larry. I just highlighting those two awards and the 1986 entry in particular.