This reminds me of the original 'Log Filesystem' research in the 80s, back when drive geometry was known to the OS, and the OS could take steps to optimize for it.
The Log Filesystem concept was to write all data sequentially to disk, and update metadata during idle times.
The basic research influenced a number of filesystems, such as NetApp's WAFL, Sun's ZFS, Linux's JFFS2, etc.
Interesting to see the concept implemented 'in hardware'.
I think there's plenty of oxygen. The trick is to find a way to extract the O2 from the water ice (hopefully there is a percentage of O2 that is simply dissolved in the ice instead of having to use electrolysis) to make it advantageous to burn it with the methane for power.
The equation should end up as the following for a viable mission using methane as a fuel:
Heat cost of producing O2 < Heat generated by burning O2 with atmospheric methane.
1) Plants are grown for their sugars to feed into this process (carbon sequestered from atmosphere into the roots left in the soil after harvesting, and into animals that the resulting cellulose will be fed to after sugar extraction). 2) Sugar (carbon) from step (1) is fed to the modified E. Coli to make gasoline. "Dead" E. Coli sludge could be used in animal feed, or processed into fertilizer for (1). 3) Gasoline from step (2) is then used for cars/trucks.
The problem is not "gasoline as a fuel", but the "extract carbon from natural sequestration to GET gasoline" that is the issue.
As an energy storage system, gasoline is REALLY hard to beat.
For its energy density, it is one of the safest energy storage systems we have.
Once upon a time (circa 2000), I wrote a Linux driver for the damned thing, fully supported by SanDisk.
It's Made of Evil. You need special software to be able to ready/write it (since it only has the read/write area that contacts the head on side 0, track 0), only transfers at 150K bits/sec (MFM encoded), and uses obsolete memory technologies.
If anyone has one of these crusty things, I might have a version of mtdtools that will work with it laying around somewhere.
If you want a copy, email jason dot mcmullan the-at-sign google's mail dot com
The Roku vidio player is an excellent example of security through "meh". It's almost an ideal box for a Boxee or MythTv frontend, but it is pretty much unhackable (cryptographically signed u-boot, kernel, and ramdisk). They've released their sources (but not their crypto key) months ago, yet not one single crack is available for it.
Why? Because (a) they don't make a big deal of the security features to the public, b) it's stupid cheap ($99 USD), and (c) It Just Works.
The combination of all three make 'meh'. Due to (a) there is no implicit challenge to the security community, (b) trumped the TiVo problem of trying to get 'more value for your money' out of an expensive piece of kit, and (c) prevents your Average Joe hacker from wanting to break a working (and useful to him) device.
Good counterexamples are TiVo, Linksys routers, and the Wii.
For TiVo, it was expensive enough that people wanted to get more value for their money, and felt it was time well spent to hack it.
With Linksys routers, It just Doesn't Work caused people to spend a lot of time finding a way to make some perfectly good equipment work at all for them.
The Wii advertised to the community that it was unhackable, which promptly cause all manner of security professionals to take up arms and figure out how to hack it.
In 'Medium' mode, splicers that have a free loot slot of their "top three", can pick up loot from the game world.
In 'Hard' mode, splicers can also pick up and change out weapons, except for the Spider Splicers.
Vita Chambers:
In 'Easy' mode, leave them as-is.
In 'Medium' mode, you drop all your loot, except for your Adam, and all your plasmids go back into the Gene Bank. Heath is 100%, Eve is 20% You need to either get new stuff, or go back to your dead body and scavenge.
In 'Hard' mode, in addition to the above, require a 'lock on to only you' hack before they're usable. Until then, all Splicers will get Vita Chambered. (I mean, story wise, why don't all the splicers get Vita Chambered when you kill them?) Also, consistent with 'Smart Splicers' behavior above, Splicers still on the level are free to grab your loot.
Quests:
No changes.
Little Sisters:
In 'Hard' mode only, Little Sisters can be killed easily while they are harvesting Adam from a corpse - makes surprise attacks on Big Daddies much harder if they are close to a Sister!
What would be really funny is if the reason for the delay is that they're trying to port Spore from Linux - and they can't get the damned thing to work on Windows!
No matter what you do, without *one original message*, you cannot
reconstruct the originals. (Now, this is cool for things like one
time pad encryption, but it won't work for network coding).
However, if you *do* have an original message, they all fall out
(assume A^B, B^C, and you have, say, 'A')
A^B ^ A = B (that's one!)
B^C ^ B = C (that's the other!)
So to reconstruct all three messages ('A', 'B', and 'C')
we need three messages - two XORed, and one original.
On the other hand, I have taken my mother's advice: "Find what you would gladly do for free, then find some sucker to pay you for it."
I love engineering, and I'd be happy if I was paid ditchdigger's wages. (And yes, I have a wife and kid, and I can afford my my house at a lower salary - I live in Pittsburgh!)
I used to work for LinuxCare, from January 2000 to Sept 2003. I have to say, to was a wild ride.
At the 'LinuxCare' phase, I mostly did contract work to write Linux device drivers for 3rd parties. (Including some absolutely evil stuff like a C++ stub for kernel modules, and a 'look like NT' wrapper for a MPEG encoder kernel module.)
In early 2000, we moved into our 'new' offices (we took up the entire basement of the huge converted warehouse building we were in), and had 'The Worlds Ugliest Mural' done by a local graffiti artist. The entire floor was carpeded with the LinuxCare 'X' logo. Yes, custom logo carpet.
Around 2001, the support business collapsed. The Founders left, except for Art, but we picked up a new CEO, some really smart IBM guys, and started working on what was to be the Levanta project. Originally targeted for IBM z/390 mainframes, it used the z/VM operating system to provide multiple 'on-demand' Linux-on-390 'partitions'. (z/VM is the mainframe equivalent to VMWare, but 20 years old!)
Akmal Khan came on board after Levanta was in full swing, and immediately took a dislike to the the distributed nature of our development group. There was Pittsburgh, doing the primary backend database; Ottawa was doing the web GUI and z/VM interface; Las Vegas handled the web infrastructure; project management in Atlanta; and San Francisco was sales and marketing. Except for SF and Ottawa, most sites telecommuted, so no 'office overhead' for those areas.
It became apparent pretty quickly that Akmal was the micromanaging type. By spring 2003, A.K. had collected his own group of technical people (very good ones, by the way) in SF, diverted all development of 'Levanta-on-Intel' to SF, and started making it pretty clear to the managers that all sites except SF would be going away.
That fall of 2003, the axe arrived for Ottawa, and I walked away from Levanta and the political mess that had developed.
I'm glad to have worked for LinuxCare, and had a ton-of-fun working on Levanta-on-z/390.
I used to work for LinuxCare, from January 2000 to Sept 2003. I have to say, to was a wild ride.
At the 'LinuxCare' phase, I mostly did contract work to write Linux device drivers for 3rd parties. (Including some absolutely evil stuff like a C++ stub for kernel modules, and a 'look like NT' wrapper for a MPEG encoder kernel module.)
In early 2000, we moved into our 'new' offices (we took up the entire basement of the huge converted warehouse building we were in), and had 'The Worlds Ugliest Mural' done by a local graffiti artist. The entire floor was carpeded with the LinuxCare 'X' logo. Yes, custom logo carpet.
Around 2001, the support business collapsed. The Founders left, except for Art, but we picked up a new CEO, some really smart IBM guys, and started working on what was to be the Levanta project. Originally targeted for IBM z/390 mainframes, it used the z/VM operating system to provide multiple 'on-demand' Linux-on-390 'partitions'. (z/VM is the mainframe equivalent to VMWare, but 20 years old!)
Akmal Khan came on board after Levanta was in full swing, and immediately took a dislike to the the distributed nature of our development group. There was Pittsburgh, doing the primary backend database; Ottawa was doing the web GUI and z/VM interface; Las Vegas handled the web infrastructure; project management in Atlanta; and San Francisco was sales and marketing. Except for SF and Ottawa, most sites telecommuted, so no 'office overhead' for those areas.
It became apparent pretty quickly that Akmal was the micromanaging type. By spring 2003, A.K. had collected his own group of technical people (very good ones, by the way) in SF, diverted all development of 'Levanta-on-Intel' to SF, and started making it pretty clear to the managers that all sites except SF would be going away.
That fall of 2003, the axe arrived for Ottawa, and I walked away from Levanta and the political mess that had developed.
I'm glad to have worked for LinuxCare, and had a ton-of-fun working on Levanta-on-z/390.
Slashdot Mirror
This reminds me of the original 'Log Filesystem' research in the 80s, back when drive geometry was known to the OS, and the OS could take steps to optimize for it.
The Log Filesystem concept was to write all data sequentially to disk, and update metadata during idle times.
The basic research influenced a number of filesystems, such as NetApp's WAFL, Sun's ZFS, Linux's JFFS2, etc.
Interesting to see the concept implemented 'in hardware'.
I think there's plenty of oxygen. The trick is to find a way to extract the O2 from the water ice (hopefully there is a percentage of O2 that is simply dissolved in the ice instead of having to use electrolysis) to make it advantageous to burn it with the methane for power.
The equation should end up as the following for a viable mission using methane as a fuel:
Heat cost of producing O2 < Heat generated by burning O2 with atmospheric methane.
Actually, this will end up as a carbon sink:
1) Plants are grown for their sugars to feed into this process (carbon sequestered from atmosphere into the roots left in the soil after harvesting, and into animals that the resulting cellulose will be fed to after sugar extraction).
2) Sugar (carbon) from step (1) is fed to the modified E. Coli to make gasoline. "Dead" E. Coli sludge could be used in animal feed, or processed into fertilizer for (1).
3) Gasoline from step (2) is then used for cars/trucks.
The problem is not "gasoline as a fuel", but the "extract carbon from natural sequestration to GET gasoline" that is the issue.
As an energy storage system, gasoline is REALLY hard to beat.
For its energy density, it is one of the safest energy storage systems we have.
http://en.wikipedia.org/wiki/Energy_density
For more details on 'step e', where the fatty aldehyde is converted to the gasoline alkanes, look up the "cer1 enzyme": http://www.ncbi.nlm.nih.gov/pmc/articles/PMC161066/
This enzyme is used by plants to create waxy coating on their leaves.
I find this to be a great example of what can happen when people are looking cross-disciple for solutions!
Like anything else, you can just buy it over the internet:
http://www.millipore.com/catalogue/item/554726-25MG?cid=bios-C-epdf-1032-1302-RC
It'll definitely get better pr0n results than 'SELECT ...'....
Of course, this is also why the IBM system only has '+' and '*' as operators.
Which severely limits the operations you can do on the encrypted system (so long as you aren't given E(-1) as an input).
If given E(-1) as input, you can recover all the integers again, and you can do any arithmetic set of operations.
I think testing for equality is strictly verboten under their system.
E(0) is E(a) - E(a)
E(1) is obtained by E(a)/E(a).
E(-1) is E(0) - E(1)
E(2) = E(1) + E(1) ...
E(n) = E(n-1) + E(1)
E(-n) = E(-(n-1)) + E(-1)
And there you go. All the integers - from one encrypted input.
And you can even recover your cyphertext *if* the homomorphic set has a 'is equal to' or a 'is not equal to' operator.
for (i = 0, ei = E(a)-E(a); ei is_equal E(a); i=i+1, ei = ei plus E(a) div E(a));
At the end of the loop, i == a.
My company just hired a guy in his 50s for a Software engineer position.
Move to Pittsburgh! Lots of tech companies here are starving for high-experience engineers!
Yes, yes we will.
mSATA is physically, but not electrically, compatible with miniPCIe slots.
It will fit, but will probably cause your system to catch on fire.
Have fun!
I've done some research on mSATA adapters:
http://tinyurl.com/682ehsd
What I find hilarious is that the mSATA is physically identical to a PCIe card edge, but is not electrically identical.
I wonder how many returns they are going to get on these.
Once upon a time (circa 2000), I wrote a Linux driver for the damned thing, fully supported by SanDisk.
It's Made of Evil. You need special software to be able to ready/write it (since it only has the read/write area that contacts the head on side 0, track 0), only transfers at 150K bits/sec (MFM encoded), and uses obsolete memory technologies.
If anyone has one of these crusty things, I might have a version of mtdtools that will work with it laying around somewhere.
If you want a copy, email jason dot mcmullan the-at-sign google's mail dot com
The Roku vidio player is an excellent example of security through "meh". It's almost an ideal box for a Boxee or MythTv frontend, but it is pretty much unhackable (cryptographically signed u-boot, kernel, and ramdisk). They've released their sources (but not their crypto key) months ago, yet not one single crack is available for it.
Why? Because (a) they don't make a big deal of the security features to the public, b) it's stupid cheap ($99 USD), and (c) It Just Works.
The combination of all three make 'meh'. Due to (a) there is no implicit challenge to the security community, (b) trumped the TiVo problem of trying to get 'more value for your money' out of an expensive piece of kit, and (c) prevents your Average Joe hacker from wanting to break a working (and useful to him) device.
Good counterexamples are TiVo, Linksys routers, and the Wii.
For TiVo, it was expensive enough that people wanted to get more value for their money, and felt it was time well spent to hack it.
With Linksys routers, It just Doesn't Work caused people to spend a lot of time finding a way to make some perfectly good equipment work at all for them.
The Wii advertised to the community that it was unhackable, which promptly cause all manner of security professionals to take up arms and figure out how to hack it.
Parts: the Clonus Horror
http://en.wikipedia.org/wiki/Parts:_The_Clonus_Horror
Great movie idea. Horrible execution.
Things I'd Change:
Splicers:
In 'Easy' mode, no changes.
In 'Medium' mode, splicers that have a free loot slot of their "top three", can pick up loot from the game world.
In 'Hard' mode, splicers can also pick up and change out weapons, except for the Spider Splicers.
Vita Chambers:
In 'Easy' mode, leave them as-is.
In 'Medium' mode, you drop all your loot, except for your Adam, and all your plasmids go back into the Gene Bank.
Heath is 100%, Eve is 20%
You need to either get new stuff, or go back to your dead body and scavenge.
In 'Hard' mode, in addition to the above, require a 'lock on to only you' hack before they're usable.
Until then, all Splicers will get Vita Chambered.
(I mean, story wise, why don't all the splicers get Vita Chambered when you kill them?)
Also, consistent with 'Smart Splicers' behavior above, Splicers still on the level are free to grab your loot.
Quests:
No changes.
Little Sisters:
In 'Hard' mode only, Little Sisters can be killed easily while they are harvesting Adam from a corpse - makes surprise attacks on Big Daddies much harder if they are close to a Sister!
What would be really funny is if the reason for the delay is that they're trying to port Spore from Linux - and they can't get the damned thing to work on Windows!
XOR would be a bad choice.
Assume you have A^B, B^C, and A^C
A^B ^ B^C = A^C
A^B ^ A^C = B^C
B^C ^ A^C = A^B
No matter what you do, without *one original message*, you cannot reconstruct the originals. (Now, this is cool for things like one time pad encryption, but it won't work for network coding).
However, if you *do* have an original message, they all fall out (assume A^B, B^C, and you have, say, 'A')
A^B ^ A = B (that's one!)
B^C ^ B = C (that's the other!)
So to reconstruct all three messages ('A', 'B', and 'C') we need three messages - two XORed, and one original.
I've used model rocket igniters, rigged up to a school desk with a nail pegboard (one nail per firework) and a iron bar 'wand'.
My 6 year old loved it! She could light all the fireworks from a safe distance by herself!
We only have 2 duds, but those were due to bad fuses on the fireworks.
On the other hand, I have taken my mother's advice: "Find what you would gladly do for
free, then find some sucker to pay you for it."
I love engineering, and I'd be happy if I was paid ditchdigger's wages. (And yes,
I have a wife and kid, and I can afford my my house at a lower salary - I live in
Pittsburgh!)
(Dang, I'm so used to Wiki...)
I used to work for LinuxCare, from January 2000 to Sept 2003. I have to say, to was a wild ride.
At the 'LinuxCare' phase, I mostly did contract work to write Linux device drivers for 3rd parties. (Including some absolutely evil stuff like a C++ stub for kernel modules, and a 'look like NT' wrapper for a MPEG encoder kernel module.)
In early 2000, we moved into our 'new' offices (we took up the entire basement of the huge converted warehouse building we were in), and had 'The Worlds Ugliest Mural' done by a local graffiti artist. The entire floor was carpeded with the LinuxCare 'X' logo. Yes, custom logo carpet.
Around 2001, the support business collapsed. The Founders left, except for Art, but we picked up a new CEO, some really smart IBM guys, and started working on what was to be the Levanta project. Originally targeted for IBM z/390 mainframes, it used the z/VM operating system to provide multiple 'on-demand' Linux-on-390 'partitions'. (z/VM is the mainframe equivalent to VMWare, but 20 years old !)
Akmal Khan came on board after Levanta was in full swing, and immediately took a dislike to the the distributed nature of our development group. There was Pittsburgh, doing the primary backend database; Ottawa was doing the web GUI and z/VM interface; Las Vegas handled the web infrastructure; project management in Atlanta; and San Francisco was sales and marketing. Except for SF and Ottawa, most sites telecommuted, so no 'office overhead' for those areas.
It became apparent pretty quickly that Akmal was the micromanaging type. By spring 2003, A.K. had collected his own group of technical people (very good ones, by the way) in SF, diverted all development of 'Levanta-on-Intel' to SF, and started making it pretty clear to the managers that all sites except SF would be going away.
That fall of 2003, the axe arrived for Ottawa, and I walked away from Levanta and the political mess that had developed.
I'm glad to have worked for LinuxCare, and had a ton-of-fun working on Levanta-on-z/390.
I used to work for LinuxCare, from January 2000 to Sept 2003. I have to say, to was a wild ride. At the 'LinuxCare' phase, I mostly did contract work to write Linux device drivers for 3rd parties. (Including some absolutely evil stuff like a C++ stub for kernel modules, and a 'look like NT' wrapper for a MPEG encoder kernel module.) In early 2000, we moved into our 'new' offices (we took up the entire basement of the huge converted warehouse building we were in), and had 'The Worlds Ugliest Mural' done by a local graffiti artist. The entire floor was carpeded with the LinuxCare 'X' logo. Yes, custom logo carpet. Around 2001, the support business collapsed. The Founders left, except for Art, but we picked up a new CEO, some really smart IBM guys, and started working on what was to be the Levanta project. Originally targeted for IBM z/390 mainframes, it used the z/VM operating system to provide multiple 'on-demand' Linux-on-390 'partitions'. (z/VM is the mainframe equivalent to VMWare, but 20 years old !) Akmal Khan came on board after Levanta was in full swing, and immediately took a dislike to the the distributed nature of our development group. There was Pittsburgh, doing the primary backend database; Ottawa was doing the web GUI and z/VM interface; Las Vegas handled the web infrastructure; project management in Atlanta; and San Francisco was sales and marketing. Except for SF and Ottawa, most sites telecommuted, so no 'office overhead' for those areas. It became apparent pretty quickly that Akmal was the micromanaging type. By spring 2003, A.K. had collected his own group of technical people (very good ones, by the way) in SF, diverted all development of 'Levanta-on-Intel' to SF, and started making it pretty clear to the managers that all sites except SF would be going away. That fall of 2003, the axe arrived for Ottawa, and I walked away from Levanta and the political mess that had developed. I'm glad to have worked for LinuxCare, and had a ton-of-fun working on Levanta-on-z/390.
Worst case scenario (unlikely), is that SCO has a legitimate claim. What are we to do?
.wav files!
Distribute kernel tarballs as
It's a sample now, in the noisiest song ever heard.
Plenty of existing legislation permitting samples....