Keeping on top of the management of a server farm may be more challenging in those specific aspects than "designing spacecraft", but if you install a patch and it hoses your server:
- The server does not explode in a ball of fire taking (B/M)ILLIONS of dollars of equipment with it. - The server is accessible and (worst case) can be fixed by having someone install new hardware and restoring from a backup. It's not a 50,000 miles away, requiring HUNDREDS of people and (B/M)ILLIONS of dollars just to get to it to replace a borked RAM chip. - When the server fails you don't lose (B/M)ILLIONS of dollars of capital investment and 300,000+ man hours (30 ppl * 2000 hrs/yr * 5 yrs) - Your bug doesn't kill 7 of your best specialized programmers who have received MILLIONS of dollars worth of training and spent YEARS training to work on your server.
I would check the horse's mouth again for bad teeth.
End of SPAM? Probably not (as SPAM mailers can authenticate themselves, and Microsoft WORMS and Viruses can hijack legitimate mail servers which authenticate themselves and send SPAM anyway) but it is a start.
True. But wouldn't being able to easily filter out mail servers without MX/PTR records, thus narrowing the SPAM sources to poorly-protected intentional mail servers, dramatically increase the usefullness of the RBL system?
Yes, but is it really a map?
on
SCOoby Snacks
·
· Score: 1
3. SCO UNIX(R) has a Committed, Well-Defined Roadmap
If there's an impact on one side of a sphere (or oblate spheroid, the Earth) shockwaves are going to travel along the surface boundary. As the waves radiate away from the impact site, they end up converging on the opposite side of the sphere (the antipode).
If the eruptions are directly linked to the impact and the tectonic stresses from it, the stress is greatest at the impact site and second greatest at the antipode of the impact site.
The new Scientific American has a great article on the K-T impact and they theorize the antipode got a double whammy as not only the shockwaves converge there, but a lot of the ejecta from the impact re-enters the atmosphere at or near the antipode. In addition, the falling debris heats the atmosphere several hundred degrees long enough to set fire to most vegetation. So you get lots of ash (and buckyballs) in the geologic record to look at as well.
I got out in 2000, but I was still in the USMC (0311, grunt, qty. 1) when we switched from the old LPC (Leather Personnel Carrier) to the new boot. I didn't get to try the desert boot, but I really liked the ICB, the first version was a bit hot/heavy, but I have heard from buddies that it is better now.
I think this make a great first step towards the complete Elevator concept.
Dr. Forward and Tethers Unlimited had advanced the idea of the "Rotovator", a spinning tether that orbits such that as teh end of the tether dips down to suborbital height it has a low enough speed that a vehicle could "catch" it and ride it up into orbit.
Is this the "killer app" that will focus NASA and teh X-PRize people on the use of reusable suborbital vehicles to get into space?
How hard would it be to implement something like the/proc virtual fs for/ldap ?
Would the ldap data be too large? Could you use the ldap data in a database and have the read/writes to/ldap actually transform into database calls for the info?
I have no idea what I am talking about, feel free respond appropriately.
dselect sucks and is not worth the pain. aptitude, on the other hand, is Very Good. (Incidentally, I wonder if on Red Hat + APT, I can browse packages like I do with aptitude's UI...)
I second this notion!
Setting up a new debian box for me usually goes:
what incentive do manufacturers have for designing protocols and building equipment for distributed routing?
NASA is funding research on this kind of mesh computing for the "sensor web" robots for exploration aren't they?
And you wanted to know WHY we need faster chips?
on
Reflections
·
· Score: 2, Interesting
"One restraint on speed is the intense data processing it requires. With current technology, higher speeds would demand chips that are too large and too power-hungry for hand-held devices."
So is this the next killer app to fund chip research?
Reporter: "The new Boeing 787 crashed on takeoff today killing 500 people, here's the lead engineer..."
Engineer: "Well, the computer said it would work. I guess we'll start over from scratch and try it again. I don't know how the computer could have been wrong."
So could this effect be applied to the skin of a Europoa ice rover that would melt its way through 2km of ice?
I always hear that using heat to melt it would be impractical, but with a drill and this electric field effect maybe something more energy efficient could be done?
Paint ALL of the asteroid. If it's rotating (and they ALL are rotating to some degree) only ~50% will be facing the sun at one time anyway, and you still get the effect of light pressure pushing it away from the sun as it rotates.
Incompetent pilot?
on
Book on NR-1
·
· Score: 3, Interesting
From the book excerpt:
The ship was so stable that it automatically tried to keep itself level, which meant that as we came down the slope, the bow tended to rise and stay even with the stern. Only the rear wheel was touching the mud as the forward part of the boat angled slightly higher. That pointed the forward television cameras up too far to see anything on the bottom. That separation of the bow from the bottom also limited the effectiveness of the sonar.
Maurer added still more water to the forward ballast tanks, which brought the bow down and put both wheels back in contact. Unfortunately, the extra ballast made the boat so heavy that the maximum upward force from the combined fore-and-aft thrusters would not be able to lift it. We rolled ahead.
Why didn't the idiot pilot add a little water to the bow tanks and release some from the aft tanks? He was either totally incompetent (highly unlikely) or this "teaser" is made up...
oh well..
Lets ASSUME, that terrorists can get at the cable at some point with a nuclear weapon. It would be nearly impossible to sneak one onto a cargo pod as all that stuff will probably be inspected and loaded by hand... So the best way is to stick a nuke in a plane and then fly it near the ribbon and BOOM. right?
So lets say they manage to do this at um... around 65000 feet (not likely) and lets say that the weapon manages to vaporize/burn/shred 100 miles of ribbon above the blast. ok?
So the terrorists have now destroyed about 113 miles of a ***60,000*** mile long ribbon. That's a whopping 0.002%...
So, assuming that NASA/ESA/Whoever was planning on doing some maintenance at SOME POINT in the life cycle of the cable, and that probably SOME of the maintenance would be on the end closer to the orbiting station, there's probably SOME ribbon repair junk up there.
So it takes them 7 days to send down a robot with enough ribbon to get the lowest 150 miles reattached and then the ground station can send up the mondo repairbots to fix the cable....
So how does this prevent the cable from being back in operation in a few weeks, maybe months at most?
And that's after some terrorists "blow it up" using a nuclear weapon!
I am **dying** to get an open source replacement that I can use in production to replace Exchange...
But if they "slap a pretty administative UI on top" like you say, I won't touch that shit with a 3m pole.
Unless they do some serious design and usability testing on the administrative interface then it will never be useful. Frankly I have doubts based on their early spec for it...
They defined 8 separate "features": 3 for logging, 2 for vacation setup, 2 for server setup, and 1 for user administration. Now I'm hardly familiar with PostFix, OpenLDAP, etc. so I don't know if that's really enough to administer the Kolab server, but I am familiar with Exchange and it looks like there is no way the current spec covers all the configuration options and tools available to an Exchange Admin.
Are there more tools that come with OpenLDAP that they are going to include in the administration?
How do you do things like search all the email on a server/mailbox to find certain text? I vaguely remember reading that all data is stored in separate files instead of a database, so do you use "find" (or whatever the console command is. i am a newbie at *nix). And if so are they going to make that accessible through the php interface...
I sure hope the spec grows a bit more from the inital one they have.
Perhaps someday there will be a $20/month service that lets me watch whatever I want to watch whever I want to watch it forever. When that day comes I'll feel like $20 per movie is a ripoff.
Try NetFlix. I've been using it for over a year and it ROCKS. Of course I live very near one of the main distribution cneters so most of my movie turn-around times are about 2 days from sending out to getting the next one on my list.
Just include a PIN in the verification phase. The article mentioned a central database that stores a bunch of input/ouput combinations to use for verification.
So I imagine it goes something like: ========== Card Reader: take a reference reading, transmit to CC Company. To identify the card.
CC Company: Verify reference reading, request predetermined (or not!) number of other readings (i.e. laser positions 23, 42, and 8974) to confirm the card. Request PIN reading.
Card Reader : request a PIN from user, use some hash to convert the PIN into a set of laser readings. Take the laser readings and transmit to CC Company.
CC Company: Compare the PIN readings to database and verify.
==========
I know that's simplistic but at least it requires 2 of 3 factors for really strong authentication (something you have, something you know, something you are). And unlike the debit cards we use now (i use anyway), the PIN can be a required part of the transaction.
First, launching from the sea has been and IS being done now. In fact if you bother to read the Highlift website summary, you'll note that Sea Launch is the basis for the tether platform. I'm not aware of any extraordinary security taken for Sea Launch that wouldn't otherwise be taken for high value items.
As for defending it from attacks, well... This is a 1 METER wide ribbon, that's going to be about as thick as a sheet of paper. In the middle of the ocean. Design to withstand meteor impacts. Do you think someone is just going to walk over and cut it down with a hacksaw? Or shoot it? Crash into it? And the worst thing you could POSSIBLY do to the ribbon is knock off a few miles of it. The rest is still in space, probably with many miles of reserve ribbon for repairs etc. ready to be lowered and have it fixed in a few months...
Losing a few miles probably won't even be that big of a deal considiering the minor percentage amount of mass that would be lost. I'm sure they'll have some mobile counter weights on the end that can just be reeled in to compensate, until the new ribbon is deployed.
Slashdot Mirror
I thought her acting "ability" was reason enough to send her into space...
That's a pretty ridiculous statement.
Keeping on top of the management of a server farm may be more challenging in those specific aspects than "designing spacecraft", but if you install a patch and it hoses your server:
- The server does not explode in a ball of fire taking (B/M)ILLIONS of dollars of equipment with it.
- The server is accessible and (worst case) can be fixed by having someone install new hardware and restoring from a backup. It's not a 50,000 miles away, requiring HUNDREDS of people and (B/M)ILLIONS of dollars just to get to it to replace a borked RAM chip.
- When the server fails you don't lose (B/M)ILLIONS of dollars of capital investment and 300,000+ man hours (30 ppl * 2000 hrs/yr * 5 yrs)
- Your bug doesn't kill 7 of your best specialized programmers who have received MILLIONS of dollars worth of training and spent YEARS training to work on your server.
I would check the horse's mouth again for bad teeth.
Well... does this really qualify as a map???
Not coincidence at all.
If there's an impact on one side of a sphere (or oblate spheroid, the Earth) shockwaves are going to travel along the surface boundary. As the waves radiate away from the impact site, they end up converging on the opposite side of the sphere (the antipode).
If the eruptions are directly linked to the impact and the tectonic stresses from it, the stress is greatest at the impact site and second greatest at the antipode of the impact site.
The new Scientific American has a great article on the K-T impact and they theorize the antipode got a double whammy as not only the shockwaves converge there, but a lot of the ejecta from the impact re-enters the atmosphere at or near the antipode. In addition, the falling debris heats the atmosphere several hundred degrees long enough to set fire to most vegetation. So you get lots of ash (and buckyballs) in the geologic record to look at as well.
You could always use this instead
Thanks for the thought, though!
I got out in 2000, but I was still in the USMC (0311, grunt, qty. 1) when we switched from the old LPC (Leather Personnel Carrier) to the new boot. I didn't get to try the desert boot, but I really liked the ICB, the first version was a bit hot/heavy, but I have heard from buddies that it is better now.
I think this make a great first step towards the complete Elevator concept.
Dr. Forward and Tethers Unlimited had advanced the idea of the "Rotovator", a spinning tether that orbits such that as teh end of the tether dips down to suborbital height it has a low enough speed that a vehicle could "catch" it and ride it up into orbit.
Is this the "killer app" that will focus NASA and teh X-PRize people on the use of reusable suborbital vehicles to get into space?
Hmm... interesting.
/proc virtual fs for /ldap ?
/ldap actually transform into database calls for the info?
How hard would it be to implement something like the
Would the ldap data be too large?
Could you use the ldap data in a database and have the read/writes to
I have no idea what I am talking about, feel free respond appropriately.
Setting up a new debian box for me usually goes:
1.) Install Debian using netinstall.
2.) # Apt-get install aptitude
3.) Continue to lock down the box, add apps, etc.
Aptitude rocks!
Dell = Fry : Delivery Boy
But IBM = Dr. Farnsworth : 100 Year old Nobel Prize winning(seeking?) mad scientist
Am I the only one that saw the title and thought, "Cool! I wonder if they'll provide a door to Warcraft or a Warcraft version TradeWars?"
Damn I spent too much time BBS'n.
NASA is funding research on this kind of mesh computing for the "sensor web" robots for exploration aren't they?
"One restraint on speed is the intense data processing it requires. With current technology, higher speeds would demand chips that are too large and too power-hungry for hand-held devices."
So is this the next killer app to fund chip research?
So the alternative is better????
Reporter: "The new Boeing 787 crashed on takeoff today killing 500 people, here's the lead engineer..."
Engineer: "Well, the computer said it would work. I guess we'll start over from scratch and try it again. I don't know how the computer could have been wrong."
So could this effect be applied to the skin of a Europoa ice rover that would melt its way through 2km of ice?
I always hear that using heat to melt it would be impractical, but with a drill and this electric field effect maybe something more energy efficient could be done?
Is this the next everquest? Will we have massive Star War fans calling in sick everyday in attempts to become a jedi?
I think the "massive" fans would be more inclined towards attempts at becoming a Hut...
Easy enough to solve that problem...
Paint ALL of the asteroid. If it's rotating (and they ALL are rotating to some degree) only ~50% will be facing the sun at one time anyway, and you still get the effect of light pressure pushing it away from the sun as it rotates.
From the book excerpt: The ship was so stable that it automatically tried to keep itself level, which meant that as we came down the slope, the bow tended to rise and stay even with the stern. Only the rear wheel was touching the mud as the forward part of the boat angled slightly higher. That pointed the forward television cameras up too far to see anything on the bottom. That separation of the bow from the bottom also limited the effectiveness of the sonar.
Maurer added still more water to the forward ballast tanks, which brought the bow down and put both wheels back in contact. Unfortunately, the extra ballast made the boat so heavy that the maximum upward force from the combined fore-and-aft thrusters would not be able to lift it. We rolled ahead.
Why didn't the idiot pilot add a little water to the bow tanks and release some from the aft tanks? He was either totally incompetent (highly unlikely) or this "teaser" is made up... oh well..
OK.
Lets ASSUME, that terrorists can get at the cable at some point with a nuclear weapon. It would be nearly impossible to sneak one onto a cargo pod as all that stuff will probably be inspected and loaded by hand... So the best way is to stick a nuke in a plane and then fly it near the ribbon and BOOM. right?
So lets say they manage to do this at um... around 65000 feet (not likely) and lets say that the weapon manages to vaporize/burn/shred 100 miles of ribbon above the blast. ok?
So the terrorists have now destroyed about 113 miles of a ***60,000*** mile long ribbon. That's a whopping 0.002%...
So, assuming that NASA/ESA/Whoever was planning on doing some maintenance at SOME POINT in the life cycle of the cable, and that probably SOME of the maintenance would be on the end closer to the orbiting station, there's probably SOME ribbon repair junk up there.
So it takes them 7 days to send down a robot with enough ribbon to get the lowest 150 miles reattached and then the ground station can send up the mondo repairbots to fix the cable....
So how does this prevent the cable from being back in operation in a few weeks, maybe months at most?
And that's after some terrorists "blow it up" using a nuclear weapon!
I am **dying** to get an open source replacement that I can use in production to replace Exchange...
But if they "slap a pretty administative UI on top" like you say, I won't touch that shit with a 3m pole.
Unless they do some serious design and usability testing on the administrative interface then it will never be useful. Frankly I have doubts based on their early spec for it...
They defined 8 separate "features": 3 for logging, 2 for vacation setup, 2 for server setup, and 1 for user administration. Now I'm hardly familiar with PostFix, OpenLDAP, etc. so I don't know if that's really enough to administer the Kolab server, but I am familiar with Exchange and it looks like there is no way the current spec covers all the configuration options and tools available to an Exchange Admin.
Are there more tools that come with OpenLDAP that they are going to include in the administration?
How do you do things like search all the email on a server/mailbox to find certain text? I vaguely remember reading that all data is stored in separate files instead of a database, so do you use "find" (or whatever the console command is. i am a newbie at *nix). And if so are they going to make that accessible through the php interface...
I sure hope the spec grows a bit more from the inital one they have.
Perhaps someday there will be a $20/month service that lets me watch whatever I want to watch whever I want to watch it forever. When that day comes I'll feel like $20 per movie is a ripoff.
Try NetFlix. I've been using it for over a year and it ROCKS. Of course I live very near one of the main distribution cneters so most of my movie turn-around times are about 2 days from sending out to getting the next one on my list.
Just include a PIN in the verification phase. The article mentioned a central database that stores a bunch of input/ouput combinations to use for verification.
So I imagine it goes something like:
==========
Card Reader: take a reference reading, transmit to CC Company. To identify the card.
CC Company: Verify reference reading, request predetermined (or not!) number of other readings (i.e. laser positions 23, 42, and 8974) to confirm the card. Request PIN reading.
Card Reader : request a PIN from user, use some hash to convert the PIN into a set of laser readings. Take the laser readings and transmit to CC Company.
CC Company: Compare the PIN readings to database and verify.
==========
I know that's simplistic but at least it requires 2 of 3 factors for really strong authentication (something you have, something you know, something you are). And unlike the debit cards we use now (i use anyway), the PIN can be a required part of the transaction.
> They do a procedure like this for certain kinds of horses all the time.
Right! So what are we going to do about the evil bastards raping all of our horses!?!?
First, launching from the sea has been and IS being done now. In fact if you bother to read the Highlift website summary, you'll note that Sea Launch is the basis for the tether platform. I'm not aware of any extraordinary security taken for Sea Launch that wouldn't otherwise be taken for high value items.
As for defending it from attacks, well... This is a 1 METER wide ribbon, that's going to be about as thick as a sheet of paper.
In the middle of the ocean.
Design to withstand meteor impacts.
Do you think someone is just going to walk over and cut it down with a hacksaw? Or shoot it? Crash into it? And the worst thing you could POSSIBLY do to the ribbon is knock off a few miles of it. The rest is still in space, probably with many miles of reserve ribbon for repairs etc. ready to be lowered and have it fixed in a few months...
Losing a few miles probably won't even be that big of a deal considiering the minor percentage amount of mass that would be lost. I'm sure they'll have some mobile counter weights on the end that can just be reeled in to compensate, until the new ribbon is deployed.