The upside is getting Solaris into lots of geeky student systems.
Todays geeky students are the people who will be making purchasing decisions in 10 years time. If they are already familiar with Solaris they will be much more receptive to moving their Oracle DB onto Sun systems.
That would be some breakthrough, interfering at a quantum level with the probability of nuclear interactions.
However, the only people likely to discover such a secret are mad geniuses bent on world domination. After all it always happens that way on the movies.
I'm continually surprised by how often these free registration programs will accept root@localhost or abuse@localhost in the e-mail field.
This has the pleasing consequence that it unless they employ someone to vet the list they are likely to end up spamming themselves or their provider. Both much more amusing than sending it to a completely fake address.
Naturally, the rest of these forms must be treated as an exercise in creativity - and we should give our creations suitable names. My favourites include Hugh Jorgens and Tess Tickle.
So, treat these forms like you treat the religious nuts who arrive on your doorstep preaching salvation - as a source of amusement.
I've been there and done that. The company I was working for was taken over by a multi-national. Another key grunt and I were transferred onto the big guys pay scales on unfriendly terms.
So I found a job which payed better then went and talked to my boss. I told him what the problem was, detailed what I had been offered and gave him my resignation with plenty of notice.
He was unable or unwilling to match the new offer to I left. Better still my colleague found himself in a very strong bargaining position and got a substantial pay rise.
In my previous job I developed code which used OLE to analyse information from an app which had an Oracle back end and build Excel spreadsheets of the results.
When I wasn't on the clients site I would run personal oracle on my SMP system along with the application and a development environment. On a uniprocessor system this was dog-slow, but on an SMP system one CPU ran Oracle, Excel and the development environment and the other ran the app. and my code. Waay more responsive and didn't need a hugely expensive Oracle server license.
I guess a gigabyte of RAM probably helped too!
For code development SMP is great, especially if your compiler supports it. You can either have compiles running twice as fast, or use the second CPU for productive work while the code builds. Only drawback - no more coffee breaks whenever you rebuild the whole project.
If you're not living in the USA you might need to take a look at your local governments encryption policies before locking down your file systems. There are some countries where using strong encryption is a crime too.
OK - so let's get geeky about this. Why do we need to broadcast continuously to disrupt mobile phones. Why not listen for outgoing packets and emit a nice big rf chirp when the base station tries to handshake.
Benefits -
prevents users dialling out prevents users accepting calls low rf power requirements reasonable battery life difficult for law enforcement to track down
Disadvantages -
illegal more difficult to design
Any final year electronics students looking for an interesting project??
Surely this is no different from the tactics used by other large corporations. Hardly big news.
Short of major changes to the political establishment in Washington it will remain in the interest of big corporations to use 'soft money' to nudge the law making process in their direction.
While we're on the subject of history, aren't 90% of the Australian population descended from immigrants. The only difference is the the Abos didn't have the ability to force the Europeans back into international waters when they arrived with shiploads of surplus population.
1) Act as an interface between the programming team and the sales/client world.
2) Control the creeping feature problem
3) Buy a big whiteboard and use it to keep the project hit-list in one corner. When a problem arises or a milestone is getting close it goes on the hit-list. The rest of the board is for the programmers to explain stuff to each other.
4) Review the interface descriptions for what people are building. You probably don't need to know exactly how things work, but you _must_ ensure that all the bits fit together
5) Plan training and career development for your team, and ensure that it happens.
6) When a project is completed on time and under budget, spend some of the remaining money on the team. If you're waaay under budget (does happen occasionally) host a party for some customer types so the programmers can show off how cool they've been.
7) Give the programmers some slack, but not too much.
The problem is that even very small fragments present a hazard to future satellites. A 0.5" fragment travelling at orbital velocity will punch a hole right through a satellite.
Satellite disposal can take two courses - drop it into the atmosphere or shoot it into space. Dropping it into the asmosphere is _much_ cheaper to implement so that's the preferred option.
My inclination would be to require the satellite designers to ensure safer re-entry characteristics, i.e. satellites which will break up into small parts early in the re-entry process. My first thought is to assemble the heavy parts of the satellite using fusible links so that before it reaches red heat these parts separate into a number of smaller fragments.
This would probably leave the main motor components (which have to operate at high temperatures anyway) as the largest piece to enter the atmosphere.
The downside to this approach is a small reduction in satellite reliability - every extra component means one more thing to go wrong.
I considered using smoothwall for my home network, but the lack of SCSI support was a problem. Instead I downloaded the Mandrake SNF firewall as an iso image. It's based on Mandrake 7.2 so it's not the latest and greatest. However it does include a facility to update its components if you don't feel like downloading the RPMs yourself.
IIRC this does implement shadow passwords as well as the SCSI support I needed and a web interface. There is also a Java encrypted terminal connection, which allows you to login securely from a browser. This is really handy for tweaking the config files without needing a screen and keyboard connected to the host.
There were two areas which needed a little manual tweaking - dhcpd.conf and lilo.conf. Once these had been fixed everything worked a treat, it even handled the VPN connection to my office seamlessly. So, nine out of ten for Mandrake SNF from me.
How many of you think that you could decipher the structure of the command (given the motivation)?
I suspect that the command structure will be fairly simple on most scientific satellites (excepting Hubble) given that there are only a handful of operations which need to be performed. I wouldn't be too surprised if the commands are just a variation of a published scheme in most cases.
The technology to send commands should be within the reach of a decent physics graduate, but I suspect that the price of a suitable transmitter will be the limiting factor for most potential hackers. By the time we're earning enough to pay for a two meter dish and a suitable rack of electronics we've lost interest in hacking satellites.
Women are great at multi-tasking. They can iron, watch tv, mind the kids and still find time to tell you all the things you have done wrong in the last two weeks.
If you're a SCSI freak you will probably need to double the budget to get ~1 terabyte of RAID5 storage. First stop is half a dozen 180Gb drives from Seagate (http://www.seagate.com/cda/products/discsales/mar keting/detail/0,1121,331,00.html) and an Ultra160 RAID controller. Add just about any processor, a stack of RAM, and as many network cards as you can afford.
I'd be interested to see them take this a step further. Add an additional servo controlled element at the foot of each foil. When the boat starts to heel over the servo reduces the lift on the upwind side and increases lift on the downwind side. In extreme conditions it might even need to generate significant downforce to keep the boat level. This is quite a big step, because most of the structural members would need to handle new loads, but it should result in a faster ride.
They could even run RT Linux on a PC104 system to monitor speed and servo angle to avoid stalling the additional surface. I'll volunteer to go along and keep an eye on it:-).
Best of all - it would look pretty damn cool at high speed with a stream of water shooting up behind the upwind 'foil. (Well that momentum has to go somewhere).
I guess someone will create a real-time kernel for the built-in hardware (something like RT Linux) when the price drops to $100. However, it still seems like complete overkill for most control applications. There is a lot of competition in the microcontroller market and some pretty good one chip controllers for less than $20.
OTOH a rack containing a few hundred of these would make a great render farm or code cracker.
Keith.
Hmm - IIRC the 'permanent gases', nitrogen, oxygen, helium,neon, etc are so called because they can't be liquified by pressure at room temperature. And, even if they could, they wouldn't be much use for _cooling_ at room temperature.
The Danish site makes it fairly clear that the liquid nitrogen is used as a heat transfer medium rather than being allowed to boil off and provide cooling that way.
On the subject of running CPUs around 77K, I would expect silicon to act as a pretty good insulator around those temperatures.
Just my friday morning thoughts,
Keith.
As I understand it, the whole point of a distributed system is that there is more than one copy of each piece of information. They are passed around the network as necessary, and cached on the machines they pass through.Ideally this will result in the keys you use most frequently being cached on machines near you.
I can see several problems with using this kind of system for public keys.
Firstly, there is no defined search path (unlike DNS) so it can take a long time to retrieve a key the first time you request it. There is also the possibility that requests might find a circular path.
Secondly, I'm not sure that the bandwidth requirements are a linear function of the number of users. Any slashdotters out there tried simulating this type of system??
Thirdly, there needs to be a way of flushing old versions of a key from all the caches without using up heaps of bandwidth. Otherwise old keys can persist almost indefinitely.
Time to get out one of those tacky sixties phones and see if you can build a web server inside it. Failing that maybe there's room in the bottom of a lava lamp.
Perhaps it's time for a/. competition for the most tasteless web server. Not just plain crude - tasteless, like your grandma wearing an Eminem T-shirt.
Keith
Slashdot Mirror
Mummy, mummy .... I want Santa to bring me a BFG!!
But Dear, you hardly play with the Chainsaw uncle Keith bought you for your birthday...
Naah - he applied to Metronomy for a PC and they informed the Army.
Metronomy are going to claim the $25M reward. That's the only way this business model makes any sense at all.
Keith.
The upside is getting Solaris into lots of geeky student systems.
Todays geeky students are the people who will be making purchasing decisions in 10 years time. If they are already familiar with Solaris they will be much more receptive to moving their Oracle DB onto Sun systems.
Definitely a long term game plan.
Keith
That would be some breakthrough, interfering at a quantum level with the probability of nuclear interactions.
However, the only people likely to discover such a secret are mad geniuses bent on world domination. After all it always happens that way on the movies.
Population Density might have something to do with it. There's a lot more potential customers per square mile in Japan than most other countries.
I'm continually surprised by how often these free registration programs will accept root@localhost or abuse@localhost in the e-mail field.
This has the pleasing consequence that it unless they employ someone to vet the list they are likely to end up spamming themselves or their provider. Both much more amusing than sending it to a completely fake address.
Naturally, the rest of these forms must be treated as an exercise in creativity - and we should give our creations suitable names. My favourites include Hugh Jorgens and Tess Tickle.
So, treat these forms like you treat the religious nuts who arrive on your doorstep preaching salvation - as a source of amusement.
I've been there and done that. The company I was working for was taken over by a multi-national. Another key grunt and I were transferred onto the big guys pay scales on unfriendly terms.
So I found a job which payed better then went and talked to my boss. I told him what the problem was, detailed what I had been offered and gave him my resignation with plenty of notice.
He was unable or unwilling to match the new offer to I left. Better still my colleague found himself in a very strong bargaining position and got a substantial pay rise.
Definitely a win-win situation.
In my previous job I developed code which used OLE to analyse information from an app which had an Oracle back end and build Excel spreadsheets of the results.
When I wasn't on the clients site I would run personal oracle on my SMP system along with the application and a development environment. On a uniprocessor system this was dog-slow, but on an SMP system one CPU ran Oracle, Excel and the development environment and the other ran the app. and my code. Waay more responsive and didn't need a hugely expensive Oracle server license.
I guess a gigabyte of RAM probably helped too!
For code development SMP is great, especially if your compiler supports it. You can either have compiles running twice as fast, or use the second CPU for productive work while the code builds. Only drawback - no more coffee breaks whenever you rebuild the whole project.
If you're not living in the USA you might need to take a look at your local governments encryption policies before locking down your file systems. There are some countries where using strong encryption is a crime too.
Keith.
OK - so let's get geeky about this. Why do we need to broadcast continuously to disrupt mobile phones. Why not listen for outgoing packets and emit a nice big rf chirp when the base station tries to handshake.
Benefits -
prevents users dialling out
prevents users accepting calls
low rf power requirements
reasonable battery life
difficult for law enforcement to track down
Disadvantages -
illegal
more difficult to design
Any final year electronics students looking for an interesting project??
Keith.
Best of all - when you re-work some components the number of lines of code may go down. Does this mean that the project is going backwards??
Keith.
Surely this is no different from the tactics used by other large corporations. Hardly big news.
Short of major changes to the political establishment in Washington it will remain in the interest of big corporations to use 'soft money' to nudge the law making process in their direction.
While we're on the subject of history, aren't 90% of the Australian population descended from immigrants. The only difference is the the Abos didn't have the ability to force the Europeans back into international waters when they arrived with shiploads of surplus population.
1) Act as an interface between the programming team and the sales/client world.
2) Control the creeping feature problem
3) Buy a big whiteboard and use it to keep the project hit-list in one corner. When a problem arises or a milestone is getting close it goes on the hit-list. The rest of the board is for the programmers to explain stuff to each other.
4) Review the interface descriptions for what people are building. You probably don't need to know exactly how things work, but you _must_ ensure that all the bits fit together
5) Plan training and career development for your team, and ensure that it happens.
6) When a project is completed on time and under budget, spend some of the remaining money on the team. If you're waaay under budget (does happen occasionally) host a party for some customer types so the programmers can show off how cool they've been.
7) Give the programmers some slack, but not too much.
Satellite disposal can take two courses - drop it into the atmosphere or shoot it into space. Dropping it into the asmosphere is _much_ cheaper to implement so that's the preferred option.
My inclination would be to require the satellite designers to ensure safer re-entry characteristics, i.e. satellites which will break up into small parts early in the re-entry process. My first thought is to assemble the heavy parts of the satellite using fusible links so that before it reaches red heat these parts separate into a number of smaller fragments. This would probably leave the main motor components (which have to operate at high temperatures anyway) as the largest piece to enter the atmosphere.
The downside to this approach is a small reduction in satellite reliability - every extra component means one more thing to go wrong.
Any space systems engineers care to comment?
Keith.
I considered using smoothwall for my home network, but the lack of SCSI support was a problem. Instead I downloaded the Mandrake SNF firewall as an iso image. It's based on Mandrake 7.2 so it's not the latest and greatest. However it does include a facility to update its components if you don't feel like downloading the RPMs yourself.
IIRC this does implement shadow passwords as well as the SCSI support I needed and a web interface. There is also a Java encrypted terminal connection, which allows you to login securely from a browser. This is really handy for tweaking the config files without needing a screen and keyboard connected to the host.
There were two areas which needed a little manual tweaking - dhcpd.conf and lilo.conf. Once these had been fixed everything worked a treat, it even handled the VPN connection to my office seamlessly. So, nine out of ten for Mandrake SNF from me.
Keith.
I suspect that the command structure will be fairly simple on most scientific satellites (excepting Hubble) given that there are only a handful of operations which need to be performed. I wouldn't be too surprised if the commands are just a variation of a published scheme in most cases.
The technology to send commands should be within the reach of a decent physics graduate, but I suspect that the price of a suitable transmitter will be the limiting factor for most potential hackers. By the time we're earning enough to pay for a two meter dish and a suitable rack of electronics we've lost interest in hacking satellites.
Just my thoughts,
Keith
I prefer "They all wait at the same speed".
Women are great at multi-tasking. They can iron, watch tv, mind the kids and still find time to tell you all the things you have done wrong in the last two weeks.
If you're a SCSI freak you will probably need to double the budget to get ~1 terabyte of RAID5 storage. First stop is half a dozen 180Gb drives from Seagate (http://www.seagate.com/cda/products/discsales/mar keting/detail/0,1121,331,00.html) and an Ultra160 RAID controller. Add just about any processor, a stack of RAM, and as many network cards as you can afford.
Just my friday afternoon thoughts,
Keith.
I'd be interested to see them take this a step further. Add an additional servo controlled element at the foot of each foil. When the boat starts to heel over the servo reduces the lift on the upwind side and increases lift on the downwind side. In extreme conditions it might even need to generate significant downforce to keep the boat level. This is quite a big step, because most of the structural members would need to handle new loads, but it should result in a faster ride.
:-).
They could even run RT Linux on a PC104 system to monitor speed and servo angle to avoid stalling the additional surface. I'll volunteer to go along and keep an eye on it
Best of all - it would look pretty damn cool at high speed with a stream of water shooting up behind the upwind 'foil. (Well that momentum has to go somewhere).
Just an idle thought,
Keith.
I guess someone will create a real-time kernel for the built-in hardware (something like RT Linux) when the price drops to $100. However, it still seems like complete overkill for most control applications. There is a lot of competition in the microcontroller market and some pretty good one chip controllers for less than $20. OTOH a rack containing a few hundred of these would make a great render farm or code cracker. Keith.
Hmm - IIRC the 'permanent gases', nitrogen, oxygen, helium,neon, etc are so called because they can't be liquified by pressure at room temperature. And, even if they could, they wouldn't be much use for _cooling_ at room temperature. The Danish site makes it fairly clear that the liquid nitrogen is used as a heat transfer medium rather than being allowed to boil off and provide cooling that way. On the subject of running CPUs around 77K, I would expect silicon to act as a pretty good insulator around those temperatures. Just my friday morning thoughts, Keith.
As I understand it, the whole point of a distributed system is that there is more than one copy of each piece of information. They are passed around the network as necessary, and cached on the machines they pass through.Ideally this will result in the keys you use most frequently being cached on machines near you.
I can see several problems with using this kind of system for public keys.
Firstly, there is no defined search path (unlike DNS) so it can take a long time to retrieve a key the first time you request it. There is also the possibility that requests might find a circular path.
Secondly, I'm not sure that the bandwidth requirements are a linear function of the number of users. Any slashdotters out there tried simulating this type of system??
Thirdly, there needs to be a way of flushing old versions of a key from all the caches without using up heaps of bandwidth. Otherwise old keys can persist almost indefinitely.
Just my slightly hung-over thoughts,
Keith.
Time to get out one of those tacky sixties phones and see if you can build a web server inside it. Failing that maybe there's room in the bottom of a lava lamp. Perhaps it's time for a /. competition for the most tasteless web server. Not just plain crude - tasteless, like your grandma wearing an Eminem T-shirt.
Keith