Three easy steps:
1) Get program to parse browser history for URL's
2) Awk output for domains and save to blocklist
3) cat blocklist | pine support@ezula.com
Just a few examples of the useless stuff that we've attained from NASA and the military. Though I should mention it after reading some of these posts... After all, pure research is worthless and expensive, and if the military does it, well then, it must be bad.
Plastic, modern ceramics, nuclear power, fuel efficient car engines, wireless anything, teflon, semiconductors and superconductors, transistors, the Internet, microwave ovens, GPS, computers, engine emissions scrubbers, and more and more....
So lets cut research, great idea. Way to invent things that people haven't already thought of, because we all know that we got were we are by improving existing things.
Sounds alright, except you never get anything unless other people have bookmarked it. So does that mean you would never find new sites? What if I were to come up with some incredible idea, and put it on the web tomorrow. How would anybody find it, except by me directly telling them? I don't publicise my sites, but Google and AltaVista and likes have them. This would never have them! Sounds like the engine would be doomed to fail from it's conception, unless they start doing something else in addition.
And if you take the digit placements (2 and 6), the difference from 0 (the first) and 2 is 2, and the difference from 2 and 6 is 4. If you invert them you get 42. Coincidence?
I don't know about the GNOME ports, but your 486/66 is a more than adequate machine. A low end 486 can easily flood a T1 or two, your cable modem isn't going to be a problem to route for. I'm using one right now for something quite similar!
The S3 Virge chipsets are supported by just about everything out there. So is the Millenium II card. Both are rather nice, the Matrox definetely being the faster of them, but if you just need good 2d, the Virge cards are cheap. If you want really nice 2d, the Millenium II is a little more expensive, but also very well supported.
SCSI allows many things that IDE doesn't do. It has a rather mature RAID controller market, not to say IDE RAID doesn't exist in hardware and software. Also, I don't remember the last hot swappable IDE drive that worked. Since I deal with many arrays of disks chained off a smaller number of computers, it would even be *possible* to have the number of discs we have. You could only get four drives in an array with IDE. That's nowhere near acceptable. Plus since you can't do command queueing and simultaneous transactions on IDE, the performance is crap for it. Plus, since the cables have length limitations, especially for ATA/66, it's no good for external bays. IDE is nice and all, but I would never rely on it.
As for WD, just had *two* drives fail within five hours of each other from them. One a year newer than the other. Don't miss them a bit, perfectly happy with Fujitsu and IBM right now.
Most Linux products have what, an average of six - ten updates a year? Take a look at Freshmeat. Some things are updated every few days. It wouldn't work very well in real life. I'd rather go with the standard versioning. It's a lot easier to deal with 4.0, 4.1, etc than 95, 95 OSR2, 95 OSR2.5, 95OSR-Your_mama. The actual version number shows how little difference there was between products: Win95 (4.0), Win98 (4.1). The same generally hold true for other things. A point release tends to not add features (well, unless it's pre v1.0), just fix problems. But a full version adds things. IE: 2.0.38 differs from 2.2.13, but much of the underlying code stays the same, it wasn't a rewrite. 3.1 is much different from 4.0/4.1. See the point? So we can have Linux 2000, Linux 2000.2.13, or Linux 2.2.13. The smaller one looks more manageable.
Quite the contrary, UDP is very useful still. With multimedia you don't need to guarentee that all the packets get to the destination, just that the packets are only used in the correct sequence. You drop packets that arrive late. With standard TCP net congestion can really kill your stream due to all the retries to make sure everything gets where it's going.
This isn't always the best, for example, if you require everything gets to the destination. Even then, though, if you have a generously sized buffer, you can store packets and request for a missed one, thus potentionally saving time on retransmits. This can be useful in pausing a stream temporarily as well.
This makes UDP very good to use for real-time streaming. Besides, any contribution to open source should be welcomed... if you don't like the contribution, just don't use it!;-) There are already a few existing servers, do a few searches on freshmeat. However, something that Apple wants to include in their flagship server software should be carefully looked at too.
The other possible reason is that you can't afford to get sued by the five homeowners between you and your access point. Also, the network admin at that access point had some sort of aversion to a static discharge (read lightning or some like punk messing with the cable) ruining his many thousands of dollars of network equipment in his network closet. What's the tradeoff here, potential for $24,000 in damage because we're lazy and run cable, or play it safe and use wireless...
Plus, you might want to ask those neighbors about your cable, or it'll be interesting the next time they get a nice we lawn and your cable shows through
A bunch of friends and I were looking into this idea. The problem comes in with a few things. First, you need chips to modulate your laser from your ethernet pulsing, and also to decode and amplify from the laser recieve photosensor. As far as line-of-sight, it isn't bad, but *nothing* can be in the way. If you use a class 3 laser (laser pointers are part of this) you are generally safe, as long as people never are able to look at the beam. If you go any higher than class 3, you have safety concerns...
Also, you need a fairly good laser diode to transmit with, just because you can't permit the beam to diffuse very much, or you'll get signal loss. This can be compensated for with different types of beam modulations, but still can be a bit daunting.
The chips we were looking at were the same they used in OC-48 fiber terminations, but unfortunately I dont' remember the company.
As fun as the talk about liquid cooling is, wouldn't it be easier and safer to use a low tempreture liquid to cool the air instead? For instance, a typical freezer uses Freon; what if you adapted a freezer to user something like liquid nitrogen? I don't know if that's easily done, but it's an idea. Then you need not worry about conductivity problems, signal degredation, and problems with timing across the board.
You could cut and seal a freezer, and seal parts of the case, to allow for use of removeable media as well. The big benefit (if you ask me) would be you wouldn't void any equipment warrenties!
Many Cray's are multiple Alpha CPU's.... and standard Intel CPU's have been cooled to below -35C. *If* the silicon were cooled enough to superconduct, they might be circuitry failures. But only if the etched circuit's walls started to conduct I would think.;-)
That is a problem. Not so much if you gradually heat up the components, the same way you'd have to cool them when you immersed them.
There is a chance that you would break a connection somewhere, but most of the components are fairly well connected. You'd have to hope that they would all cool fast enough to not be a problem.
The motherboard probably would break, but regardless you'd have to be very careful. If you were putting the amount of money necessary to do this into a system, I have a feeling you wouldn't have to worry about it though.;-)
Either way, liquid cooling using N or He isn't really feasible for almost anyone.:)
Definitely true, cost would be ridiculous for using helium. A liter of 3He is about $100,000... but an interesting idea!
The N2 solution would be incredibly less expensive, but still, the required components for containing, cooling, and safety would probably be quite a lot of money!
Think of using something like this with some high-output TEK (Peltier) panels... that would be quite a cool[sic] system.
Well, you don't really have to replace the nitrogen. The hard part would be cycling it around through a cooling system... the heat let off when it's recooled would probably be quite tremendous. This is used to some extent already in many different applications. If you are going for this level of cooling, why not use liquid helium? At that tempreture your computer would nearly superconduct, and performance approximately doubles. Also, helium boils at -268.6C, vs -195.8C for nitrogen.
This sounds suspiciously like the process that SAP uses, only with "web" attached. Send a request to a central server. It then checks for the least busy ring server, and tranfers the connection to that. Occasionally the ring servers query the central server for data...
So they're going to kill the ability to play MP3's in hardware? Hmm... interesting. Just like Sony stopped anyone from playing out-of-region games or copied games on the Playstation, eh?
Here's an idea, how about we don't use TCP/IP to talk to devices in a house. How about we just give a house an IP address, and let some computer handle figuring out devices. Or we could just use TCP/IP and assing a device to a port on that IP. There are an awful lot of ports above port 1024 anyway.
Just a thought... but since the ruling said that "Movie Buff" would be OK, but not "moviebuff", I'm just wondering one thing. How in the world do you get the domain "movie buff.com", complete with the space?;-)
Slashdot Mirror
Three easy steps: 1) Get program to parse browser history for URL's 2) Awk output for domains and save to blocklist 3) cat blocklist | pine support@ezula.com
To be complaint to standards, the shortest the track can be is 4 seconds... anything different and it violates the standard.
Just a few examples of the useless stuff that we've attained from NASA and the military. Though I should mention it after reading some of these posts... After all, pure research is worthless and expensive, and if the military does it, well then, it must be bad.
Plastic, modern ceramics, nuclear power, fuel efficient car engines, wireless anything, teflon, semiconductors and superconductors, transistors, the Internet, microwave ovens, GPS, computers, engine emissions scrubbers, and more and more....
So lets cut research, great idea. Way to invent things that people haven't already thought of, because we all know that we got were we are by improving existing things.
Cant' remember OC48 pricing, but T1 runs around 1100$/mo, and OC3 is around 80,000$/mo (1.54Mb vs 155Mb).
http://wire-head.org/linux-2.4.0.tar.bz2
Sounds alright, except you never get anything unless other people have bookmarked it. So does that mean you would never find new sites? What if I were to come up with some incredible idea, and put it on the web tomorrow. How would anybody find it, except by me directly telling them? I don't publicise my sites, but Google and AltaVista and likes have them. This would never have them! Sounds like the engine would be doomed to fail from it's conception, unless they start doing something else in addition.
And if you take the digit placements (2 and 6), the difference from 0 (the first) and 2 is 2, and the difference from 2 and 6 is 4. If you invert them you get 42. Coincidence?
I don't know about the GNOME ports, but your 486/66 is a more than adequate machine. A low end 486 can easily flood a T1 or two, your cable modem isn't going to be a problem to route for. I'm using one right now for something quite similar!
The S3 Virge chipsets are supported by just about everything out there. So is the Millenium II card. Both are rather nice, the Matrox definetely being the faster of them, but if you just need good 2d, the Virge cards are cheap. If you want really nice 2d, the Millenium II is a little more expensive, but also very well supported.
SCSI allows many things that IDE doesn't do. It has a rather mature RAID controller market, not to say IDE RAID doesn't exist in hardware and software. Also, I don't remember the last hot swappable IDE drive that worked. Since I deal with many arrays of disks chained off a smaller number of computers, it would even be *possible* to have the number of discs we have. You could only get four drives in an array with IDE. That's nowhere near acceptable. Plus since you can't do command queueing and simultaneous transactions on IDE, the performance is crap for it. Plus, since the cables have length limitations, especially for ATA/66, it's no good for external bays. IDE is nice and all, but I would never rely on it.
As for WD, just had *two* drives fail within five hours of each other from them. One a year newer than the other. Don't miss them a bit, perfectly happy with Fujitsu and IBM right now.
Most Linux products have what, an average of six - ten updates a year? Take a look at Freshmeat. Some things are updated every few days. It wouldn't work very well in real life. I'd rather go with the standard versioning. It's a lot easier to deal with 4.0, 4.1, etc than 95, 95 OSR2, 95 OSR2.5, 95OSR-Your_mama. The actual version number shows how little difference there was between products: Win95 (4.0), Win98 (4.1). The same generally hold true for other things. A point release tends to not add features (well, unless it's pre v1.0), just fix problems. But a full version adds things. IE: 2.0.38 differs from 2.2.13, but much of the underlying code stays the same, it wasn't a rewrite. 3.1 is much different from 4.0/4.1. See the point? So we can have Linux 2000, Linux 2000.2.13, or Linux 2.2.13. The smaller one looks more manageable.
Quite the contrary, UDP is very useful still. With multimedia you don't need to guarentee that all the packets get to the destination, just that the packets are only used in the correct sequence. You drop packets that arrive late. With standard TCP net congestion can really kill your stream due to all the retries to make sure everything gets where it's going.
;-) There are already a few existing servers, do a few searches on freshmeat. However, something that Apple wants to include in their flagship server software should be carefully looked at too.
This isn't always the best, for example, if you require everything gets to the destination. Even then, though, if you have a generously sized buffer, you can store packets and request for a missed one, thus potentionally saving time on retransmits. This can be useful in pausing a stream temporarily as well.
This makes UDP very good to use for real-time streaming. Besides, any contribution to open source should be welcomed... if you don't like the contribution, just don't use it!
The other possible reason is that you can't afford to get sued by the five homeowners between you and your access point. Also, the network admin at that access point had some sort of aversion to a static discharge (read lightning or some like punk messing with the cable) ruining his many thousands of dollars of network equipment in his network closet. What's the tradeoff here, potential for $24,000 in damage because we're lazy and run cable, or play it safe and use wireless...
Plus, you might want to ask those neighbors about your cable, or it'll be interesting the next time they get a nice we lawn and your cable shows through
A bunch of friends and I were looking into this idea. The problem comes in with a few things. First, you need chips to modulate your laser from your ethernet pulsing, and also to decode and amplify from the laser recieve photosensor. As far as line-of-sight, it isn't bad, but *nothing* can be in the way. If you use a class 3 laser (laser pointers are part of this) you are generally safe, as long as people never are able to look at the beam. If you go any higher than class 3, you have safety concerns...
Also, you need a fairly good laser diode to transmit with, just because you can't permit the beam to diffuse very much, or you'll get signal loss. This can be compensated for with different types of beam modulations, but still can be a bit daunting.
The chips we were looking at were the same they used in OC-48 fiber terminations, but unfortunately I dont' remember the company.
As fun as the talk about liquid cooling is, wouldn't it be easier and safer to use a low tempreture liquid to cool the air instead? For instance, a typical freezer uses Freon; what if you adapted a freezer to user something like liquid nitrogen? I don't know if that's easily done, but it's an idea. Then you need not worry about conductivity problems, signal degredation, and problems with timing across the board.
You could cut and seal a freezer, and seal parts of the case, to allow for use of removeable media as well. The big benefit (if you ask me) would be you wouldn't void any equipment warrenties!
Just a thought though...
Many Cray's are multiple Alpha CPU's.... and standard Intel CPU's have been cooled to below -35C. *If* the silicon were cooled enough to superconduct, they might be circuitry failures. But only if the etched circuit's walls started to conduct I would think. ;-)
I am probably wrong, so correct me!
What if something *did* go wrong in this setup? That would be a pretty dangerous "oops"...
That is a problem. Not so much if you gradually heat up the components, the same way you'd have to cool them when you immersed them.
;-)
There is a chance that you would break a connection somewhere, but most of the components are fairly well connected. You'd have to hope that they would all cool fast enough to not be a problem.
The motherboard probably would break, but regardless you'd have to be very careful. If you were putting the amount of money necessary to do this into a system, I have a feeling you wouldn't have to worry about it though.
Either way, liquid cooling using N or He isn't really feasible for almost anyone. :)
;-)
Definitely true, cost would be ridiculous for using helium. A liter of 3He is about $100,000... but an interesting idea!
The N2 solution would be incredibly less expensive, but still, the required components for containing, cooling, and safety would probably be quite a lot of money!
Think of using something like this with some high-output TEK (Peltier) panels... that would be quite a cool[sic] system.
(NMR.. yummy
Well, you don't really have to replace the nitrogen. The hard part would be cycling it around through a cooling system... the heat let off when it's recooled would probably be quite tremendous. This is used to some extent already in many different applications. If you are going for this level of cooling, why not use liquid helium? At that tempreture your computer would nearly superconduct, and performance approximately doubles. Also, helium boils at
-268.6C, vs -195.8C for nitrogen.
This sounds suspiciously like the process that SAP uses, only with "web" attached. Send a request to a central server. It then checks for the least busy ring server, and tranfers the connection to that. Occasionally the ring servers query the central server for data...
--AaronL
So they're going to kill the ability to play MP3's in hardware? Hmm... interesting. Just like Sony stopped anyone from playing out-of-region games or copied games on the Playstation, eh?
Here's an idea, how about we don't use TCP/IP to talk to devices in a house. How about we just give a house an IP address, and let some computer handle figuring out devices. Or we could just use TCP/IP and assing a device to a port on that IP. There are an awful lot of ports above port 1024 anyway.
Great, now it'll be illegal to write code without a license? I can't change a room in my house without government approval as it is!
Just a thought... but since the ruling said that "Movie Buff" would be OK, but not "moviebuff", I'm just wondering one thing. How in the world do you get the domain "movie buff.com", complete with the space? ;-)
-Aaron