devices is that they change the source address of the packets they send so that it looks like the data all comes from the device itself.
The thing is that for instance a Linux box normally start using port number 1024 and up. So if you see port numbers in the 1500's that's cool. If they go up to 5000, you say "wow that machine has a long uptime".
But when masquerading the Linux kernel uses port numbers above 60000. So if you see port numbers 60123 and the like, it's very likely that the person is using a Linux machine to do Masquerading. How other routers/NAT systems would do this, I don't know.
Of course, I'm also the guy that turned down a job offer because they wanted me to get a haircut. I just can't stand to work in any sort of environment that doesn't value the individual.
There are several things about the question: "Could you get a haircut?".
First, they are indirectly asking you if you'd be willing to conform on little things to the company culture. It can very well be that "Sure, if you really, really want to" is a valid answer, and that actually getting the haircut is not required.
I'm told that you'd end up getting the haircut anyway 6 to 12 months later, but that is NOT because your boss asks you to.
On the other hand, it is an indication of the kind of company culture that they have. And you're right that you can take it as a warning.
If it's a national security issue then the machines shouldn't be on the internet.
No, no. It's not "being on the internet" that's a problem. It's that some highly untrusted individual had a chance to insert malicious code into the OS.
Like "if you dertermine yourself to be in a position to fire a missile: fire the missile".
The danger is that this might go undetected for years until we learn "the hard way".
There would be so much memory to swap with a (relativly) very small disk bandwidth
Ehmm. Very small disk bandwidth? I expect every node to have its own physical disk should there be a need to add swap. In that case you can expect every node to have about 40Mb per second worth of swapping potential. The machine would have a total of about 50Gb per second of disk throughput.
(Aren't you happy that your PC can do 800Mb per second to RAM? this thing does 60 times that to DISK!)
Speed of light: 3*10^8 m/s (approx)
Clock tick: 1*10^-9 s (1 GHz)
Distance traveled: 3*10^8*10^-9 = 0.3 m
Right! But electrical signals travelling electrical wires don't do "speed-of-light". In my student room, about 10 years ago, I measured about one third the speed-of-light, but most people quote about twice that.
When you turn on a magnet, electrically conducting objects will have to encompass the magnetic field. This induces a current. If the electrically conducting object is superconducting the electricity that will flow will completely cancel the magnetic field, and the electric current will continue to repell the magnetic field until something external happens. You can balance a superconducting piece of material above a magnet using this effect.
If you have a non-superconducting piece of iron, and you turn a strong magnet on, the object will jolt. This is a force between the magnet and the iron. So the magnet will be repelled by the iron and vice versa. Nothing mysterious. Nothing that will drive a spaceship.
Some guy saying that "there is unresolved physics" does not make me dismiss a century of well-verified physics.
Huh? Most modern CDs are flexible enough to wrap around a soca can. Try it.
and be inserted into magazines without breaking,
Exactly! Ever found a broken CD in a magazine? Didn't think so!
Roger.
Re:Suprise isn't necessary, but worthy ending is.
on
Review: "Unbreakable"
·
· Score: 1
...endings....... the best movies to come out recently tho: Seven, Fight Club, the Sixth Sense.
In the Sixth Sense, then ending is good, because suddenly it changes the meaning of the whole movie.
For example, I had remembered that the Doc drove the kid to the funeral in his car. Had to go see the movie a second time to realize that the kid and the doc took the bus.....
Maybe "on a global scale, the internet will remain connected", however, as a user, my ISP has one or a few connections "to THE INTERNET". If that node breaks down (here in the netherlands: "the amsterdam Internet Exchange") I'm cut off from the net.
So, what use is a "globally connected net" if 99% of the users are cut off, as well as 99% of the web servers?
There is a 99% chance that google is cut off, so that you can't search the net with google anymore. If you're lucky, google still works, but returns 99% broken links.
Maybe from a theoretical viewpoint there will still be a link from Europe to America, however, "the net" will be completely useless.
(argument: but many ISPs have more links to the internet than one. Sure, ok, but there are more nodes than one involved before you get a useful link through a secondary link. In the end I estimate that still chances are less than 1% that you'll be connected, even if you have an ISP that has multiple redundant links....)
There is an electrical effect called the "transmission line effect".
A cable can be modelled (for high-speed signals) as a series of inductors, resistors and capacitors.
If a certain relation between the values of the inductors, and capacitors exists, the signal will surprisingly come out the other side, very similar to how it went in. Take a garden hose and shake it up and down. You can make the waves "walk" through the hose. That's the effect.
This effect is most clearly observed with coax cables. There are several "waves" of TV signal on the wire, even if there is only a few yards of cable between your wall-socket and the TV.
Now, in a ribbon cable, a signal that has a "ground" conductor on each side of a signal, this effect is noticable enought to be useful.
For UDMA33, and especially higher, the transmission line effect becomes noticable, and required for correct operation.
If you just pack a bunch of lose cables in a bundle with tiewraps or stuff like that, this will not guarantee that there are ground conductors next to the signal wires.
As an experiment: Try making two round cables. One with the odd and one with the even numbered conductors. You certainly shouldn't be able to run UDMA66 anymore.
I really don't know who you talked to at WD. They DO have people who know this, but apparently the guy you talked to didn't.
Oh, you will probably be able to get away with splitting the cable in pairs. But for the tranmsission line effect to really work, the fact that there is a ground line on the other side too really helps.
I'm sorry, but I've been wondering about this for a few years now(*). Suppose I were to route the whole IP number space as class C networks. That means 2^24, or 16M of "routes". Now even if my router happens to have about 100 different network interfaces, I can still hold that in 8 bits. So with 16Mb of memory I can hold my routing table.
Now a route lookup is equivalent to
itf = route_table[dest_ip >> 8];
That's going to take around 60ns on a modern PC.
So if that's all, we'd be able to do around 13M routing decisions per second. That's not bad. (you'd be routing over a gigabyte per second by the time that this could start to become a bottleneck...)
The only problem with this method is that when a class-A route changes, you have to update 65536 routing table entries. This can be solved by having a multi-level table.
You'll probably have to have a few "exceptions": Someone is bound to have split up a C class network that you route it over different interfaces. Simple: An exception "interface" that indicates: "try the exceptions routing table".
Roger.
(*) This subject keeps popping up in the media every year or so...
software is loaded at boot time by the BIOS and then the "door is closed" so that no other changes can occur until the system is reset.
Nope. Intel Microcode loads work as long as the new microcode has a higher revision than the old one.
Oh, and the processor verifies the "intel signature" on the microcode before it accepts the new microcode. But I doubt that they have RSA public key cryptography in those processors, so I would expect simple cryptanalysis to reveal the signature.
I have more problems with SPAM from large corporations than I do from subject: XXX
I have a mailbox labelled "spam" that my mail-filter delivers all "probably spam" stuff into.
These big corporations that end up spamming me end up with a special procmail rule that diverts their shit into that mailbox. If I really want I can still read it. I don't lose email when it misfires and it's satisifying to open that mailbox and see a list of mails that I would've discarded without reading anyway.
IBM, SGI, Trendmicro, itselling, webknowledgebase, shopping planet all have a private entry...
Well, I guess you all see why the Prof gave me a C in that class.
Hihi. I got a grade (6) similar to a C too. I got "best 5%" on the test, but didn't get the "almost for free" bonus points that you got by writing a report about solving one of the problems in the book with a group of 3 students.
I solved a problem NOT in the book (i.e. a NEW problem) by myself and they didn't give credit for that. I'm a guy who doesn't get upset about that.
What a load. Pennies and nickels are legal tender according to popular myth debunking site, snopes.com.
Hmmm. Here in the Neterhlands, I thought that you can pay up to about hfl 10,- in dimes. So for instance if you get a speeding ticket and want to pay that, you'll not be allowed to pay that in dimes and nickels (as the amout is going to be well above hfl 10,-)
I'm usually at the other end than the person who asked the question here.
I get asked to write Linux drivers for various hardware. The guys who made the chip and their colleagues have the closest feeling with the chip and its interfaces. However, I have intimate knowledge of Linux.
So when you develop "in-house", there is one advantage, if you leave it to the "linux experts" there is another advantage.
So the question is: which way do the scales tip?
Brook's law doesn't apply if you add programmers at the beginning. Thinking about the project with a medium-sized team will make the design better. This will reduce development time.
If you start with one, two or three programmers, and later start adding more and more programmers because you're running late, you'll find out first hand about brook's law.
Also, a company wanting to start supporting Linux should hire us because if you hire a person to "also" do the Linux driver, soon, he'll be doing nothing else. So supporting Linux would cost you a full year-salary per year. We offer MUCH cheaper maintenance and support contracts. And for that money you can have one or two drivers developed every year too!
Roger.
If there is no clear maintainer for the part you're patching, try this:
(Otherwise you're dealing with a "random" maintainer, who is usually more responsive than Linus)
- Post on Linux-kernel. Ask people to test your
patch.
- wait a week
- post patch again, this time with "so-many people tested it. Please apply." CC Linus or Alan this time.
- Wait a week. Or at least 2 new (minor) releases. This prevents you from sending things twice when the maintainer happened to be "out of the office" for a moment.
- If it's still not included, send again.
Alan will usually respond with "this is bullshit" if it is, on the first try. Linus kind of assumes that you know that, and ignores you. Politely keep on asking for feedback. Send it again. And again (at the right pace: 2 releases or a week in between). He'll get annoyed at you, and finally tell you to shut up because your patch is bullshit. When he gets annoyed at having to throw your Email away every time, he'll give you feedback about what's wrong, and you'll be able to fix your patch.
So, if you hurry, you'll have feedback after about 3 weeks, and a "fixed" patch by week 4. It does take some "time" in that you have to keep coming back every week to see if your patch got accepted. One Email in december, one in june is "slow" enough to give the impression that you don't really care, and you might get ignored.
Also, Linus and Alan get soooo much Email that if they go away for a few days, they will have a hard time catching up, so are coping by just deleting everything that came in while they were away. If it was important, you're supposed to resend it.
Slashdot Mirror
devices is that they change the source address of the packets they send so that it looks like the data all comes from the device itself.
The thing is that for instance a Linux box normally start using port number 1024 and up. So if you see port numbers in the 1500's that's cool. If they go up to 5000, you say "wow that machine has a long uptime".
But when masquerading the Linux kernel uses port numbers above 60000. So if you see port numbers 60123 and the like, it's very likely that the person is using a Linux machine to do Masquerading. How other routers/NAT systems would do this, I don't know.
Roger.
Of course, I'm also the guy that turned down a job offer because they wanted me to get a haircut. I just can't stand to work in any sort of environment that doesn't value the individual.
There are several things about the question: "Could you get a haircut?".
First, they are indirectly asking you if you'd be willing to conform on little things to the company culture. It can very well be that "Sure, if you really, really want to" is a valid answer, and that actually getting the haircut is not required.
I'm told that you'd end up getting the haircut anyway 6 to 12 months later, but that is NOT because your boss asks you to.
On the other hand, it is an indication of the kind of company culture that they have. And you're right that you can take it as a warning.
Just my two cents.
Roger.
Trademarks have to be protected, no matter how little you care, or else they will become invalid and anyone can use them.
So... Grant them a licence.
I once got a funny postcard: A cola-can made flat.
the guy sending it to me had indeed put on enough postage.
The fun is that when it got here, it had obvious marks on it that it had gone through an automated mail-sorting machine...... and jammed there.
:-)
Roger.
This sounds like that story in which RAM was made that could store 4 values instead of 2
by using the same technology described here.
That didn't make it. Will this?
Didn't make it? If I understand things correclty, this is very common for flash chips nowadays!
Roger.
If it's a national security issue then the machines shouldn't be on the internet.
No, no. It's not "being on the internet" that's a problem. It's that some highly untrusted individual had a chance to insert malicious code into the OS.
Like "if you dertermine yourself to be in a position to fire a missile: fire the missile ".
The danger is that this might go undetected for years until we learn "the hard way".
Roger.
There would be so much memory to swap with a (relativly) very small disk bandwidth
Ehmm. Very small disk bandwidth? I expect every node to have its own physical disk should there be a need to add swap. In that case you can expect every node to have about 40Mb per second worth of swapping potential. The machine would have a total of about 50Gb per second of disk throughput.
(Aren't you happy that your PC can do 800Mb per second to RAM? this thing does 60 times that to DISK!)
Roger.
Says a factor of 1000 every 20 years. That comes to a factor of 10 every 6 years or so.
Pushing the issue, Intel will claim they hit 1G last year (in the lab).
I predict that 10G will be "in the shop" around summer 2007.
Roger.
Speed of light: 3*10^8 m/s (approx)
Clock tick: 1*10^-9 s (1 GHz)
Distance traveled: 3*10^8*10^-9 = 0.3 m
Right! But electrical signals travelling electrical wires don't do "speed-of-light". In my student room, about 10 years ago, I measured about one third the speed-of-light, but most people quote about twice that.
Roger.
This is completely bullshit.
When you turn on a magnet, electrically conducting objects will have to encompass the magnetic field. This induces a current. If the electrically conducting object is superconducting the electricity that will flow will completely cancel the magnetic field, and the electric current will continue to repell the magnetic field until something external happens. You can balance a superconducting piece of material above a magnet using this effect.
If you have a non-superconducting piece of iron, and you turn a strong magnet on, the object will jolt. This is a force between the magnet and the iron. So the magnet will be repelled by the iron and vice versa. Nothing mysterious. Nothing that will drive a spaceship.
Some guy saying that "there is unresolved physics" does not make me dismiss a century of well-verified physics.
Roger.
is also flexible enough to wrap around soda cans
Huh? Most modern CDs are flexible enough to wrap around a soca can. Try it.
and be inserted into magazines without breaking,
Exactly! Ever found a broken CD in a magazine? Didn't think so!
Roger.
...endings... .... the best movies to come out recently tho: Seven, Fight Club, the Sixth Sense.
In the Sixth Sense, then ending is good, because suddenly it changes the meaning of the whole movie.
For example, I had remembered that the Doc drove the kid to the funeral in his car. Had to go see the movie a second time to realize that the kid and the doc took the bus.....
Roger.
Maybe "on a global scale, the internet will remain connected", however, as a user, my ISP has one or a few connections "to THE INTERNET". If that node breaks down (here in the netherlands: "the amsterdam Internet Exchange") I'm cut off from the net.
So, what use is a "globally connected net" if 99% of the users are cut off, as well as 99% of the web servers?
There is a 99% chance that google is cut off, so that you can't search the net with google anymore. If you're lucky, google still works, but returns 99% broken links.
Maybe from a theoretical viewpoint there will still be a link from Europe to America, however, "the net" will be completely useless.
(argument: but many ISPs have more links to the internet than one. Sure, ok, but there are more nodes than one involved before you get a useful link through a secondary link. In the end I estimate that still chances are less than 1% that you'll be connected, even if you have an ISP that has multiple redundant links....)
Roger.
There is an electrical effect called the "transmission line effect".
A cable can be modelled (for high-speed signals) as a series of inductors, resistors and capacitors.
If a certain relation between the values of the inductors, and capacitors exists, the signal will surprisingly come out the other side, very similar to how it went in. Take a garden hose and shake it up and down. You can make the waves "walk" through the hose. That's the effect.
This effect is most clearly observed with coax cables. There are several "waves" of TV signal on the wire, even if there is only a few yards of cable between your wall-socket and the TV.
Now, in a ribbon cable, a signal that has a "ground" conductor on each side of a signal, this effect is noticable enought to be useful.
For UDMA33, and especially higher, the transmission line effect becomes noticable, and required for correct operation.
If you just pack a bunch of lose cables in a bundle with tiewraps or stuff like that, this will not guarantee that there are ground conductors next to the signal wires.
As an experiment: Try making two round cables. One with the odd and one with the even numbered conductors. You certainly shouldn't be able to run UDMA66 anymore.
I really don't know who you talked to at WD. They DO have people who know this, but apparently the guy you talked to didn't.
Oh, you will probably be able to get away with splitting the cable in pairs. But for the tranmsission line effect to really work, the fact that there is a ground line on the other side too really helps.
Roger.
Note: This article is a piece of satire meant to brighten your day.
I hope everyone picked up on that.
I didn't notice it until you told me. I stopped reading after two paragraphs: "Bill isn't this stupid, he can't have said this".
Roger.
I'm sorry, but I've been wondering about this for a few years now(*). Suppose I were to route the whole IP number space as class C networks. That means 2^24, or 16M of "routes". Now even if my router happens to have about 100 different network interfaces, I can still hold that in 8 bits. So with 16Mb of memory I can hold my routing table.
Now a route lookup is equivalent to
itf = route_table[dest_ip >> 8];
That's going to take around 60ns on a modern PC.
So if that's all, we'd be able to do around 13M routing decisions per second. That's not bad. (you'd be routing over a gigabyte per second by the time that this could start to become a bottleneck...)
The only problem with this method is that when a class-A route changes, you have to update 65536 routing table entries. This can be solved by having a multi-level table.
You'll probably have to have a few "exceptions": Someone is bound to have split up a C class network that you route it over different interfaces. Simple: An exception "interface" that indicates: "try the exceptions routing table".
Roger.
(*) This subject keeps popping up in the media every year or so...
software is loaded at boot time by the BIOS and then the "door is closed" so that no other changes can occur until the system is reset.
Nope. Intel Microcode loads work as long as the new microcode has a higher revision than the old one.
Oh, and the processor verifies the "intel signature" on the microcode before it accepts the new microcode. But I doubt that they have RSA public key cryptography in those processors, so I would expect simple cryptanalysis to reveal the signature.
Roger.
I have more problems with SPAM from large corporations than I do from subject: XXX
I have a mailbox labelled "spam" that my mail-filter delivers all "probably spam" stuff into.
These big corporations that end up spamming me end up with a special procmail rule that diverts their shit into that mailbox. If I really want I can still read it. I don't lose email when it misfires and it's satisifying to open that mailbox and see a list of mails that I would've discarded without reading anyway.
IBM, SGI, Trendmicro, itselling, webknowledgebase, shopping planet all have a private entry...
Roger.
A PIC is genearally a one-shot deal anyway.. you can't 're-program' it.
No. A PIC is usually a reprogrammable device. However, they may use the one-time-programmable version in a VCR due to the high volumes....
Roger..
Well, I guess you all see why the Prof gave me a C in that class.
Hihi. I got a grade (6) similar to a C too. I got "best 5%" on the test, but didn't get the "almost for free" bonus points that you got by writing a report about solving one of the problems in the book with a group of 3 students.
I solved a problem NOT in the book (i.e. a NEW problem) by myself and they didn't give credit for that. I'm a guy who doesn't get upset about that.
Roger.
The *only* place where 'storage' is measured constantly in powers of two is on-chip memory.
You mean you've never seen a 65Mb RAM machine advertized?
Roger.
A natural gas power plant would be very very lucky to hit 40 % efficiency
I learnt that in school 20 years ago too.
Nowadays they seem to be able to reach 50% efficiency with a bit of luck.
Roger.
What a load. Pennies and nickels are legal tender according to popular myth debunking site, snopes.com.
Hmmm. Here in the Neterhlands, I thought that you can pay up to about hfl 10,- in dimes. So for instance if you get a speeding ticket and want to pay that, you'll not be allowed to pay that in dimes and nickels (as the amout is going to be well above hfl 10,-)
Roger.
I get asked to write Linux drivers for various hardware. The guys who made the chip and their colleagues have the closest feeling with the chip and its interfaces. However, I have intimate knowledge of Linux.
So when you develop "in-house", there is one advantage, if you leave it to the "linux experts" there is another advantage.
So the question is: which way do the scales tip?
Brook's law doesn't apply if you add programmers at the beginning. Thinking about the project with a medium-sized team will make the design better. This will reduce development time.
If you start with one, two or three programmers, and later start adding more and more programmers because you're running late, you'll find out first hand about brook's law.
Also, a company wanting to start supporting Linux should hire us because if you hire a person to "also" do the Linux driver, soon, he'll be doing nothing else. So supporting Linux would cost you a full year-salary per year. We offer MUCH cheaper maintenance and support contracts. And for that money you can have one or two drivers developed every year too! Roger.
If there is no clear maintainer for the part you're patching, try this:
(Otherwise you're dealing with a "random" maintainer, who is usually more responsive than Linus)
- Post on Linux-kernel. Ask people to test your
patch.
- wait a week
- post patch again, this time with "so-many people tested it. Please apply." CC Linus or Alan this time.
- Wait a week. Or at least 2 new (minor) releases. This prevents you from sending things twice when the maintainer happened to be "out of the office" for a moment.
- If it's still not included, send again.
Alan will usually respond with "this is bullshit" if it is, on the first try. Linus kind of assumes that you know that, and ignores you. Politely keep on asking for feedback. Send it again. And again (at the right pace: 2 releases or a week in between). He'll get annoyed at you, and finally tell you to shut up because your patch is bullshit. When he gets annoyed at having to throw your Email away every time, he'll give you feedback about what's wrong, and you'll be able to fix your patch.
So, if you hurry, you'll have feedback after about 3 weeks, and a "fixed" patch by week 4. It does take some "time" in that you have to keep coming back every week to see if your patch got accepted. One Email in december, one in june is "slow" enough to give the impression that you don't really care, and you might get ignored.
Also, Linus and Alan get soooo much Email that if they go away for a few days, they will have a hard time catching up, so are coping by just deleting everything that came in while they were away. If it was important, you're supposed to resend it.
Roger.