Intel would love to use a CPU ID to help us. This has so many problems that I'm just not going to go into it. But it would work.
Many boxes already have HostID numbers, but the PC world has not had this, traditionally. I have not heard any complaints from Sun users, apart from when they need to fool a license server somewhere - but in a PC a hostid is suddenly something bad? (Well, you can change the ID in a Sun, but if it's etched into the CPU it might be harder to change.)
But what I was going to say was that it wouldn't work, since a CPUID/HostID identifies a machine, and not a user. Think about multiuser environments. Think about typical university setups, where you can sit at any one of hundreds of machines. Or when you set up individual accounts for each member of your family on your home box.
No, we absolutely need something that follows the user, and not the machine. I don't see any better solutions than cookies, but I'm sure that there are some.
"The 600-series had rubber skin, we spotted them easy..."
I say, these snakes need glowing red eyes! Don't use rubber skin, though, or they'll be spotted immediately. But they don't look like they'll ever be able to hold a plasma rifle, snakes typically have too few hands...
That documentary was shown here in Sweden too (in "Vetenskapens Värld", a while ago). It was a lot of stuff on how the cars got safer over time - for a long time the attitude was along the lines of "if the driver is stupid enough to crash, of course he must be hurt". All the blame was put on the driver, even if it wasn't his fault at all.
Nowadays, however, cars are designed in a much more sensible way, that is, "if you crash, the car will protect you as much as possible". It might not have been your fault that you crashed, after all...
One language which actually is translated depending on which language version you run is Visual Basic for Applications. If you have the English Office package, the command for opening a file is FileOpen(), but if you have the Swedish Office package, it is called FilÖppna() or something (I'm not really sure here, I don't use it myself).
This internationalization of the language might have seemed like a good idea at the time, but it makes it hopeless to create macros that will run on any version of Micros~1 Office. Incidentally, it also prevents some macro virii, which will crash when the functions aren't named the same...:-)
I have tried some Swedishified languages, and I really hate them. Your'e just too used to writing "ifxgotoy" instead of "omxgå tilly". The fact that you sometimes use Swedish identifiers is another story altogether, though...
I didn't understand this at all, until I read the article and found out that they didn't talk about my part of the world, but rather an object in space with a very similar name.
FYI, "Europe" is spelled "Europa" in Swedish (and a few other languages too).
Did anyone else notice the bit at the end of the license, going on about Gnomovision, Yoyodyne, James Hacker and Ty Coon etc?
At least I found this very amusing!:-) I mean, 3Com is a big company, and one would assume that someone from their legal department would at least read the license which they are using...
First of all, the power lines are not intended for data transfer. They are intended for power transfer. This means that:
There is no shielding. Disturbances leak in at free will, and also they leak out (which means radio interference).
There are no filters. A TV, microwave oven or a welding unit generate lots and lots of disturbances, which goes out on the grid unfiltered (well, to the nearest transformer station anyway).
Different conditions for each house. The line from the last transformer station (where the data is moved to/from a real data communications line) to the individual houses can be ten meters or 500 meters. There can be one household or thirty on the same line. This means that the conditions vary wildly from house to house, and it is very hard to create a transmitter/receiver that will work everywhere, since they are typically optimized for a fixed line condition.
These things, and several others, means that yes, you can communicate over the power lines, if you have a short line to the transformer, and if your neighbour doesn't turn on his TV, and if you don't care about airplanes dropping from the sky because you confused their tracking system, and so on.
There are a few test systems in use here in southern Sweden, just intended for reading the power meter remotely. These systems communicate with around 1200 bits/second, and use small packages (I think it was less than 40 bytes). Each packet is retransmitted up to four times. Even so, they have an average packet drop ratio of 40 - 100 percent. Yes, that's 100% corrupted packages.
My opinion about all this is: Keep power in power lines, data in data lines and phone calls in phone lines. That's what they were designed for.
I just wonder why the geographical domains are not more enforced than they are? The.us domain is heavily underused. Suggestion: Make US-based companies use domains like.ca.us,.mn.us,.tx.us and so on, depending on where they have their headquarters. That would resolve a lot of the ambiguities out there. Reserve the.com domain for companies with representation in at least n countries, where n could be somewhere around 10 or so.
A similar system is actually used here in Sweden, "small" companies can only register under.a.se,.m.se and so on (where the letter indicates in which part of the country the company is located). Likewise, private persons can only register under.pp.se . Seems to work so far.
Another thing which would make the domains more easily handled is to make it harder to register product names as domains. Is it, for example, that much harder to write "www.company.com/product" than to write "www.product.com"? I certainly prefer the first variant. Some companies seem to register each and every product as their own domains, and that leads to a very cluttered namespace.
Bottom line is, you just can't allow anybody to register any domain they want like now - there are only so many letter combinations before they start getting incomprehensible.
I find my left hand aiming for a non-existant "Front" key much too often. It's the best key ever invented! Just point at your window and press "Front", no need to click in some stupid title bar or anything, which may not even be visible. Press "Front" once more, and back goes your window. I love it.
And what happened to "Do", "Help", "Prior", "Next" and "Setup"? I used DECStations some years ago, and had "Do" mapped to M-x in Emacs. Cool.
My ideal keyboard would be something like a standard 105-key PC keyboard, with the left F-key row from a Sun 5c and a few others from a standard DEC keyboard, with the feeling of the IBM XT/AT-keyboards, and with some custom keymapping... Well, one can always dream.
Just a minor correction: That should read "and | I get with AltGr+<" (the key between left shift and Z).
Also, AltGr is (was?) labled "Alt Graph" on some keyboards, for "alternate graphics". That's how I always pronounced it.
But it does make it a pain to get to the prompt in telnet et al, which requires that you press Ctrl-AltGr-9 to get the ^] combo. Some keyboard can't handle this, and only beep at you when you try to press it. Mostly true for older keyboards, though.
Whenever I hack C code, I use the CapsLock key when I have some DEFINED_LONG_UGLY_CONSTANT to be written. It does save a lot of time and finger strain to just toggle the capses.
Unfortunately they only sell the new IBM-style keyboards with the US key layout. If they made a Swedish (or Finnish, for that matter) 102-key keyboard, I would order one right now.
I used IBM PS/2 model 30 (and the original IBM XT, with a 20 MB HD (full-height, do you ever see a full-height unit nowadays?) and an 8088) around 1990-1992, and I loved the keyboards.
There is plenty of reasons for going above 128MB swap. I have (well, the department here has, but I take care of them) a cluster of dual-PII:s with 256MB physical memory and 1GB swap each. I have seen many programs here eating around 600-700 MB of memory, the top notation was a Matlab script using close to 1 GB of memory. Yes, the script was carefully written to avoid thrashing. On the other hand, I have also seen programs using "only" about 400 MB that thrashed like h*ll.
Yes, we're about to buy 1 GB of memory per machine, we really need it. Problem is that with these fscking PC:s you can't have more than 1 GB. We don't really want to buy a Sun Enterprise or an Alpha, since we're trying to move to all PC:s, so I guess we'll have to wait for the Xeon machines to become common.
Our department is into research in coding theory and cryptography among other things, and the simulations for this kind of stuff are huge. One of our professors want to run a program which needs 4 GB of memory and 20 GB swap, and so far we haven't found any machine which can manage that.
Special tuning has always been a part of benchmarking. There are many compilers out there with special optimizations designed only to improve the performance of a specific benchmark (detection of certain matrix operations and other snippets of code which almost exclusively occur in benchmark programs). Many compilers, for example, recognize this piece of code:
x = sqrt(exp(log(x)/t1))
which gets transformed to
x = exp(log(x)/(2*t1))
thereby skipping the sqrt calculation. Since this code almost only occurs in the Whetstone benchmark, in the general case it will never get used. (Hennessy & Patterson, CA:AQA, chap. 1.8)
Some of the compiler optimizations out there will even generate faulty code if they are switched on when not compiling a specific benchmark program. Might be worth thinking about.
And don't think that only software can be optimized. Many of the processors out there have some special hardware hacks designed only to improve some benchmark program, and which almost never is used in "real world" programs.
Bottom line is, don't be surprised that they use some special tuning, it has always been this way and always will.
Not trying to sound ignorant or anything, but what is this "IPO" thingy? And the talk about "S-1"?
I understand it has something to do with stocks, but I don't know anything about how this stuff works over there. If someone could just expand the acronyms for me it might help...
Well, actually I do have a copy of Win3.1/DOS 5, but hey, I was young and foolish:-) And Linux was pretty unknown in 1992, anyway.
When I bought a new PPro a couple of years ago, I specifically asked for "No operating system please, thank you very much!", and they agreed without questions.
Slashdot Mirror
Intel would love to use a CPU ID to help us. This has so many problems that I'm just not going to go into it. But it would work.
Many boxes already have HostID numbers, but the PC world has not had this, traditionally. I have not heard any complaints from Sun users, apart from when they need to fool a license server somewhere - but in a PC a hostid is suddenly something bad? (Well, you can change the ID in a Sun, but if it's etched into the CPU it might be harder to change.)
But what I was going to say was that it wouldn't work, since a CPUID/HostID identifies a machine, and not a user. Think about multiuser environments. Think about typical university setups, where you can sit at any one of hundreds of machines. Or when you set up individual accounts for each member of your family on your home box.
No, we absolutely need something that follows the user, and not the machine. I don't see any better solutions than cookies, but I'm sure that there are some.
CyberSnake Model 101, anyone?
"The 600-series had rubber skin, we spotted them easy ..."
I say, these snakes need glowing red eyes! Don't use rubber skin, though, or they'll be spotted immediately. But they don't look like they'll ever be able to hold a plasma rifle, snakes typically have too few hands...
(Apologies to James Cameron)
Forbidden
You don't have permission to access / on this server.
Apache/1.3.9 Server at gpltrans.zzweb.com Port 80
Same result with the other link... This makes it a wee bit hard to check out the site. Are there any mirrors out there?
That documentary was shown here in Sweden too (in "Vetenskapens Värld", a while ago). It was a lot of stuff on how the cars got safer over time - for a long time the attitude was along the lines of "if the driver is stupid enough to crash, of course he must be hurt". All the blame was put on the driver, even if it wasn't his fault at all.
Nowadays, however, cars are designed in a much more sensible way, that is, "if you crash, the car will protect you as much as possible". It might not have been your fault that you crashed, after all...
One language which actually is translated depending on which language version you run is Visual Basic for Applications. If you have the English Office package, the command for opening a file is FileOpen(), but if you have the Swedish Office package, it is called FilÖppna() or something (I'm not really sure here, I don't use it myself).
This internationalization of the language might have seemed like a good idea at the time, but it makes it hopeless to create macros that will run on any version of Micros~1 Office. Incidentally, it also prevents some macro virii, which will crash when the functions aren't named the same... :-)
I have tried some Swedishified languages, and I really hate them. Your'e just too used to writing "if x goto y" instead of "om x gå till y". The fact that you sometimes use Swedish identifiers is another story altogether, though...
I didn't understand this at all, until I read the article and found out that they didn't talk about my part of the world, but rather an object in space with a very similar name.
FYI, "Europe" is spelled "Europa" in Swedish (and a few other languages too).
Did anyone else notice the bit at the end of the license, going on about Gnomovision, Yoyodyne, James Hacker and Ty Coon etc?
At least I found this very amusing! :-) I mean, 3Com is a big company, and one would assume that someone from their legal department would at least read the license which they are using...
First of all, the power lines are not intended for data transfer. They are intended for power transfer. This means that:
Disturbances leak in at free will, and also they leak out (which means radio interference).
A TV, microwave oven or a welding unit generate lots and lots of disturbances, which goes out on the grid unfiltered (well, to the nearest transformer station anyway).
The line from the last transformer station (where the data is moved to/from a real data communications line) to the individual houses can be ten meters or 500 meters. There can be one household or thirty on the same line. This means that the conditions vary wildly from house to house, and it is very hard to create a transmitter/receiver that will work everywhere, since they are typically optimized for a fixed line condition.
These things, and several others, means that yes, you can communicate over the power lines, if you have a short line to the transformer, and if your neighbour doesn't turn on his TV, and if you don't care about airplanes dropping from the sky because you confused their tracking system, and so on.
There are a few test systems in use here in southern Sweden, just intended for reading the power meter remotely. These systems communicate with around 1200 bits/second, and use small packages (I think it was less than 40 bytes). Each packet is retransmitted up to four times. Even so, they have an average packet drop ratio of 40 - 100 percent. Yes, that's 100% corrupted packages.
My opinion about all this is: Keep power in power lines, data in data lines and phone calls in phone lines. That's what they were designed for.
I just wonder why the geographical domains are not more enforced than they are? The .us domain is heavily underused. Suggestion: Make US-based companies use domains like .ca.us, .mn.us, .tx.us and so on, depending on where they have their headquarters. That would resolve a lot of the ambiguities out there. Reserve the .com domain for companies with representation in at least n countries, where n could be somewhere around 10 or so.
A similar system is actually used here in Sweden, "small" companies can only register under .a.se, .m.se and so on (where the letter indicates in which part of the country the company is located). Likewise, private persons can only register under .pp.se . Seems to work so far.
Another thing which would make the domains more easily handled is to make it harder to register product names as domains. Is it, for example, that much harder to write "www.company.com/product" than to write "www.product.com"? I certainly prefer the first variant. Some companies seem to register each and every product as their own domains, and that leads to a very cluttered namespace.
Bottom line is, you just can't allow anybody to register any domain they want like now - there are only so many letter combinations before they start getting incomprehensible.
I find my left hand aiming for a non-existant "Front" key much too often. It's the best key ever invented! Just point at your window and press "Front", no need to click in some stupid title bar or anything, which may not even be visible. Press "Front" once more, and back goes your window. I love it.
And what happened to "Do", "Help", "Prior", "Next" and "Setup"? I used DECStations some years ago, and had "Do" mapped to M-x in Emacs. Cool.
My ideal keyboard would be something like a standard 105-key PC keyboard, with the left F-key row from a Sun 5c and a few others from a standard DEC keyboard, with the feeling of the IBM XT/AT-keyboards, and with some custom keymapping... Well, one can always dream.
Just a minor correction: That should read "and | I get with AltGr+<" (the key between left shift and Z).
Also, AltGr is (was?) labled "Alt Graph" on some keyboards, for "alternate graphics". That's how I always pronounced it.
But it does make it a pain to get to the prompt in telnet et al, which requires that you press Ctrl-AltGr-9 to get the ^] combo. Some keyboard can't handle this, and only beep at you when you try to press it. Mostly true for older keyboards, though.
Whenever I hack C code, I use the CapsLock key when I have some DEFINED_LONG_UGLY_CONSTANT to be written. It does save a lot of time and finger strain to just toggle the capses.
Unfortunately they only sell the new IBM-style keyboards with the US key layout. If they made a Swedish (or Finnish, for that matter) 102-key keyboard, I would order one right now.
I used IBM PS/2 model 30 (and the original IBM XT, with a 20 MB HD (full-height, do you ever see a full-height unit nowadays?) and an 8088) around 1990-1992, and I loved the keyboards.
There is plenty of reasons for going above 128MB swap. I have (well, the department here has, but I take care of them) a cluster of dual-PII:s with 256MB physical memory and 1GB swap each. I have seen many programs here eating around 600-700 MB of memory, the top notation was a Matlab script using close to 1 GB of memory. Yes, the script was carefully written to avoid thrashing. On the other hand, I have also seen programs using "only" about 400 MB that thrashed like h*ll.
Yes, we're about to buy 1 GB of memory per machine, we really need it. Problem is that with these fscking PC:s you can't have more than 1 GB. We don't really want to buy a Sun Enterprise or an Alpha, since we're trying to move to all PC:s, so I guess we'll have to wait for the Xeon machines to become common.
Our department is into research in coding theory and cryptography among other things, and the simulations for this kind of stuff are huge. One of our professors want to run a program which needs 4 GB of memory and 20 GB swap, and so far we haven't found any machine which can manage that.
Special tuning has always been a part of benchmarking. There are many compilers out there with special optimizations designed only to improve the performance of a specific benchmark (detection of certain matrix operations and other snippets of code which almost exclusively occur in benchmark programs). Many compilers, for example, recognize this piece of code:
which gets transformed to thereby skipping the sqrt calculation. Since this code almost only occurs in the Whetstone benchmark, in the general case it will never get used. (Hennessy & Patterson, CA:AQA, chap. 1.8)Some of the compiler optimizations out there will even generate faulty code if they are switched on when not compiling a specific benchmark program. Might be worth thinking about.
And don't think that only software can be optimized. Many of the processors out there have some special hardware hacks designed only to improve some benchmark program, and which almost never is used in "real world" programs.
Bottom line is, don't be surprised that they use some special tuning, it has always been this way and always will.
Not trying to sound ignorant or anything, but what is this "IPO" thingy? And the talk about "S-1"?
I understand it has something to do with stocks, but I don't know anything about how this stuff works over there. If someone could just expand the acronyms for me it might help...
Did they remove the limit?
>you should also add some sort of weather box.
How about an xearth with customizable view coordinates, so you can easily check whether the sun is up or not where you are...
Well, actually I do have a copy of Win3.1/DOS 5, but hey, I was young and foolish :-) And Linux was pretty unknown in 1992, anyway.
When I bought a new PPro a couple of years ago, I specifically asked for "No operating system please, thank you very much!", and they agreed without questions.
/B