In fact, mature projects should have fewer developers since there's much less left to do.
Hah, hah.
Pick a project. Compare the feature set for version 1.0 (the first officially mature release) with the feature set for version 2.0.
In many projects, release 2 is comparitivly huge. (Read "The Mythical Man-Month" - the 3 version generally has to clean up the mess.)
Additionally, release 2 is about the time that a project adds extension capabilities that allow independent development: permitting many more developers to work at once without tripping over conflicts in each other's changes.
If you look at Perl as a "typical" example, there are hundreds of people who have done some development work on the language directly; but thousands who have developed library modules.
If a cracker gets control of a system that is capable of decrypting, she can trace the code and find the key.
If the decryption is done on a separate system, she can pretend to be a system that is allowed to ask for decryption services.
If the key is manually entered by an operator and only kept in the memory of a running decryption program, she can trace through the memory.
You can't make it perfect; you'll have to be satisfied with making it hard and separately working hard to prevent the system from being cracked in the first place.
(how many times have you gotten a perl error saying "You left a quote off up there at line 123"?)
That's a bad example (or a good example depending upon your point of view). Perl tells you "... probably a runaway quote starting on line 123." or words to that effect. Perl can tell you where the likely unterminated string starts but it is really hard to figure out where it should have ended.
The danger is that sometimes the auto-correction will be wrong. Not wrong enough that you notice it, just wrong enough to cause subtle bugs in the code. Refusing to compile such errors makes the programmer figure out exactly what was wrong, no matter how obvious it seems to be.
PL/1 used to fix all sorts of errors for you. It was great when it worked but it really cost you badly when it seemed to work.
You miss an important point. He provides a tool to facilitate anonymous posting.
- Spammers can misuse this to send spam.
- Whistleblowers and humans rights activists and oppressed people can use this to pass out information without getting shot or identified. There are parts of the world where complaining about your situation is a capital offence. Even is the "free" world, complaining about mistakes in a company can cost a job and your livelihood. (It is not "his friends" that he provides this facility for - so a large number of other people have also missed the point.)
Napster/Gnutella provide a tool to facilitate copying files.
- Some people can misuse them to bypass copyright restrictions.
- Other people can use them to copy files for legitimate purposes.
How is a general purpose tool that can allow publishing information about attrocities less worthy of your acceptance than a general purpose tool for file copying?
I'm seeing lots of people attacking the idea of them obfuscating GPL'd source code that they have gotten form somewhere else and used. That is irrelevant - the original GPL source is still available from whereever the company picked it up.
The key issue is what it means if they obfuscate the additions and modifications that they applied to the original GPL source code to create their own product. They could do that while leaving the original GPL code unchanged. (The argument about preferred form for editting seems to show that this is not allowed.)
Re:Subject to the ``Skating Force'' of LP days
on
Perpetual Skislope
·
· Score: 1
If you've never operated an LP phonograph -- the skating force is due to the differential friction on opposite sides of the needle on a phonograph, and tends to draw the needle inward toward the center of the record.
No, skating does not necessarily pull towards the center.
On the turntable, the record pulls the needle along the tangent, but the the arm is attached to the base at an angle to that pull. The skating force comes because of that angular difference.
Imagine that you are holding onto the doorknob of an open door. If you try to walk directly away from the hinges, the knob will stop you from moving. If instead, you walk perpendicular to the door, you will be able to walk forward for a while, being pulled sideways until you bump into the wall or get stopped because the door has turned enough that you are now pulling directly away from the hinges.
There were a number of tricks used to deal with skating. One company used a longer arm with no bend in it located so that when it was in the middle of playing the record (i.e. halfway between the edge and the hole) the arm was pointing straight along the tangent. That meant that the skating was pulling the needle forward (toward the hole) for the first half of the record and backward for the last half. The standard shorter arms would always pull towards the hole - but the amount of skating at the edge of the record was much greater (because it was bigger difference in angle). Another record arm moved completely away from the fixed rotational hinge and had the base of the arm mounted on a track. The base of the arm was moved as the record was played so that the arm was always parallel to the groove and so there was no skating.
A skier on one of these turntable slopes is not going to be attached to a fixed point, so the forces operate a bit differently. There is the rotating hill which pushes the skier around the center, and gravity which pushes the skier toward the bottom of the hill. For the top portion of the hill the difference between these forces pushes the skier horizontally in the direction of the center of the rotation. By halfway down, there is no sideways force. For the bottom portion of the hill, the difference in forces is pushing the skier away from the center.
The whole DeCSS issue involves hardware that didn't do what the owner wanted, either limiting the capabilities to what the manufacturer wanted to provide or was willing to provide, and so some of the owners provided alternate support software for the device.
The patent/IP aspect made distributing such mods illegal. An unencumbering "standard" would not have prevented that distribution.
Before people get too upset about some of the "obviously wrong" bits in this story, the radio editorial by Gordon Sinclair was given early in the 1970's.
At the time, America-bashing was fashionable - in particular, the Viet Nam war was unpopular in many places.
Sinclair's editorial struck a strong resonance against the bashing fad; leading many Canadians to think again about the many positive aspects of our neighbours and to stop thinking exclusively about their few failings.
So, the references to draft dodgers was current when the editorial was written; the Boeing Jumbo Jet was at the top of the aircraft industry, and so on.
Steganography is a general concept with myriad ways of being implemented.
You can twiddle the low bits of a JPEG, you can add a bit of low level "static" to sound, you can choose how many words are in each sentence of your message. The form of steganography I played with as a kid (well before personal computers) was a piece of paper with rectangular holes. You would write your message through the holes onto another piece of paper, remove the template, and then fill in additional words around the hidden message to cover the whole page with an innocent sounding mesage. Only someone with the same template could read the embedded "real" message.
So, you might find a filtering technique that would destroy hidden steganographic messages of one particular encoding, but you would have no assurance of destroying all possible steganographic forms.
Such filters would also announce themselves. Treated messages, whether or not they had contained any hidden stegangraphic message, would be changed - a simple checksum would determine whether that had happened. So, it would be easy to determine which routes were being protected by such filtering, and which were not.
This may be an historic first. When has there ever before been an article written about Transmeta that failed to mention Linus? One writer, at least, thinks that the whole world has gotten it finally.
"According to commentators, whoever finds the Higgs first will probably win a Nobel Prize."
Actually, they got the Nobel prize, but instead of getting it for finding the Higgs it was for inventing the Tachyon Tardyon Collider. See Robert Sawyer's Flash Forward.
Are there other employment and economic possibilities beyond gaming in the concept of virtual property?
He said inexperienced players needed to be especially careful these days. "I lost 700,000 Ultimate Online gold pieces in an attempt to purchase a Tower, a structure it's no longer possible to build on UO because there's no more open land.
But suppose the programmers created a way to insert more open land into Ultima? (I haven't played Ultima, but the same sort of change might happen in any virtual economy.) Suddenly, all those virtual properties that players have been accumulating lose their status of no longer possible to build. The "Ultima artifact" monetary system suffers a sudden devaluation. This sort of changing of the rules which defined the value on a monetary system is much like printing money without limit - it can lead to runaway inflation and people leaving that economy.
There could end up being a conflict between the "virtual economy" and the "usage" view of a virtual item. (Adding new land to Ultima might make perfect sense as far as making the game more playable for the current number of people involved, yet it would devalue the surrounding virtual economy.)
There could also be the sort of "insider trading" issues that affect all commodity markets. "Sell your properties now, they're going to open a new continent next week."
... any company wanting to export strong crypto must do so in source form
That's actually not much of a restriction - those 3 line RSA programs that people have tattooed on their bodies are source code. To export strong encryption into a binary, you "just" have to make the binary able to use an interpreter for some specialized purposes. Lots of programs already have the concept of plug-ins. An encryption plug-in in an interpreted language is exportable; and only needs to be replaced with a binary plug-in subject to export controls if there are speed problems.
It is yet to be seen whether it gets appealed again (don't they have to show that there was some flaw in the process so far to make a higher appeal?).
However, I wonder how broadly this can be applied. The rulings of the 9'th Circuit court are not binding for other circuits (AKAIK, IANAL) but rather can be used in other circuits as being suggestive only.
Perhaps the FUD'iest thing they could do would be to NOT appeal, but to charge someone else in a different jurisdiction.
Katz writes: Why are so many of these killers male and middle-class, rather than the poor or the underclass?
If a statistically significant number are middle-class, there actually could be a correlation to game-playing and on-line browsing - which would also seem to be related to middle-class (and up).
While it's a huge jump to attribute any sort of cause and effect relationship, there is something here worthy of further study.
Slashdot Mirror
In fact, mature projects should have fewer developers since there's much less left to do.
Hah, hah.
Pick a project. Compare the feature set for version 1.0 (the first officially mature release) with the feature set for version 2.0.
In many projects, release 2 is comparitivly huge. (Read "The Mythical Man-Month" - the 3 version generally has to clean up the mess.)
Additionally, release 2 is about the time that a project adds extension capabilities that allow independent development: permitting many more developers to work at once without tripping over conflicts in each other's changes.
If you look at Perl as a "typical" example, there are hundreds of people who have done some development work on the language directly; but thousands who have developed library modules.
If a cracker gets control of a system that is capable of decrypting, she can trace the code and find the key.
If the decryption is done on a separate system, she can pretend to be a system that is allowed to ask for decryption services.
If the key is manually entered by an operator and only kept in the memory of a running decryption program, she can trace through the memory.
You can't make it perfect; you'll have to be satisfied with making it hard and separately working hard to prevent the system from being cracked in the first place.
The is the Really Dangerous Linux Virus. Your system has been attacked.
This virus works on the honour system.
Please send a copy of it to all of your friends and then delete all of the files on your computer.
> > Copyright law doesn't prohibit decompilation
> Yes it does; the decompiled source is a derivative work.
So? A student's notes made while studying a work are also a derivative work.
Copyright law prohibits publishing/selling a derivative work. Fair use permits creating a derivative work for your own use.
Decompiling is not prohibited by copyright law.
(how many times have you gotten a perl error saying "You left a quote off up there at line 123"?)
That's a bad example (or a good example depending upon your point of view). Perl tells you "... probably a runaway quote starting on line 123." or words to that effect. Perl can tell you where the likely unterminated string starts but it is really hard to figure out where it should have ended.
The danger is that sometimes the auto-correction will be wrong. Not wrong enough that you notice it, just wrong enough to cause subtle bugs in the code. Refusing to compile such errors makes the programmer figure out exactly what was wrong, no matter how obvious it seems to be.
PL/1 used to fix all sorts of errors for you. It was great when it worked but it really cost you badly when it seemed to work.
... is that some of your employees may have a taste for steak.
You miss an important point. He provides a tool to facilitate anonymous posting.
- Spammers can misuse this to send spam.
- Whistleblowers and humans rights activists and oppressed people can use this to pass out information without getting shot or identified. There are parts of the world where complaining about your situation is a capital offence. Even is the "free" world, complaining about mistakes in a company can cost a job and your livelihood. (It is not "his friends" that he provides this facility for - so a large number of other people have also missed the point.)
Napster/Gnutella provide a tool to facilitate copying files.
- Some people can misuse them to bypass copyright restrictions.
- Other people can use them to copy files for legitimate purposes.
How is a general purpose tool that can allow publishing information about attrocities less worthy of your acceptance than a general purpose tool for file copying?
I'm seeing lots of people attacking the idea of them obfuscating GPL'd source code that they have gotten form somewhere else and used. That is irrelevant - the original GPL source is still available from whereever the company picked it up.
The key issue is what it means if they obfuscate the additions and modifications that they applied to the original GPL source code to create their own product. They could do that while leaving the original GPL code unchanged. (The argument about preferred form for editting seems to show that this is not allowed.)
If you've never operated an LP phonograph -- the skating force is due to the differential friction on opposite sides of the needle on a phonograph, and tends to draw the needle inward toward the center of the record.
No, skating does not necessarily pull towards the center.
On the turntable, the record pulls the needle along the tangent, but the the arm is attached to the base at an angle to that pull. The skating force comes because of that angular difference.
Imagine that you are holding onto the doorknob of an open door. If you try to walk directly away from the hinges, the knob will stop you from moving. If instead, you walk perpendicular to the door, you will be able to walk forward for a while, being pulled sideways until you bump into the wall or get stopped because the door has turned enough that you are now pulling directly away from the hinges.
There were a number of tricks used to deal with skating. One company used a longer arm with no bend in it located so that when it was in the middle of playing the record (i.e. halfway between the edge and the hole) the arm was pointing straight along the tangent. That meant that the skating was pulling the needle forward (toward the hole) for the first half of the record and backward for the last half. The standard shorter arms would always pull towards the hole - but the amount of skating at the edge of the record was much greater (because it was bigger difference in angle). Another record arm moved completely away from the fixed rotational hinge and had the base of the arm mounted on a track. The base of the arm was moved as the record was played so that the arm was always parallel to the groove and so there was no skating.
A skier on one of these turntable slopes is not going to be attached to a fixed point, so the forces operate a bit differently. There is the rotating hill which pushes the skier around the center, and gravity which pushes the skier toward the bottom of the hill. For the top portion of the hill the difference between these forces pushes the skier horizontally in the direction of the center of the rotation. By halfway down, there is no sideways force. For the bottom portion of the hill, the difference in forces is pushing the skier away from the center.
As any long-time Dungeons and Dragons player knows, AC is Armour Class.
That's why as you progress through increasing AC levels (e.g. -AC4, -AC5, etc.) the kernel gets stronger and less easily attacked.
The whole DeCSS issue involves hardware that didn't do what the owner wanted, either limiting the capabilities to what the manufacturer wanted to provide or was willing to provide, and so some of the owners provided alternate support software for the device.
The patent/IP aspect made distributing such mods illegal. An unencumbering "standard" would not have prevented that distribution.
At the time, America-bashing was fashionable - in particular, the Viet Nam war was unpopular in many places.
Sinclair's editorial struck a strong resonance against the bashing fad; leading many Canadians to think again about the many positive aspects of our neighbours and to stop thinking exclusively about their few failings.
So, the references to draft dodgers was current when the editorial was written; the Boeing Jumbo Jet was at the top of the aircraft industry, and so on.
You can twiddle the low bits of a JPEG, you can add a bit of low level "static" to sound, you can choose how many words are in each sentence of your message. The form of steganography I played with as a kid (well before personal computers) was a piece of paper with rectangular holes. You would write your message through the holes onto another piece of paper, remove the template, and then fill in additional words around the hidden message to cover the whole page with an innocent sounding mesage. Only someone with the same template could read the embedded "real" message.
So, you might find a filtering technique that would destroy hidden steganographic messages of one particular encoding, but you would have no assurance of destroying all possible steganographic forms.
Such filters would also announce themselves. Treated messages, whether or not they had contained any hidden stegangraphic message, would be changed - a simple checksum would determine whether that had happened. So, it would be easy to determine which routes were being protected by such filtering, and which were not.
It's not dead yet. In fact, it's only just barely alive - so it's still barely capable of trying to gum you to death or to bleed all over you.
This may be an historic first. When has there ever before been an article written about Transmeta that failed to mention Linus? One writer, at least, thinks that the whole world has gotten it finally.
"According to commentators, whoever finds the Higgs first will probably win a Nobel Prize."
Actually, they got the Nobel prize, but instead of getting it for finding the Higgs it was for inventing the Tachyon Tardyon Collider. See Robert Sawyer's Flash Forward.
Are there other employment and economic possibilities beyond gaming in the concept of virtual property?
He said inexperienced players needed to be especially careful these days. "I lost 700,000 Ultimate Online gold pieces in an attempt to purchase a Tower, a structure it's no longer possible to build on UO because there's no more open land.
But suppose the programmers created a way to insert more open land into Ultima? (I haven't played Ultima, but the same sort of change might happen in any virtual economy.) Suddenly, all those virtual properties that players have been accumulating lose their status of no longer possible to build. The "Ultima artifact" monetary system suffers a sudden devaluation. This sort of changing of the rules which defined the value on a monetary system is much like printing money without limit - it can lead to runaway inflation and people leaving that economy.
There could end up being a conflict between the "virtual economy" and the "usage" view of a virtual item. (Adding new land to Ultima might make perfect sense as far as making the game more playable for the current number of people involved, yet it would devalue the surrounding virtual economy.)
There could also be the sort of "insider trading" issues that affect all commodity markets. "Sell your properties now, they're going to open a new continent next week."
... any company wanting to export strong crypto must do so in source form
That's actually not much of a restriction - those 3 line RSA programs that people have tattooed on their bodies are source code. To export strong encryption into a binary, you "just" have to make the binary able to use an interpreter for some specialized purposes. Lots of programs already have the concept of plug-ins. An encryption plug-in in an interpreted language is exportable; and only needs to be replaced with a binary plug-in subject to export controls if there are speed problems.
It is yet to be seen whether it gets appealed again (don't they have to show that there was some flaw in the process so far to make a higher appeal?).
However, I wonder how broadly this can be applied. The rulings of the 9'th Circuit court are not binding for other circuits (AKAIK, IANAL) but rather can be used in other circuits as being suggestive only.
Perhaps the FUD'iest thing they could do would be to NOT appeal, but to charge someone else in a different jurisdiction.
Katz writes: Why are so many of these killers male and middle-class, rather than the poor or the underclass?
If a statistically significant number are middle-class, there actually could be a correlation to game-playing and on-line browsing - which would also seem to be related to middle-class (and up).
While it's a huge jump to attribute any sort of cause and effect relationship, there is something here worthy of further study.