Re-implement? It isn't even possible to implement it ONCE!! The format that MS is calling OOXML doesn't meet the standard that they bought and paid for. (Except in a truly trivial sense that allows a zipped Czechoslovakian translation of a Japanese document to be legitimately called an OOXML version. You wouldn't BELIEVE the exceptions that thing allows! Just make it a binary blob with a one-time pad and it fits as legitimate OOXML...not that anything could read it. [Note: The particular thing that I'm thinking of could only save that one document, nothing else, and there isn't a program that could read it.])
Sorry, correction: In Soviet America TV watches you.
This correction inspired by an invention that the cable companies recently voted invention of the year. It allows them to tell which channel is currently being watched by which TVs.
Lots of people are seduced by eye-candy. But those same people are often very price sensitive. And sometimes they NEED a computer that will run off batteries for a long time.
The thing I'm dubious about is that Wi-Fi access is common enough that there's a large market for something that requires web access to function reasonably. If it doesn't have, say, OpenOffice (or equivalent) installed, then I'm not sure who they're selling it to. I suppose there's no reason they couldn't have OpenOffice installed, but they never seem to advertise it, and I'm not sure how heavy an application that is in CPU usage. AbiWord, perhaps? And some light spreadsheet...perhaps GnuMeric?
So it all seems doable, but it's not clear that this is the direction they're heading. (Still, if Ubuntu's installed, the only question is "Will this [light] app work reasonably with the given screen size?".)
They're aiming at the NetBook market-space, so much of that wouldn't be a problem. Flash might be a problem, but they may have an answer (either a deal with Adobe or an enhanced version of gnash or some such). I think you can forget WINE & Mono. This is a netbook, after all. Not a full fledged computer. By the time the chips get powerful enough to make the fancy stuff practical it could be ported. (Though I *do* hope that Mono gets left behind, and I suspect that WINE is impossible without a vm.)
I suspect that most of the FireFox extensions wouldn't require much more than recompilation.
So I don't see the same problems that you do. But it still might just be a bargaining chip.
That's not what geneticists say. They say that the rate of human evolution has increased markedly. Of course, they have a hard time measuring periods shorter than a few thousand years, but that's long enough for the effects of city life to have shown up. And a few other things. Lots of evolution in the immune system and in the expression of neural proteins in the brain.
This doesn't mean that they know what it means. There's certainly a huge amount of hybridization going on. But this means that evolution isn't slowing down. And it seems as if most of it's happening in locations rather then between locations. (I.e., professions or social classes becoming specialized rather than countries or continents.)
Then again, remember the coarse filter being used. Anything within the last 200 years will be totally invisible.
Yes, I do believe that historically the christian churches have caused more harm to society than any other single force. And I disapprove of ALL dictatorships.
One could mention several attempts to equal the harm done by the christian churches...Genghis Khan, Tamerlane, Hitler, Pot Pol, etc., but they have all been but short term affairs. To equal the evil of the churches requires staying power.
I have a friend who argues that Soviet Communism was basically a Christian heresy. I haven't found a good argument that he's wrong. In both cases there's the worship of an infallible authority. (The Protestant Churches claim that he's not incarnate, the Catholics say that the Pope is infallible. So some Protestant churches are tolerable. If you claim that there exists an infallible authority, but that nobody knows what it really wants, then you aren't doing much harm.)
OK. Everywhere I've worked those offices have been held by the same person as a single position. (The trouble was finding unused hardware on which to perform the backup validations. Sometimes it didn't happen for awhile, and once that lead to problems.)
If you can identify them uniquely within your database, why do you need another key? You can store all the data in your database.
If you *DO* need such a thing, how about email address + date of birth. SSN could be a fallback in case they didn't HAVE an email address.
But *DON'T* have that be your unique ID primary key. Allow it to be blank. Even if someone has a SSN and wants you to store it, they might not remember it.
There's one that was actually canceled, though I don't remember where I found that information.
You might try 123-45-6789. I have a suspicion, though, that it wouldn't be accepted. It would probably fail a validity check. Still, by that time you might have already been served. .
At one time one was not supposed to use the SSN for anything not involving the Social Security Administration. That was a long time ago. I was told that it was originally illegal, but I don't know that this was really so.
N.B.: This was specifically the SSN. Don't generalize it to other kinds of data, which have largely never been regulated.
This is the systems administrator we're talking about here. He manages the systems backup, and where the off-site back up is stored and how it's accessed. He has total access to EVERYTHING that you didn't encrypt. And he can get to it from off-site. He can copy everything during a "routine verification of backup readability". And he'd BETTER do those routine verifications!
But I *don't* accept their arguments. And I question either the sense or the honesty of anyone who does. Including them.
What they are talking about is copyright infringement. Sometimes they have a point. Other times they're accusing innocent people with shoddy or even constructed evidence.
Then there's the assertion that it "hurts no one". I'm not certain. E.g., I have the weird belief that much of the music they provide is damaging to those who hear it. I feel the same way about TV and radio. It's a medium that is designed to prevent you from having a chance to sit back and evaluate what you're experiencing. This dynamic flow is, admittedly, attractive. That doesn't inherently mean that it's not injurious, and I suspect that it is, that the damage in each separate exposure is quite mild, and probably indetectable, but that it's cumulative over time. I know of no studies that have addressed this point, so I won't assert it as a fact. But I believe the downloads to frequently be injurious to the recipients.
There was this case where a California hospital contracted for the entry of medical data into a database with a Texas company, who subcontracted it to a Florida company who subcontracted it to an Indian company who subcontracted it to a lady in Afghanistan. When she didn't get paid, she auctioned off the data to pay for her time, effort and costs.
I never did hear how that turned out, but it's hard to call the Afghan lady dishonest. The Indian company, yest. Probably the Florida company. And the Texas company broke it's agreement on confidentiality...
You've *GOT* to trust your systems admin. You really don't have any choice. If you choose to BE the systems admin, then you've got to trust yourself.
THEREFORE: When you chose to outsource your systems administration, you chose to trust your system to people you don't have any direct oversight over. (Do you really want strangers trooping through your company every time you have any sort of system problem? That opens other kinds of security holes.)
Should you trust a distant remote administrator for your system? I'd say no. But then I'd say your choice to outsource was a gross mistake, also. Once you've made that choice, you are stuck with the results. And one result is that you don't know who is administering your system, or what they're doing. And you can only find out what they choose to tell you. At minimum you can expect them to be playing CYA, so you definitely can't expect honest answers. Employees play CYA too, but if you can watch them, you can get an idea what's going on. If they say "It's really nothing" in between yelling into the phone at a parts supplier, you have an idea what kind of nothing they mean. If they're remote, you don't have that information.
If you want the remote systems manager to send in an on-site engineer to administer your system, be prepared to pay IBM systems prices. (And IBM is very skilled at providing that kind of service. And they have a track record that can't be beaten. [It would take decades to establish a record to equal theirs, and while those decades are passing, IBM would be improving their current record.] But they aren't cheap.)
If your code is publicly published, then I agree, the BSD license does what you say. But just as with the GPL, BSD code doesn't always end up publicly published.
I can't speak as to how "most people who use the GPL" expect with any authority. I don't know most of them. In my experience most of the developers I know/have known that develop code under the GPL expect continuing access to updates. I'll agree that the GPL doesn't guarantee this, but it makes it a more reasonable expectation.
Projects generally have an optimal number of contributors. What the optimum is depends partially on the managerial skill of the directors and partially on the technical skills of the contributors. And a whole lot on the degree of agreement over goals of the project community. At the moment my project appears to me to have an optimal size of 1. I don't have the code is good enough shape that I can demonstrate what I'm trying to do, and I keep changing my mind about details of implementation to achieve the goal. Fundamental details.
I *would* like to be able to talk about it with others, but I don't know any place where I can. My friends all seem to end up being system administrators or mathematicians, and I can't really talk about it to them.
You are clearly an ardent supporter, and I'm largely ignorant. All I know is what I read on the llvm web page...and past experience with virtual machines. (If it generates native code, that takes it out of the virtual machine category, though.)
You may be right. Certainly D has an llvm compiler coming along, and the gcc one seems to have gone into stasis. (Since, however, the gcc based D compiler had only one maintainer, I don't consider this significant evidence.)
When the D based llvm compiler starts supporting D2.x (the development version) then I'll start seriously looking at llvm.
OTOH, why is llvm better than Parrot? (Can Parrot generate x86 code? Don't know why it couldn't, but perhaps it doesn't.)
There's more things to be looked at than there's time for looking in any detail. But nobody's going to be "stuck" with gcc. At each major release cycle they can choose whether to use gcc or something else as their major compiler. So far nothing's been a credible contender against gcc...that didn't BECOME gcc. (Here I'm thinking of EGCS.)
YES!! MS Word 5.1 a for the Mac LC (2?) was the best word processor I have ever used, even until now. The computer wasn't quite up to some of the documents I threw at it, but the word processor has yet to be equaled.
There probably isn't much that can be done. English is, unless you are very careful, ambiguous on this point. And you aren't going to get most people to be very careful.
Some people prefer to speak of Libre Software (Software Libre?), but that never caught on. Others talk of free-beer vs. free-freedom software. (That became a sort of joke with the GPL beer license.) But that's also too clumsy for normal conversation.
So I guess we just live with the confusion, and deal with it as best we can. An when those appear who intentionally conflate the meanings, I guess we'll argue with them for awhile, and then flame them and ignore them as trolls.
The unfortunate truth is that the confusion isn't all deliberate. And there is a real overlap in the meaning. A very large overlap. E.g., "Software is free if it costs less than $0.01/MB to distribute it." True or false? Well, most GPL software costs less than $0.01/MB to distribute, and we say that it's free. We *mean* free as in freedom, but a result is free as in beer. The GPL doesn't forbid charging to distribute the software, but it establishes a floor price of $0.00, because everyone who receives a copy has the right to distribute it for any price they choose, and it's really cheap to distribute software.
So it's not really surprised that many people are honestly confused.
P.S.: Re:This isn't sensationalist, it's the truth
on
Leaving the GPL Behind
·
· Score: 1
I absolutely HATE the current preview mode. By eliminating paragraph spacing it renders the post unreadable. So I usually don't preview successfully. Like this time.
Change: publish the code BSD, an my code would be a trivial part of
To: publish the code BSD, and my code would be a trivial part of
OK. Use BSD then. Most GPL or LGPL libraries have rough equivalents that are BSD. Some of them need a bit of work, but that's expectable, because fewer people choose to work on BSD libraries.
You use what works well for you, and I'll use what works well for me. For me it's GPL. I'll admit I'm still dithering between GPLv3 and AGPL, but if you're going to use BSD my dithering won't affect you.
If you want a different license, contact me. I'll make you an offer I consider reasonable. If you're proposing to publish the code BSD, an my code would be a trivial part of your project, I'll probably just want an exchange (so that I'll be certain I get a copy of your code under a GPL compatible license).
You'll note I didn't mention any names? It seems plausible to me that most GPL projects would reason this way. I certainly would.
Why do you appear to suppose that Linux doesn't support the LLVM? This is incorrect.
As to inferior... LLVM is, as it's name implies, a virtual machine. As such it's probably slower than good assembly code of a particular langauge written for a compiler that was customized to produce code targeted to the hardware on which it runs.
I *do* understand that LLVM is considerably faster than the JVM. That doesn't mean it's as fast as good native code. And it definitely doesn't mean that it has less overhead.
(It also appears to use GCC as a front-end for it's C compiler. Just FYI.)
Slashdot Mirror
Re-implement? It isn't even possible to implement it ONCE!! The format that MS is calling OOXML doesn't meet the standard that they bought and paid for. (Except in a truly trivial sense that allows a zipped Czechoslovakian translation of a Japanese document to be legitimately called an OOXML version. You wouldn't BELIEVE the exceptions that thing allows! Just make it a binary blob with a one-time pad and it fits as legitimate OOXML...not that anything could read it. [Note: The particular thing that I'm thinking of could only save that one document, nothing else, and there isn't a program that could read it.])
Sorry, correction:
In Soviet America TV watches you.
This correction inspired by an invention that the cable companies recently voted invention of the year. It allows them to tell which channel is currently being watched by which TVs.
Lots of people are seduced by eye-candy. But those same people are often very price sensitive. And sometimes they NEED a computer that will run off batteries for a long time.
The thing I'm dubious about is that Wi-Fi access is common enough that there's a large market for something that requires web access to function reasonably. If it doesn't have, say, OpenOffice (or equivalent) installed, then I'm not sure who they're selling it to. I suppose there's no reason they couldn't have OpenOffice installed, but they never seem to advertise it, and I'm not sure how heavy an application that is in CPU usage.
AbiWord, perhaps? And some light spreadsheet...perhaps GnuMeric?
So it all seems doable, but it's not clear that this is the direction they're heading. (Still, if Ubuntu's installed, the only question is "Will this [light] app work reasonably with the given screen size?".)
They're aiming at the NetBook market-space, so much of that wouldn't be a problem. Flash might be a problem, but they may have an answer (either a deal with Adobe or an enhanced version of gnash or some such). I think you can forget WINE & Mono. This is a netbook, after all. Not a full fledged computer. By the time the chips get powerful enough to make the fancy stuff practical it could be ported. (Though I *do* hope that Mono gets left behind, and I suspect that WINE is impossible without a vm.)
I suspect that most of the FireFox extensions wouldn't require much more than recompilation.
So I don't see the same problems that you do. But it still might just be a bargaining chip.
That's not what geneticists say. They say that the rate of human evolution has increased markedly. Of course, they have a hard time measuring periods shorter than a few thousand years, but that's long enough for the effects of city life to have shown up. And a few other things. Lots of evolution in the immune system and in the expression of neural proteins in the brain.
This doesn't mean that they know what it means. There's certainly a huge amount of hybridization going on. But this means that evolution isn't slowing down. And it seems as if most of it's happening in locations rather then between locations. (I.e., professions or social classes becoming specialized rather than countries or continents.)
Then again, remember the coarse filter being used. Anything within the last 200 years will be totally invisible.
Yes, I do believe that historically the christian churches have caused more harm to society than any other single force. And I disapprove of ALL dictatorships.
One could mention several attempts to equal the harm done by the christian churches...Genghis Khan, Tamerlane, Hitler, Pot Pol, etc., but they have all been but short term affairs. To equal the evil of the churches requires staying power.
I have a friend who argues that Soviet Communism was basically a Christian heresy. I haven't found a good argument that he's wrong. In both cases there's the worship of an infallible authority. (The Protestant Churches claim that he's not incarnate, the Catholics say that the Pope is infallible. So some Protestant churches are tolerable. If you claim that there exists an infallible authority, but that nobody knows what it really wants, then you aren't doing much harm.)
OK. Everywhere I've worked those offices have been held by the same person as a single position. (The trouble was finding unused hardware on which to perform the backup validations. Sometimes it didn't happen for awhile, and once that lead to problems.)
If you can identify them uniquely within your database, why do you need another key? You can store all the data in your database.
If you *DO* need such a thing, how about email address + date of birth. SSN could be a fallback in case they didn't HAVE an email address.
But *DON'T* have that be your unique ID primary key. Allow it to be blank. Even if someone has a SSN and wants you to store it, they might not remember it.
There's one that was actually canceled, though I don't remember where I found that information.
You might try 123-45-6789. I have a suspicion, though, that it wouldn't be accepted. It would probably fail a validity check. Still, by that time you might have already been served.
.
At one time one was not supposed to use the SSN for anything not involving the Social Security Administration. That was a long time ago. I was told that it was originally illegal, but I don't know that this was really so.
N.B.: This was specifically the SSN. Don't generalize it to other kinds of data, which have largely never been regulated.
I see. And because he didn't choose to accept MS' EULA, he's bound by it??
Somehow that seems a bit of strange reasoning.
Digital Research & CP/M.
But I think you bought your copies.
Are you figuring odd per minute or odds per century?
He just said odds. He didn't quantify it. I'd guess he meant odds/year, not odds/day.
You're kidding, right?
This is the systems administrator we're talking about here. He manages the systems backup, and where the off-site back up is stored and how it's accessed. He has total access to EVERYTHING that you didn't encrypt. And he can get to it from off-site. He can copy everything during a "routine verification of backup readability". And he'd BETTER do those routine verifications!
But I *don't* accept their arguments. And I question either the sense or the honesty of anyone who does. Including them.
What they are talking about is copyright infringement. Sometimes they have a point. Other times they're accusing innocent people with shoddy or even constructed evidence.
Then there's the assertion that it "hurts no one". I'm not certain. E.g., I have the weird belief that much of the music they provide is damaging to those who hear it. I feel the same way about TV and radio. It's a medium that is designed to prevent you from having a chance to sit back and evaluate what you're experiencing. This dynamic flow is, admittedly, attractive. That doesn't inherently mean that it's not injurious, and I suspect that it is, that the damage in each separate exposure is quite mild, and probably indetectable, but that it's cumulative over time. I know of no studies that have addressed this point, so I won't assert it as a fact. But I believe the downloads to frequently be injurious to the recipients.
Or even people are aren't dishonest.
There was this case where a California hospital contracted for the entry of medical data into a database with a Texas company, who subcontracted it to a Florida company who subcontracted it to an Indian company who subcontracted it to a lady in Afghanistan. When she didn't get paid, she auctioned off the data to pay for her time, effort and costs.
I never did hear how that turned out, but it's hard to call the Afghan lady dishonest. The Indian company, yest. Probably the Florida company. And the Texas company broke it's agreement on confidentiality...
Ditto.
You've *GOT* to trust your systems admin. You really don't have any choice. If you choose to BE the systems admin, then you've got to trust yourself.
THEREFORE:
When you chose to outsource your systems administration, you chose to trust your system to people you don't have any direct oversight over. (Do you really want strangers trooping through your company every time you have any sort of system problem? That opens other kinds of security holes.)
Should you trust a distant remote administrator for your system? I'd say no. But then I'd say your choice to outsource was a gross mistake, also. Once you've made that choice, you are stuck with the results. And one result is that you don't know who is administering your system, or what they're doing. And you can only find out what they choose to tell you. At minimum you can expect them to be playing CYA, so you definitely can't expect honest answers. Employees play CYA too, but if you can watch them, you can get an idea what's going on. If they say "It's really nothing" in between yelling into the phone at a parts supplier, you have an idea what kind of nothing they mean. If they're remote, you don't have that information.
If you want the remote systems manager to send in an on-site engineer to administer your system, be prepared to pay IBM systems prices. (And IBM is very skilled at providing that kind of service. And they have a track record that can't be beaten. [It would take decades to establish a record to equal theirs, and while those decades are passing, IBM would be improving their current record.] But they aren't cheap.)
If your code is publicly published, then I agree, the BSD license does what you say. But just as with the GPL, BSD code doesn't always end up publicly published.
I can't speak as to how "most people who use the GPL" expect with any authority. I don't know most of them. In my experience most of the developers I know/have known that develop code under the GPL expect continuing access to updates. I'll agree that the GPL doesn't guarantee this, but it makes it a more reasonable expectation.
Projects generally have an optimal number of contributors. What the optimum is depends partially on the managerial skill of the directors and partially on the technical skills of the contributors. And a whole lot on the degree of agreement over goals of the project community. At the moment my project appears to me to have an optimal size of 1. I don't have the code is good enough shape that I can demonstrate what I'm trying to do, and I keep changing my mind about details of implementation to achieve the goal. Fundamental details.
I *would* like to be able to talk about it with others, but I don't know any place where I can. My friends all seem to end up being system administrators or mathematicians, and I can't really talk about it to them.
You are clearly an ardent supporter, and I'm largely ignorant. All I know is what I read on the llvm web page...and past experience with virtual machines. (If it generates native code, that takes it out of the virtual machine category, though.)
You may be right. Certainly D has an llvm compiler coming along, and the gcc one seems to have gone into stasis. (Since, however, the gcc based D compiler had only one maintainer, I don't consider this significant evidence.)
When the D based llvm compiler starts supporting D2.x (the development version) then I'll start seriously looking at llvm.
OTOH, why is llvm better than Parrot? (Can Parrot generate x86 code? Don't know why it couldn't, but perhaps it doesn't.)
There's more things to be looked at than there's time for looking in any detail. But nobody's going to be "stuck" with gcc. At each major release cycle they can choose whether to use gcc or something else as their major compiler. So far nothing's been a credible contender against gcc...that didn't BECOME gcc. (Here I'm thinking of EGCS.)
Judging by recent news you can strike:
A) Is it legal?
and substitute
A) If it's not legal, is it worth the risk of getting caught?
YES!! MS Word 5.1 a for the Mac LC (2?) was the best word processor I have ever used, even until now. The computer wasn't quite up to some of the documents I threw at it, but the word processor has yet to be equaled.
There probably isn't much that can be done. English is, unless you are very careful, ambiguous on this point. And you aren't going to get most people to be very careful.
Some people prefer to speak of Libre Software (Software Libre?), but that never caught on. Others talk of free-beer vs. free-freedom software. (That became a sort of joke with the GPL beer license.) But that's also too clumsy for normal conversation.
So I guess we just live with the confusion, and deal with it as best we can. An when those appear who intentionally conflate the meanings, I guess we'll argue with them for awhile, and then flame them and ignore them as trolls.
The unfortunate truth is that the confusion isn't all deliberate. And there is a real overlap in the meaning. A very large overlap. E.g., "Software is free if it costs less than $0.01/MB to distribute it." True or false? Well, most GPL software costs less than $0.01/MB to distribute, and we say that it's free. We *mean* free as in freedom, but a result is free as in beer. The GPL doesn't forbid charging to distribute the software, but it establishes a floor price of $0.00, because everyone who receives a copy has the right to distribute it for any price they choose, and it's really cheap to distribute software.
So it's not really surprised that many people are honestly confused.
I absolutely HATE the current preview mode. By eliminating paragraph spacing it renders the post unreadable. So I usually don't preview successfully. Like this time.
Change:
publish the code BSD, an my code would be a trivial part of
To:
publish the code BSD, and my code would be a trivial part of
OK. Use BSD then. Most GPL or LGPL libraries have rough equivalents that are BSD. Some of them need a bit of work, but that's expectable, because fewer people choose to work on BSD libraries.
You use what works well for you, and I'll use what works well for me. For me it's GPL. I'll admit I'm still dithering between GPLv3 and AGPL, but if you're going to use BSD my dithering won't affect you.
If you want a different license, contact me. I'll make you an offer I consider reasonable. If you're proposing to publish the code BSD, an my code would be a trivial part of your project, I'll probably just want an exchange (so that I'll be certain I get a copy of your code under a GPL compatible license).
You'll note I didn't mention any names? It seems plausible to me that most GPL projects would reason this way. I certainly would.
Why do you appear to suppose that Linux doesn't support the LLVM? This is incorrect.
As to inferior... LLVM is, as it's name implies, a virtual machine. As such it's probably slower than good assembly code of a particular langauge written for a compiler that was customized to produce code targeted to the hardware on which it runs.
I *do* understand that LLVM is considerably faster than the JVM. That doesn't mean it's as fast as good native code. And it definitely doesn't mean that it has less overhead.
(It also appears to use GCC as a front-end for it's C compiler. Just FYI.)