First off, as I've said before, this is GOING to happen, like it or not. Every trend in the industry is pointing that way. Texas licenses software engineers and other states will eventually follow; post 9/11 the govt. is very concerned about security and more inclined to legislate it; and consumer advocates have been pushing for limiting or doing away w/ warranty disclaimers for some time. Everyone here doing software development for a living (whether for a giant corp or self-employed developers doing consulting gigs) better either prepare for liability for faulty software, get out the pocketbooks and lawyers to start lobbying madly, or find a new career.
WRT Open Source software, I see no a priori reason why OSS developers should be any less liable than commercial software companies, PROVIDED that certain reasonable guidelines apply:
liability should never exceed the amount of money the developer/company *received* from the customer or class of customers unless gross negligence can be proved;
in cases of gross negligence, the liability should coincide with the amount of *actual damages*; i.e., you don't get a million bucks because someone was able to read your web documents unless that act actually cost you $1 million in losses;
developers should be reasonably shielded from liability in cases where the customer/user *actually* modified the software (not just *had* source available) -- if the modifications had a substantive affect on the security or safety of the product;
parties can enter binding legal contracts to alter the balance of liability -- in instances where the customer *plans* to alter the software, whether they end up doing it or not. CLICKWRAP LICENSES DON'T COUNT.
These measures will only benefit the the software industry; serious programmers will have the satisfaction of working in a climate where time to market takes a back seat to quality (because the law penalizes nonconformance to this norm); software processes in the aggregate will improve for the same reason; customers and users will have a better experience with software in general and will have more respect for practioners who take the profession seriously.
And people who lack confidence in their abilities to generate bug-free code can buy liability insurance, just like many other professionals currently do. In other words, software professionals can finally expect to *earn* the title!
I take it from that fragment that the claim they're making is that all of the boundaries between layers are "opaque;" piece A can't see anything about the implementation of piece B, but communicates w/ it via a blackbox interface.
It is arguably a stretch to call that "object-oriented" since OO has come to mean so many things besides data encapsulation (though in my view that's the most important OO feature, the part which allows you to replace the implementation at any time). But hey, it's certainly less of a stretch than Oracle calling 9i "unbreakable."
Darwin is a single-server mach-based unix layer, Hurd is a multi-server unix layer. I.e., Darwin, like all other unices, provides a system-wide view of all resources (filesystems, priviledges, etc). A multi-server environment lets you compartmentalize things to a much higher degree. Each user can replace whole chunks of functionality in a multi-server OS (i.e., their own private filesystem, their own swapping algorithms perhaps). I'm not familiar w/ the details of the Hurd's approach, and what can actually be customized.
RMS specifically addresses this issue in his "philosophy" paper, IIRC. Basically, people will pay competent programmers to produce software.
The fact is, most businesses gain value from software from *using* it, not selling it. So, they will always need people to build software for them, and they'll pay for it. But the only people who win when the software is proprietary are the *few* companies who sell it (and probably relatively few of the employees of those few compaines as well).
AFAIK is not "mathematically impossible" to break even the strongest crypto available. It is "computationally infeasable." I.e., it's mathematically possible (by factoring all the large primes that could have been used for the key, for instance), but you can't afford the time/money (mostly time) required.
Re:Plain, not strict, interpretation
on
Abusing the GPL?
·
· Score: 2
No, I believe that compilers have a specific exception in the GPL along w/ OS's and standard libraries.
The bottom line AFAICT is you need to distribute all the files needed to build it with a certain baseline set of tools assumed.
one must go on a fact-finding parade to measure industry practice
Since YAAL, please explain why the answer to the question "Is the file you distributed what your engineers modify when they fix bugs or introduce features to the program" is not conclusive for the purposes of determining "preferred form" here?
Plain, not strict, interpretation
on
Abusing the GPL?
·
· Score: 2
Under the strict interpretation of the GPL people are proposing here, it would not be sufficient to simply release the source code to this application - all the associated project files that the IDE manages would also have to be released
From the GPL, section 3:
"The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable."
It specifically mentions the compile and install scripts. So it's the plain interpretation, not a strict interpretation, that leads to the conclusion that one must include the project files, etc. You need to distribute whatever is needed to transform the source -- in the form that *you* the developer normally use to modify it -- into the object files you distribute.
Note, I don't think that there's anything wrong with this or controversial at all. How useful would GCC be to you without all of the Makefiles? Not very.
A while ago I mentioned my interest in setting up a PAC to fight things like this. I got a few responses and some supportive comments from fellow/.ers.
This is the final straw which has caused me to actually get off my ass and gather all of the legal info I could find about setting up a PAC (i.e, should be a 503 (1)(c)(4) nonprofit, and all of the accounting and tax rules). This is all from 26 USC 501 & 527.
I hope to read through all of this information, hopefully this weekend. If I think I understand it sufficiently I'll go through the process of actually registering a corporation (by default here in PA, unless anyone has any good reason it should be somewhere else and practical advice on doing it that way).
Assuming I get that far, I'll submit a story to Slashdot w/ all of the particulars inlcuding what I've done so far and what I think I'll need in terms of support resources (people, services, etc).
Anyone w/ any more ideas, info & suggestions is of course free to post here or email them to me.
As you might recall, the speculation prior to the appeals court ruling was that MS had little chance because appeals courts were generally reluctant to overturn the finding of facts of a lower court.
That same appeals court had previously done so for MS concerning the 1995 consent decree.
Legislative branch: passes the laws (antitrust laws obviously included), which generally include specific penalties or ranges of penalties.
Executive branch: enforces the laws passed by Congress. This includes bringing actions against violators and suggesting from among the statutory remedies.
Judicial branch: interprets laws passed by the legislative branch, and meets out punishment within the boundaries established in those laws (which are often left to the discretion of the judge).
I.e., MS is full of shit. The judge is free to completely disregard the opinions of the executive branch if the executive branch abrogates its responsibility to pursue thorough enforcement of the laws passed by the legislature. For example, plea agreements in criminal cases must be approved by the judge, and are in part accepted or denied on the basis of whether or not the executive branch is likely to prevail at trial on the original charge; the government can't (or rather, isn't supposed to be able to)-- if it has a good case -- just decide to let the defendant plead to a lesser charge (subject to statutory exceptions -- i.e., in exchange for testimony against another individual).
As for anyone who has specific case law which they think invalidates what I've said above in this specific context, well, IANALAIYAE,TSUB (I am not a lawyer and if you aren't either, then shut up, beeyatch).
While Jackson's comments were inflammatory to be sure, IIRC they did not go to the finding of facts or the findings of law at issue in the case, but rather to the demeanor and conduct of MS' officers and counsel during the trial.
While surely they could give the impression of bias, it would be difficult to imagine anyone other than the defendants in the case who wasn't inclined to share Judge Jackson's opinions about MS' behaviour in the courtroom. Meanwhile, the Judge's finding of facts and conclusions of law were largely upheld by the (previously pro-MS) appeals court, which should speak volumes about the soundness of the judgement in light of the facts.
There will still be a pool of employable people; the malingerers who were smart enough to survive by digging through dumpters, innovating WRT to shelter, and possibly resorting to theft and murder. In short, scavengers, visionaries, and criminals.
Of course, one would have to offer these people quite a bit to get them to leave their niche and join the rolls of the gainfully employed, but that's no problem as they'd obviously primarily be management candidates anyway.
That said, I don't see how we can't trust any normal citizen (not minors, criminals, illegals, etc.) not to make the decision to purchase firearms. So, ultimately, I don't really understand why that first statement of yours is notable.
That's the parent poster's point. In one case (guns) -- where the possibility of harm includes depriving another citizen of one of their supposedly inalienable rights (life) -- we trust people absent any clear indication of their intent to harm others.
But for anything involving copyright and digital media -- where the worst harm imagineable doesn't impinge upon anyone's inalienable rights -- everyone is assumed to be a criminal, and is treated accordingly.
As for Enron, et. al - I have no sympathy for the sheeple that went along with the dumb idea of putting *all* their 401(k) money into their own company's stock... I put no more than 5% of my 401(k) money back into my own employer's stock. I guess those people never heard of the idea of diversification, huh?
Enron's matching for 401k contributions was in Enron stock, so even if any employee was "smart" enough to put *none* of their own money in the stock, a significant enough portion was in that form to merit being angry at its dissolution...
They were stupid and deserve all the misery they got.
So, workers without much experience in financial markets deserve to be screwed over, is that it?
The fact is, Enron lied to all of its shareholders. It was not possible to make a rational decision regarding the value of their stock because the true value of the company was hidden.
The Old Boys may be crooked, but at least they aren't complete morons.
While you are enough of an elitist to gain comfort from the intelligence of your abusers, I suspect that for most people the fact that they were ripped of by geniuses is cold comfort indeed.
ECMA rules don't say that a member must disclose pending patents, only issued ones. And in the case of either, ECMA rules only insist on RAND licensing, which means free software developers can be hit with licensing requirements at a later date.
This is how we will be screwed; Microsoft submarine patents. And Miguel can claim as much as he wants that "lawyers" have told him that the patent issue can be avoided, but that kind of assumes that one knows about the patents -- which in the case of "submarine" patents is not the case. Even if a future MS patent on this stuff turns out to be invalid due to prior art, WHO IS GOING TO COURT AGAINST M$$$$$ FOR YEARS AND MUCHO $$$$$ TO PROVE IT?
I think the crux of the issue is that there is one set of (somewhat difficult) skills, the mastery of which is a requisite to being a good engineer, another set for being a good manager. Good technical managers need to either:
have all of the skills required to be a good manager and at least some of the skills (engineering judgement) of the engineer; or
must be willing to defer to the expertise of the technical experts who work for them when it's appropriate.
As far as the first option goes, people that demonstrably have one or the other set of skills are hard enough to find; imagine how hard it must be to find people who have both! The second requires a kind of wisdom that I'd say most managers, like most people don't have. People in general are control freaks, professional managers even more so -- since a willingness to acceed control can be perceived as weakness, and often times what's required isn't just a willingness to say "yes" to those below, but the courage to say "no" to those above.
I've said it before (elsewhere), and I'll say it again here:
Patents are going to screw any free software/OSS implementation of.NET, and anyone (i.e. GNOME) depending on it is going to be f**ked.
Microsoft has flat out said that it will use patents to defend it's.NET investment. Miguel claims that lawyer(s) have told him that the patent issue can be avoided.
DON'T BELIEVE IT (I don't; here's why):
ECMA rules stipulate RAND licensing of member's patents, and don't require disclosure of pending patents. Remember Miguel's quote about those, spiffy -- novel -- assemblies? What if MS has a broad patent filed to cover those? What if MS has a bunch of other overly broad patents in the pipeline to stuff that may seem non-novel, obvious, or based on prior art to you and I, but which a judge or patent examiner will give the benefit of the doubt?
Microsoft will use patents for things like this to force all implementers of the standard beside themselves to pay, OSS/Free Software or not.
And that's the good scenario! Suppose MS does a RAMBUS and says that they aren't LEGALLY bound by ECMA RAND policies? Then they could refuse to allow an OSS/FS implementation at all, and even if it wouldn't ultimately stand up in court, what free software company, developer or user can afford to fight one of the most deep-pocketed companies in existence in court on this one?
This would also favour large corporates. Since they tend to have more options for hiding income.
Actually, I suppose I meant that the liability should be limited to the amount of damages to the customer or the amount the customer paid for the software, whichever is smaller -- not including punitive damages for negiligence. I think that's a somewhat standard formulation. Of course, the liability could approach the total value realized from all sales of the software if they were sued by a class of customers, rather than a single customer.
"Fourth, if these little electric cars are supposed to provide inner city transportation, can they handle the hills in Pittsburgh?"
Ever hear of San Francisco? The electric buses seem to climb over the hills okay. (The bus system sucks, of course, but it's not a technological problem.)
To be fair, there are a lot of parallels between Pittsburgh & SF topologically, but the weather in Pittsburgh makes the hills quite a bit trickier in the winter; SF is hilly but on a grid anyway. If that were true in Pittsburgh, a lot of roads would be close to impassable during parts of the winter. However, there's got to be a better solution that what we have now!
Slashdot Mirror
WRT Open Source software, I see no a priori reason why OSS developers should be any less liable than commercial software companies, PROVIDED that certain reasonable guidelines apply:
- liability should never exceed the amount of money the developer/company *received* from the customer or class of customers unless gross negligence can be proved;
- in cases of gross negligence, the liability should coincide with the amount of *actual damages*; i.e., you don't get a million bucks because someone was able to read your web documents unless that act actually cost you $1 million in losses;
- developers should be reasonably shielded from liability in cases where the customer/user *actually* modified the software (not just *had* source available) -- if the modifications had a substantive affect on the security or safety of the product;
- parties can enter binding legal contracts to alter the balance of liability -- in instances where the customer *plans* to alter the software, whether they end up doing it or not. CLICKWRAP LICENSES DON'T COUNT.
These measures will only benefit the the software industry; serious programmers will have the satisfaction of working in a climate where time to market takes a back seat to quality (because the law penalizes nonconformance to this norm); software processes in the aggregate will improve for the same reason; customers and users will have a better experience with software in general and will have more respect for practioners who take the profession seriously.And people who lack confidence in their abilities to generate bug-free code can buy liability insurance, just like many other professionals currently do. In other words, software professionals can finally expect to *earn* the title!
I do have an 8500, but it's relegated to being my wife's web browsing machine.
It is arguably a stretch to call that "object-oriented" since OO has come to mean so many things besides data encapsulation (though in my view that's the most important OO feature, the part which allows you to replace the implementation at any time). But hey, it's certainly less of a stretch than Oracle calling 9i "unbreakable."
BTW, I don't think the Hurd is OO.
I once bought a PowerMac 7200 so that I would be able to run Apple's forthcoming next generation OS.
The fact is, most businesses gain value from software from *using* it, not selling it. So, they will always need people to build software for them, and they'll pay for it. But the only people who win when the software is proprietary are the *few* companies who sell it (and probably relatively few of the employees of those few compaines as well).
Except the mark isn't "666", it's "VISA."
AFAIK is not "mathematically impossible" to break even the strongest crypto available. It is "computationally infeasable." I.e., it's mathematically possible (by factoring all the large primes that could have been used for the key, for instance), but you can't afford the time/money (mostly time) required.
The bottom line AFAICT is you need to distribute all the files needed to build it with a certain baseline set of tools assumed.
"The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable."
It specifically mentions the compile and install scripts. So it's the plain interpretation, not a strict interpretation, that leads to the conclusion that one must include the project files, etc. You need to distribute whatever is needed to transform the source -- in the form that *you* the developer normally use to modify it -- into the object files you distribute.
Note, I don't think that there's anything wrong with this or controversial at all. How useful would GCC be to you without all of the Makefiles? Not very.
Read this before you start writing checks (unless you're loaded, in which case write as many as you want).
This is the final straw which has caused me to actually get off my ass and gather all of the legal info I could find about setting up a PAC (i.e, should be a 503 (1)(c)(4) nonprofit, and all of the accounting and tax rules). This is all from 26 USC 501 & 527.
I hope to read through all of this information, hopefully this weekend. If I think I understand it sufficiently I'll go through the process of actually registering a corporation (by default here in PA, unless anyone has any good reason it should be somewhere else and practical advice on doing it that way).
Assuming I get that far, I'll submit a story to Slashdot w/ all of the particulars inlcuding what I've done so far and what I think I'll need in terms of support resources (people, services, etc).
Anyone w/ any more ideas, info & suggestions is of course free to post here or email them to me.
Executive branch: enforces the laws passed by Congress. This includes bringing actions against violators and suggesting from among the statutory remedies.
Judicial branch: interprets laws passed by the legislative branch, and meets out punishment within the boundaries established in those laws (which are often left to the discretion of the judge).
I.e., MS is full of shit. The judge is free to completely disregard the opinions of the executive branch if the executive branch abrogates its responsibility to pursue thorough enforcement of the laws passed by the legislature. For example, plea agreements in criminal cases must be approved by the judge, and are in part accepted or denied on the basis of whether or not the executive branch is likely to prevail at trial on the original charge; the government can't (or rather, isn't supposed to be able to)-- if it has a good case -- just decide to let the defendant plead to a lesser charge (subject to statutory exceptions -- i.e., in exchange for testimony against another individual).
As for anyone who has specific case law which they think invalidates what I've said above in this specific context, well, IANALAIYAE,TSUB (I am not a lawyer and if you aren't either, then shut up, beeyatch).
While surely they could give the impression of bias, it would be difficult to imagine anyone other than the defendants in the case who wasn't inclined to share Judge Jackson's opinions about MS' behaviour in the courtroom. Meanwhile, the Judge's finding of facts and conclusions of law were largely upheld by the (previously pro-MS) appeals court, which should speak volumes about the soundness of the judgement in light of the facts.
Of course, one would have to offer these people quite a bit to get them to leave their niche and join the rolls of the gainfully employed, but that's no problem as they'd obviously primarily be management candidates anyway.
But for anything involving copyright and digital media -- where the worst harm imagineable doesn't impinge upon anyone's inalienable rights -- everyone is assumed to be a criminal, and is treated accordingly.
The fact is, Enron lied to all of its shareholders. It was not possible to make a rational decision regarding the value of their stock because the true value of the company was hidden.
While you are enough of an elitist to gain comfort from the intelligence of your abusers, I suspect that for most people the fact that they were ripped of by geniuses is cold comfort indeed.This is how we will be screwed; Microsoft submarine patents. And Miguel can claim as much as he wants that "lawyers" have told him that the patent issue can be avoided, but that kind of assumes that one knows about the patents -- which in the case of "submarine" patents is not the case. Even if a future MS patent on this stuff turns out to be invalid due to prior art, WHO IS GOING TO COURT AGAINST M$$$$$ FOR YEARS AND MUCHO $$$$$ TO PROVE IT?
Abandon hope all ye who enter here...
I think the crux of the issue is that there is one set of (somewhat difficult) skills, the mastery of which is a requisite to being a good engineer, another set for being a good manager. Good technical managers need to either:
- have all of the skills required to be a good manager and at least some of the skills (engineering judgement) of the engineer; or
- must be willing to defer to the expertise of the technical experts who work for them when it's appropriate.
As far as the first option goes, people that demonstrably have one or the other set of skills are hard enough to find; imagine how hard it must be to find people who have both! The second requires a kind of wisdom that I'd say most managers, like most people don't have. People in general are control freaks, professional managers even more so -- since a willingness to acceed control can be perceived as weakness, and often times what's required isn't just a willingness to say "yes" to those below, but the courage to say "no" to those above.Patents are going to screw any free software/OSS implementation of .NET, and anyone (i.e. GNOME) depending on it is going to be f**ked.
Microsoft has flat out said that it will use patents to defend it's .NET investment. Miguel claims that lawyer(s) have told him that the patent issue can be avoided.
DON'T BELIEVE IT (I don't; here's why):
ECMA rules stipulate RAND licensing of member's patents, and don't require disclosure of pending patents. Remember Miguel's quote about those, spiffy -- novel -- assemblies? What if MS has a broad patent filed to cover those? What if MS has a bunch of other overly broad patents in the pipeline to stuff that may seem non-novel, obvious, or based on prior art to you and I, but which a judge or patent examiner will give the benefit of the doubt?
Microsoft will use patents for things like this to force all implementers of the standard beside themselves to pay, OSS/Free Software or not.
And that's the good scenario! Suppose MS does a RAMBUS and says that they aren't LEGALLY bound by ECMA RAND policies? Then they could refuse to allow an OSS/FS implementation at all, and even if it wouldn't ultimately stand up in court, what free software company, developer or user can afford to fight one of the most deep-pocketed companies in existence in court on this one?
Abandon hope all ye who enter here...