Whenever someone points out that the web suffers link-rot, they demonstrate that they're not thinking clearly about robustness in large distributed systems.
You don't have the choice between link-rot and utopia. Your choice is between a single centralised point of failure, or many points of 'partial failure'. Thankfully, the web gives us the second option. We even have archive.org to take the edge off. Unfortunately, of course, we now have silos, which take us back to the first option.
That's one of the problems IPFS addresses by making links based on a content hash, not the current storage location.
Eh? So if I make a correction, the address of the resource changes? On the web, you have the choice. e.g. on Wikipedia, https://en.wikipedia.org/wiki/InterPlanetary_File_System gives you the latest page, and https://en.wikipedia.org/w/index.php?title=InterPlanetary_File_System&oldid=818883368 will always give you the snapshot from the 6th of January 2018. Similar schemes exist on GitHub and BitBucket.
(Disclaimer: I'm just being snarky and don't really know much about Xanadu, or distributed databases, or IPFS. I'd be glad to be corrected.)
the responsibility of government to function within whatever means the people allow it. That's the true Libertarian utopia.
That would be democracy, not libertarianism.
You seem to want to imply that big government is necessarily an over-reach beyond what people want. Not so.
I'm British. I like that my country has a 'big government'. I like that we have an NHS, public transport, libraries, museums, paid for with my tax pounds. I oppose the recent slash-and-burn strategy of our Conservative government.
I'm a civil libertarian, to be sure, but I'm no libertarian. I'm glad there's a big government, and that everyone is made to pay their share (yes, by force). The real results are what matter.
You just invoked an anecdote as if it were a refutation of an statistical claim. Not any better than making a statistical claim without citation. Neither of you are worth taking seriously. You ACs never change.
Crowing "they're trying to control us!" over a law you dont like ignores this very simple fact of reality.
Or, to someone with a libertarian streak, it's precisely the opposite: when making laws, people too quickly forget that, by definition, every law infringes on liberty.
But Pascal, C, Java, C++ and Ada all belong to the same programming paradigm, with or without the object orientation twist.
But there are very significant differences regarding how bug-prone the languages are. Unlike C and C++, Java has no concept of undefined behaviour, and does a far more complete job of insulating the programmer from platform-specific quirks. It's night and day.
Unlike C/C++, Java enforces left-to-right evaluation order, always uses 2's complement arithmetic (not target-platform specific), always uses big endian (not target-platform specific), always initialises to zero (never UB), always performs array bounds checking (never UB), always checks for null when dereferencing (never UB), always checks for invalid casts at runtime (never UB), the list goes on and on.
Yes, both C++ and Java have object-orientation, but their philosophies are completely different. (Perhaps I'm just talking past you here though - you're right of course that they're both procedural/OO at the end of the day.)
You were going pretty strong up to the last line. (I happen to disagree that proprietary software is inherently immoral, but that's not the point here.)
Those business models are legitimate and I completely support making money from FOSS that way, but it's not right to call it 'selling free software'.
It seems to me that web caching of this sort is essentially the only argument for using plain old HTTP.
TheRaven has already made a strong case against the think-of-the-CPU-overhead argument, but breaking web caching is a legitimate downside of HTTPS. Ideally there'd be an automatic checksum check after downloading over plain HTTP. Fun fact: this is precisely what Steam does. If they used a proprietary protocol, or HTTPS, then caching (whether by ISPs or by 'local' sysadmins) wouldn't be possible, to the detriment of both the customers and Valve.
Ha, you got me, I'm a Slashdotter. And now I see the whole section of the article on exactly that topic:-P
Tanenbaum seems to be using a curious line of reasoning there, though.
secretive use of Free Software inside what Tanenbaum himself calls a “spy engine”, can also be said to explicitly contradict the intent of the author of the Free Software. As Tanenbaum writes: “I certainly wouldn’t have cooperated [...]” One might therefore expect that it would have been costly for Intel to buy Tanenbaum’s cooperation for secret distribution of Minix 3, had it been at all possible.
I don't see why we should care about how much Intel would have had to have paid Tanenbaum to get him onboard. Their mistake was 'only' in failing to attribute, not in failing to hire him as a consultant.
the amount of damages that can be argued for, is remarkable
They can argue whatever they like, but there's no precedent here as far as I'm aware.
ACs would post intelligent comments, everyone would agree that Vim beats Emacs, that the GPL is better than BSD, that KDE is better than Gnome, and that low user IDs deserve respect.
Great, you just invented a punishment out of whole cloth. That's now how the law works.
What sort of damages can Minix claim, here?
This is one of the problems with copyleft/copycenter licences: they're based on copyright, which is itself based on the profit motive. The software was never being sold in the first place, so damages don't apply in the same way.
Slashdot Mirror
Yes, yes. There's a reason Ted Nelson's Xanadu idea never panned out.
Whenever someone points out that the web suffers link-rot, they demonstrate that they're not thinking clearly about robustness in large distributed systems.
You don't have the choice between link-rot and utopia. Your choice is between a single centralised point of failure, or many points of 'partial failure'. Thankfully, the web gives us the second option. We even have archive.org to take the edge off. Unfortunately, of course, we now have silos, which take us back to the first option.
That's one of the problems IPFS addresses by making links based on a content hash, not the current storage location.
Eh? So if I make a correction, the address of the resource changes? On the web, you have the choice. e.g. on Wikipedia, https://en.wikipedia.org/wiki/InterPlanetary_File_System gives you the latest page, and https://en.wikipedia.org/w/index.php?title=InterPlanetary_File_System&oldid=818883368 will always give you the snapshot from the 6th of January 2018. Similar schemes exist on GitHub and BitBucket.
(Disclaimer: I'm just being snarky and don't really know much about Xanadu, or distributed databases, or IPFS. I'd be glad to be corrected.)
You can't? Here in the UK you can bet on just about anything. I had a few pounds on Clinton - far shorter than 2:1 odds though.
If it's that easy, then why don't more people do it?
A lot of people treat their 'news' sources as a medium of entertainment.
The bookies were offering something like 7 to 1 on Trump. There's not much more to say.
the responsibility of government to function within whatever means the people allow it. That's the true Libertarian utopia.
That would be democracy, not libertarianism.
You seem to want to imply that big government is necessarily an over-reach beyond what people want. Not so.
I'm British. I like that my country has a 'big government'. I like that we have an NHS, public transport, libraries, museums, paid for with my tax pounds. I oppose the recent slash-and-burn strategy of our Conservative government.
I'm a civil libertarian, to be sure, but I'm no libertarian. I'm glad there's a big government, and that everyone is made to pay their share (yes, by force). The real results are what matter.
If only western governments were starved of the tax funds they need in order to function! We'd finally have our libertarian utopia!
True, but as far as I can see Apple have never done that.
You have no credibility.
You just invoked an anecdote as if it were a refutation of an statistical claim. Not any better than making a statistical claim without citation. Neither of you are worth taking seriously. You ACs never change.
Crowing "they're trying to control us!" over a law you dont like ignores this very simple fact of reality.
Or, to someone with a libertarian streak, it's precisely the opposite: when making laws, people too quickly forget that, by definition, every law infringes on liberty.
But Pascal, C, Java, C++ and Ada all belong to the same programming paradigm, with or without the object orientation twist.
But there are very significant differences regarding how bug-prone the languages are. Unlike C and C++, Java has no concept of undefined behaviour, and does a far more complete job of insulating the programmer from platform-specific quirks. It's night and day.
Unlike C/C++, Java enforces left-to-right evaluation order, always uses 2's complement arithmetic (not target-platform specific), always uses big endian (not target-platform specific), always initialises to zero (never UB), always performs array bounds checking (never UB), always checks for null when dereferencing (never UB), always checks for invalid casts at runtime (never UB), the list goes on and on.
Yes, both C++ and Java have object-orientation, but their philosophies are completely different. (Perhaps I'm just talking past you here though - you're right of course that they're both procedural/OO at the end of the day.)
You were going pretty strong up to the last line. (I happen to disagree that proprietary software is inherently immoral, but that's not the point here.)
Those business models are legitimate and I completely support making money from FOSS that way, but it's not right to call it 'selling free software'.
That's the spirit! Light up the flame war and let's all gather round. 'tis the season!
Do we have to continue pretending that 'selling' applies to Free software in the same way it applies to payware?
If you're not required to pay in order to get access to the product, that's not what people normally mean by 'selling' your product.
With the Internet, distributing software is trivial. Does anyone make money selling FOSS? (And no, that doesn't count support contracts.)
It seems to me that web caching of this sort is essentially the only argument for using plain old HTTP.
TheRaven has already made a strong case against the think-of-the-CPU-overhead argument, but breaking web caching is a legitimate downside of HTTPS. Ideally there'd be an automatic checksum check after downloading over plain HTTP. Fun fact: this is precisely what Steam does. If they used a proprietary protocol, or HTTPS, then caching (whether by ISPs or by 'local' sysadmins) wouldn't be possible, to the detriment of both the customers and Valve.
Pretty safe bet, now that Intel CPU + AMD GPU on a chip has been announced.
Not the same thing.
Interesting ideas. Is there precedence here? This can't be the first time a product has been shipped that infringes copyrights.
Ha, you got me, I'm a Slashdotter. And now I see the whole section of the article on exactly that topic :-P
Tanenbaum seems to be using a curious line of reasoning there, though.
secretive use of Free Software inside what Tanenbaum himself calls a “spy engine”, can also be said to explicitly contradict the intent of the author of the Free Software. As Tanenbaum writes: “I certainly wouldn’t have cooperated [...]” One might therefore expect that it would have been costly for Intel to buy Tanenbaum’s cooperation for secret distribution of Minix 3, had it been at all possible.
Except of course for Freedom 0. Neither the GPL nor BSD-style licences put any stock in the author's distaste for potential applications of their work. If they did, they wouldn't qualify as Free Software licences.
I don't see why we should care about how much Intel would have had to have paid Tanenbaum to get him onboard. Their mistake was 'only' in failing to attribute, not in failing to hire him as a consultant.
the amount of damages that can be argued for, is remarkable
They can argue whatever they like, but there's no precedent here as far as I'm aware.
Keep it simple: compile a custom kernel!
ACs would post intelligent comments, everyone would agree that Vim beats Emacs, that the GPL is better than BSD, that KDE is better than Gnome, and that low user IDs deserve respect.
But no.
Thanks, systemd.
One shouldn't blame the AC for contributing precisely nothing to the conversation. It is their natural manner.
(I realise the irony in my own empty snark, here. Ah well.)
Net Neutrality is basically government creating monopolies
What on Earth are you talking about? The ISP monopolies have been around far longer than the recent net neutrality debate.
What the government isn't doing is actually promoting competition by deregulating.
Net neutrality is good. Abusive monopolies are bad. It's not hard.
Great, you just invented a punishment out of whole cloth. That's now how the law works.
What sort of damages can Minix claim, here?
This is one of the problems with copyleft/copycenter licences: they're based on copyright, which is itself based on the profit motive. The software was never being sold in the first place, so damages don't apply in the same way.
...what?
Good point.