I certainly hope so. I've been using random access voicemail here in Japan for the past year and it rocks. I dread my return back to Canada and the stone age of cellular...
<sarcasm>But that CAN'T BE! Apple INVENTED THIS! Just like Apple INVENTED touch screen phones, multitouch, windows, mice, and all other things good. You MUST be MISTAKEN!</sarcasm>
The legal test of a threat is whether a "reasonable person" would be threatened, ruling out paranoid personalities from taking over the legal system.
Well, it's not reasonable to be feel threatened by such threats. How many people post blogs like that each year? And how many people get killed by such bloggers? Now, compare that to the risk of crossing the street or getting into your car. It's not reasonable to be scared by threatening on-line speech.
I think the world (real as well as on line) would be far better off if free speech stopped just beyond my right to call somebody a moron and prevent me from doing anything or offering to do anything about it.
I don't think it would be better off, because "just beyond" is far too vague a definition; people could suppress inconvenient speech by simply tying up others in legislation forever.
IANAL, but defamation (and libel) require evidence that the guilty party have some knowledge that their statements are not factual.
I think it's quite obvious that the pictures and statements about sexual desires of the victim here were not factual and that the blogger should have known that. So, this may count as defamation. But it shouldn't count as a serious threat.
Yes, that language is disturbing, and the guy may even mean it. But what's the point of debating whether it is "acceptable"? If we made these kinds of people shut up, they'd still harbor the same thoughts and still be dangerous. The only difference would be that they'd be even harder to catch because there is no public record.
For any of us, there are people around who might do us harm. That's just a fact of life, and we can't change that by limiting speech. Either you face that, or you have to construct your life to be anonymous. People make both choices. Sierra has chosen a public life, and that puts her at greater risk, but it also has a lot of benefits.
I think the current lines are pretty well drawn in the US: if people defame you, you can sue and have them correct it. If people make clear, specific, immediate threats, you can get protection. If people make non-specific threats, however, there is little that you can do, and that's because restraining those people wouldn't help much, but putting restraints on that kind of speech might be abused by people who want to silence legitimate speech. In different words, much as I sympathize, "I felt threatened" shouldn't be sufficient to restrain speech.
"The Internet used to be a university. Then it became a shopping mall. But now, it's a war zone."
The Internet used to be populated by university students; that didn't make it "a university" any more than a frat party is "a university". Even 20 years ago, the Internet had big security holes, porn, death threats, flame wars, and all the other stuff that people get so upset about today. There was just less of it, there were fewer users who cared, and the legal system just didn't care.
It's not clear what is supposed to be new or innovative about the iPhone anyway. Touch-screen only phones have been around for a couple of years. A slim touch-screen-only phone that's even slimmer and sexier than the iPhone had won design prizes months before the iPhone was even announced.
The Gnome desktop has search for ODF built in, and mplayer probably has the largest range of format support of any multimedia player.
Where exactly is Google supposed to help?
Now, I think it would be nice if they ported Google Talk and Google Desktop Search, but I think the holdup there isn't their unwillingness, it's probably just that they are finding it tough to do and have other things to do.
If dir/ already exists, that will move newdir/ INTO dir/.
Of course it will. So what? The code I wrote illustrated how your requirement can be satisfied. In fact, this is not the way real UNIX programs usually do atomic operations (there are far better ways).
(I'm not sure about the use cases of this, if they are actually helpful in any real world programs, but that's not what I'm arguing.)
Yes, there are many kinds of operations you cannot make atomic with the UNIX file system primitives. My point is that you don't actually need those. Therefore, adding the bloat and complexity of a general purpose transaction manager to the kernel simply is not necessary.
He got the hang of things for a while, but then he interrupted a dpkg process when Synaptic was running by hitting the power button [...] The problem with these kinds of things is that if even one little glitch happens like this, the user gets stuck and then usually gives up and goes back to Windows.
Microsoft Windows also wedges when you hit the power button at the wrong point during an upgrade or install. In fact, Microsoft Windows wedges and messes up its configuration sooner or later even when you don't.
So, yes, Synaptic should recognize this case and handle it automatically. No, it's not a showstopper; it happens rarely enough and it's easy to fix.
Give me a program that will create two files in a directory while guaranteeing that no program can see only one of them.
mkdir newdir; touch newdir/file1; touch newdir/file2; mv newdir dir
On UNIX, dir/file1 and dir/file2 both appear at once, atomically.
This is only one of a large number of design possibilities for atomic updates/changes on UNIX. People tend to pick designs that offer robustness not only against interruptions, but also against other kinds of corruptions, which is probably why UNIX systems tend to be more robust than Windows.
The video was simply something designed to hit people's emotional buttons and make Clinton unappealing, using video material copied from an Apple ad; it was derivative and contained no intelligent thought or argument. You could make the same video about Obama or anybody else if you wanted to make them look bad.
If this kind of propaganda is what "citizen participation in the political process" is going to be, I want no part in it. This kind of ad would have been a low point even if it had been made by a political opponent, which is probably why Obama is distancing himself from it. Fortunately, it looks like this is not actually citizen participation, it's a political propagandist using his skills for his private amusement.
I've probably used half a dozen different parallel extensions of C and C++ over the years. This one doesn't look revolutionary, it merely looks painful.
This issue crops up in just about every major software system: database servers, web servers, mail server, multimedia servers, package install managers, etc. UNIX has been used for these kinds of applications for decades. Do you seriously think UNIX systems would be lacking a general purpose kernel transaction manager if it was useful for solving this issue? It's not rocket science to write one, and if it's not there, it's a good bet that it's not needed.
The functionality you want can be encapsulated neatly in user code, using standard UNIX file system semantics, with the same guarantees and behavior as if you had a kernel transaction manager.
The example in TFA that an entire install can be atomic: multiple filesystems, registry, everything appears complete and as requested, all at once, or it never happened.
UNIX has been able to do atomic installs/updates without a general purpose kernel transaction manager for decades; you simply don't need it for that. More importantly, a kernel transaction manager simply doesn't solve the problem because partial/inconsistent installs can exist for many other reasons besides interruptions; assuming that this case doesn't happen because you're using a transaction manager is a surefire way of making the system less robust.
Whether it belongs in the kernel or not is all but irrelevant: so what if it could be implemented as a userland service? Where they choose to put their code is up to them.
Well, it's "irrelevant" only if you don't care about things like performance, bug count, or reliability.
If it is, it will make building absolutely-bulletproof applications a whole hell of a lot easier. I know something about that. Being able to say ~`if (!quickcheck()) die(fromhere());`~ without leaving a mess means, just for starters, that you don't have to concoct a file format for complex data; you can just use the filesystem, and that choice won't complicate your life. Big win. Big.
If you write code like that, I hope you're actually just a CS undergraduate with an overactive imagination and people let you nowhere near writing applications that need to be "bullet-proof".
But being blinded by our own little wishes, and getting anti-thrust laws to be respected, is two things, EVEN in Europe.
Those aren't "our own little wishes". Microsoft has demonstrated over 20 years that they are able to manipulate and squash any non-open-source competitor through their business practices. That's why it is Microsoft's "little wish" to limit all licenses to non-open-source competitors.
The EU commission certainly has the right and the power to force Microsoft to make their protocols entirely public domain. There is no conceivable justification for keeping those protocols trade secret since the sole purpose of keeping them trade secret is for Microsoft to maintain its advantage that it has achieved through unfair business practices in the first place. Nobody, open source project or commercial competitor, should have to sign any agreement with Microsoft to interoperate with their systems.
Of course, you can do it in user space. You could do that in user space already in UNIX V7, and it's been used for anything from atomic system upgrades to maintaining gigabyte mail spool directories without data loss (something that Exchange doesn't seem to be able to achieve to this day). You really don't need a transaction manager in the kernel for that; any one of a number of existing kernel operations is sufficient.
There is little reason to put these kinds of transactional services into the kernel: they don't involve security or user permissions and they must be efficiently implementable in user code anyway (otherwise, most databases wouldn't work well on NT). So, I'd classify this as "kernel bloat".
the question is if MS distributing SuSE has impact on MS's claims against people using a version of Linux that they haven't indemnified
Yes. The question is whether that works. The GPL is pretty clear about the fact that the rights are transferable, so if Microsoft is considered to have "distributed" the software in some form, then the recipients will have the necessary rights (and, thereby, effectively, the whole community). The question is whether Microsoft can successfully argue that they aren't really "distributing" the software. I'm saying it's not at all clear that they can get away with that argument.
No - you don't "give up the right to assert a patent" through inaction.
That's correct. The point you're missing is that Microsoft has actually acted: they have distributed the software under the GPL themselves (that's part of the Novell/Microsoft agreement, and Microsoft is actually required to distribute the software). It looks like Microsoft therefore has granted transferable rights to all applicable intellectual property to the recipients of the software.
By selling access to their protocoles, they allow other companies to enter the market, which is the point in these suits.
They only allow closed-source companies to enter the market. No open source company can compete with Microsoft under these terms.
Since open source companies are actually the biggest competition for Microsoft at this point, that basically means that a closed-source only licensing policy is not a sufficient remedy for Microsoft's anti-competitive behavior. And there's a good chance that this won't be lost on the EU commission and that they will demand more.
This is about Microsoft's anti-competitive behaviors, not about Microsoft not being what we want.
Quite right. And that's why Microsoft needs to be forced to open up to the only real competition they have--open source companies.
The conclusion of the meeting? Nothing good is coming from this deal between Microsoft and Novell.
My understanding is that, as part of the deal, Microsoft is actually distributing SuSE Linux.
Doesn't this mean that they themselves are distributing the software they might be claiming patents on? And doesn't that mean that, for practical purposes, have given up their right to assert the patents against any GPL'ed software that is part of SuSE Linux?
I'm sure this wasn't Microsoft's intention, but it looks to me like it's a result of this deal.
Companies like Microsoft and Sun are fond of calling things "open source" that really aren't open source in the usual sense. So, which license is this going to be?
I think that's pretty obvious: investing 1h now into organizing things you're never going to look at again is likely wasted. OTOH, if you can't do your taxes because you lost all your tax documents, need to spend many hours trying to get new copies, and pay stiff penalties, that's no good either.
So, you really need "just enough" organization to get the things done you need to get done. That means that most paperwork can just go into a big filing box, but a few things need to be kept separate.
However, the term "slob" has connotations of dirt and bad odors; if all your interactions are on-line, it may be efficient not to worry about that either, but if you occasionally want to meet up with people in person, some degree of cleanliness is probably a good idea.
Music players are trivial applications with a trivial UI. Show me a web browser or mail client written in Apollo.
And, in any case, even if Apollo were as good or better than XULrunner, it's proprietary, which automatically disqualifies it as a reasonable long-term choice for cross-platform runtime.
Slashdot Mirror
There is a Carbon port of OO.o in development.
Why this duplication of effort? NeoOffice works great, and it actually has more functionality than OOo.
I certainly hope so. I've been using random access voicemail here in Japan for the past year and it rocks. I dread my return back to Canada and the stone age of cellular...
<sarcasm>But that CAN'T BE! Apple INVENTED THIS! Just like Apple INVENTED touch screen phones, multitouch, windows, mice, and all other things good. You MUST be MISTAKEN!</sarcasm>
The legal test of a threat is whether a "reasonable person" would be threatened, ruling out paranoid personalities from taking over the legal system.
Well, it's not reasonable to be feel threatened by such threats. How many people post blogs like that each year? And how many people get killed by such bloggers? Now, compare that to the risk of crossing the street or getting into your car. It's not reasonable to be scared by threatening on-line speech.
I think the world (real as well as on line) would be far better off if free speech stopped just beyond my right to call somebody a moron and prevent me from doing anything or offering to do anything about it.
I don't think it would be better off, because "just beyond" is far too vague a definition; people could suppress inconvenient speech by simply tying up others in legislation forever.
IANAL, but defamation (and libel) require evidence that the guilty party have some knowledge that their statements are not factual.
I think it's quite obvious that the pictures and statements about sexual desires of the victim here were not factual and that the blogger should have known that. So, this may count as defamation. But it shouldn't count as a serious threat.
Yes, that language is disturbing, and the guy may even mean it. But what's the point of debating whether it is "acceptable"? If we made these kinds of people shut up, they'd still harbor the same thoughts and still be dangerous. The only difference would be that they'd be even harder to catch because there is no public record.
For any of us, there are people around who might do us harm. That's just a fact of life, and we can't change that by limiting speech. Either you face that, or you have to construct your life to be anonymous. People make both choices. Sierra has chosen a public life, and that puts her at greater risk, but it also has a lot of benefits.
I think the current lines are pretty well drawn in the US: if people defame you, you can sue and have them correct it. If people make clear, specific, immediate threats, you can get protection. If people make non-specific threats, however, there is little that you can do, and that's because restraining those people wouldn't help much, but putting restraints on that kind of speech might be abused by people who want to silence legitimate speech. In different words, much as I sympathize, "I felt threatened" shouldn't be sufficient to restrain speech.
"The Internet used to be a university. Then it became a shopping mall. But now, it's a war zone."
The Internet used to be populated by university students; that didn't make it "a university" any more than a frat party is "a university". Even 20 years ago, the Internet had big security holes, porn, death threats, flame wars, and all the other stuff that people get so upset about today. There was just less of it, there were fewer users who cared, and the legal system just didn't care.
It's not clear what is supposed to be new or innovative about the iPhone anyway. Touch-screen only phones have been around for a couple of years. A slim touch-screen-only phone that's even slimmer and sexier than the iPhone had won design prizes months before the iPhone was even announced.
So, what's there to "upstage" anyway?
The Gnome desktop has search for ODF built in, and mplayer probably has the largest range of format support of any multimedia player.
Where exactly is Google supposed to help?
Now, I think it would be nice if they ported Google Talk and Google Desktop Search, but I think the holdup there isn't their unwillingness, it's probably just that they are finding it tough to do and have other things to do.
If dir/ already exists, that will move newdir/ INTO dir/.
Of course it will. So what? The code I wrote illustrated how your requirement can be satisfied. In fact, this is not the way real UNIX programs usually do atomic operations (there are far better ways).
(I'm not sure about the use cases of this, if they are actually helpful in any real world programs, but that's not what I'm arguing.)
Yes, there are many kinds of operations you cannot make atomic with the UNIX file system primitives. My point is that you don't actually need those. Therefore, adding the bloat and complexity of a general purpose transaction manager to the kernel simply is not necessary.
He got the hang of things for a while, but then he interrupted a dpkg process when Synaptic was running by hitting the power button [...] The problem with these kinds of things is that if even one little glitch happens like this, the user gets stuck and then usually gives up and goes back to Windows.
Microsoft Windows also wedges when you hit the power button at the wrong point during an upgrade or install. In fact, Microsoft Windows wedges and messes up its configuration sooner or later even when you don't.
So, yes, Synaptic should recognize this case and handle it automatically. No, it's not a showstopper; it happens rarely enough and it's easy to fix.
Did you file a bug report?
Give me a program that will create two files in a directory while guaranteeing that no program can see only one of them.
mkdir newdir; touch newdir/file1; touch newdir/file2; mv newdir dir
On UNIX, dir/file1 and dir/file2 both appear at once, atomically.
This is only one of a large number of design possibilities for atomic updates/changes on UNIX. People tend to pick designs that offer robustness not only against interruptions, but also against other kinds of corruptions, which is probably why UNIX systems tend to be more robust than Windows.
The video was simply something designed to hit people's emotional buttons and make Clinton unappealing, using video material copied from an Apple ad; it was derivative and contained no intelligent thought or argument. You could make the same video about Obama or anybody else if you wanted to make them look bad.
If this kind of propaganda is what "citizen participation in the political process" is going to be, I want no part in it. This kind of ad would have been a low point even if it had been made by a political opponent, which is probably why Obama is distancing himself from it. Fortunately, it looks like this is not actually citizen participation, it's a political propagandist using his skills for his private amusement.
I've probably used half a dozen different parallel extensions of C and C++ over the years. This one doesn't look revolutionary, it merely looks painful.
But who calls that livin'
When no gal will give in
To no man what's nine hundred years
This issue crops up in just about every major software system: database servers, web servers, mail server, multimedia servers, package install managers, etc. UNIX has been used for these kinds of applications for decades. Do you seriously think UNIX systems would be lacking a general purpose kernel transaction manager if it was useful for solving this issue? It's not rocket science to write one, and if it's not there, it's a good bet that it's not needed.
The functionality you want can be encapsulated neatly in user code, using standard UNIX file system semantics, with the same guarantees and behavior as if you had a kernel transaction manager.
The example in TFA that an entire install can be atomic: multiple filesystems, registry, everything appears complete and as requested, all at once, or it never happened.
UNIX has been able to do atomic installs/updates without a general purpose kernel transaction manager for decades; you simply don't need it for that. More importantly, a kernel transaction manager simply doesn't solve the problem because partial/inconsistent installs can exist for many other reasons besides interruptions; assuming that this case doesn't happen because you're using a transaction manager is a surefire way of making the system less robust.
Whether it belongs in the kernel or not is all but irrelevant: so what if it could be implemented as a userland service? Where they choose to put their code is up to them.
Well, it's "irrelevant" only if you don't care about things like performance, bug count, or reliability.
If it is, it will make building absolutely-bulletproof applications a whole hell of a lot easier. I know something about that. Being able to say ~`if (!quickcheck()) die(fromhere());`~ without leaving a mess means, just for starters, that you don't have to concoct a file format for complex data; you can just use the filesystem, and that choice won't complicate your life. Big win. Big.
If you write code like that, I hope you're actually just a CS undergraduate with an overactive imagination and people let you nowhere near writing applications that need to be "bullet-proof".
But being blinded by our own little wishes, and getting anti-thrust laws to be respected, is two things, EVEN in Europe.
Those aren't "our own little wishes". Microsoft has demonstrated over 20 years that they are able to manipulate and squash any non-open-source competitor through their business practices. That's why it is Microsoft's "little wish" to limit all licenses to non-open-source competitors.
The EU commission certainly has the right and the power to force Microsoft to make their protocols entirely public domain. There is no conceivable justification for keeping those protocols trade secret since the sole purpose of keeping them trade secret is for Microsoft to maintain its advantage that it has achieved through unfair business practices in the first place. Nobody, open source project or commercial competitor, should have to sign any agreement with Microsoft to interoperate with their systems.
Of course, you can do it in user space. You could do that in user space already in UNIX V7, and it's been used for anything from atomic system upgrades to maintaining gigabyte mail spool directories without data loss (something that Exchange doesn't seem to be able to achieve to this day). You really don't need a transaction manager in the kernel for that; any one of a number of existing kernel operations is sufficient.
There is little reason to put these kinds of transactional services into the kernel: they don't involve security or user permissions and they must be efficiently implementable in user code anyway (otherwise, most databases wouldn't work well on NT). So, I'd classify this as "kernel bloat".
the question is if MS distributing SuSE has impact on MS's claims against people using a version of Linux that they haven't indemnified
Yes. The question is whether that works. The GPL is pretty clear about the fact that the rights are transferable, so if Microsoft is considered to have "distributed" the software in some form, then the recipients will have the necessary rights (and, thereby, effectively, the whole community). The question is whether Microsoft can successfully argue that they aren't really "distributing" the software. I'm saying it's not at all clear that they can get away with that argument.
No - you don't "give up the right to assert a patent" through inaction.
That's correct. The point you're missing is that Microsoft has actually acted: they have distributed the software under the GPL themselves (that's part of the Novell/Microsoft agreement, and Microsoft is actually required to distribute the software). It looks like Microsoft therefore has granted transferable rights to all applicable intellectual property to the recipients of the software.
By selling access to their protocoles, they allow other companies to enter the market, which is the point in these suits.
They only allow closed-source companies to enter the market. No open source company can compete with Microsoft under these terms.
Since open source companies are actually the biggest competition for Microsoft at this point, that basically means that a closed-source only licensing policy is not a sufficient remedy for Microsoft's anti-competitive behavior. And there's a good chance that this won't be lost on the EU commission and that they will demand more.
This is about Microsoft's anti-competitive behaviors, not about Microsoft not being what we want.
Quite right. And that's why Microsoft needs to be forced to open up to the only real competition they have--open source companies.
The conclusion of the meeting? Nothing good is coming from this deal between Microsoft and Novell.
My understanding is that, as part of the deal, Microsoft is actually distributing SuSE Linux.
Doesn't this mean that they themselves are distributing the software they might be claiming patents on? And doesn't that mean that, for practical purposes, have given up their right to assert the patents against any GPL'ed software that is part of SuSE Linux?
I'm sure this wasn't Microsoft's intention, but it looks to me like it's a result of this deal.
Companies like Microsoft and Sun are fond of calling things "open source" that really aren't open source in the usual sense. So, which license is this going to be?
(Apologies if this in TFA, but I didn't see it.)
I think that's pretty obvious: investing 1h now into organizing things you're never going to look at again is likely wasted. OTOH, if you can't do your taxes because you lost all your tax documents, need to spend many hours trying to get new copies, and pay stiff penalties, that's no good either.
So, you really need "just enough" organization to get the things done you need to get done. That means that most paperwork can just go into a big filing box, but a few things need to be kept separate.
However, the term "slob" has connotations of dirt and bad odors; if all your interactions are on-line, it may be efficient not to worry about that either, but if you occasionally want to meet up with people in person, some degree of cleanliness is probably a good idea.
Music players are trivial applications with a trivial UI. Show me a web browser or mail client written in Apollo.
And, in any case, even if Apollo were as good or better than XULrunner, it's proprietary, which automatically disqualifies it as a reasonable long-term choice for cross-platform runtime.