Software is explicitely excluded from patentability by article 52-2c of the 1973 European Patent Convention, as are games and mathematics.
Therefore, there are no software patents in any of the member countries.
Thanks for the link, I was not aware of that.
But then you say:
Now if the European Patent Office could stop to grant software patents - against its own rules ! - that have no legal value and aren't enforceable
So which is it?:-)
I'm from EU (France) and not US, and am not aware of all subtilities of the US legal system, but it would seem logical that when a case is handled by a federal court, federal law prevails state law, right ?
I'm not from the US either, that was for the benefit of the (mostly US-based, I presume) majority of/. readers. What I meant was that even though there is no EU law (= federal law) saying <something>, there still might be a national law (= state law) that does...
What intellectual property rights ? The EU Commision didn't ask for the source code (copyright), and software patents have no legal value in Europe...
Even without any EU directive on software patents, we might still have software patents in the member countries. Compare with federal/state law in the US, for example.
So I guess that nerds (who can't get a girlfriend as easily as everybody else) deviate their attention to two things: Videogames, tech stuff, and porn.
And with faster net connections it's easier than ever to DL and have a look!
In slashdot's defense, the odds (seem to be so far that there) will be some moderator who recognizes the subtle humor and moderates appropriatley. To end users, the "right" answer will probably eventually appear.
Point taken.
Also, a couple of my comments just got modded up, so I have to agree.:-)
How's this going to work? hehe, How long before we buy our PCs from China?
Another scenario for the conspiracy theorists:
1. Implement legally unexportable cryptographic algorithm in hardware.
2. Lobby for said algorithm/hardware to be mandated for government agencies' equipment. (because it's about homeland security, you know)
3. Profit!!!
Interesting report but I would like to see more details, what type of encryption is being used?
From TFA:
Secure Blue requires a few circuits to be added to a microprocessor, taking up a small percentage of the overall silicon real estate, according to IBM. The encryption and decryption happens on-the-fly, without any processor overhead, the company said.
(emphasis added by me)
It would seem to me that the highlighted phrases above would set some sort of bound on how sophisticated this encryption can be. Besides, there's probably a bigger incentive to offer encryption that is "good enough for DRM" than there is for offering encryption that is "uncrackable within the nearest 1000 years, even by governments". (Yeah, I realize I'm not adding much useful saying that...)
Besides, it just looks like something to dangle in front of "content providers" (encrypted all the way up to the last step, etc). Paranoid geeks are, I guess, more likely to depend on software tools (PGP/GPG, encrypted partitions) to keep data safe.
I will only address one point of the optimistic idealism I saw in several parts of the article, there are others:
"A company that sues competitors for patent infringement is like a a defender who has been beaten so thoroughly that he turns to plead with the referee."
This point is made in the context of other statements that indicate this is the main reason that a company starts suing for patent infringement. The reality is there are companies that have no developers at all, just lawyers, whose sole purpose is to seek out and buy patents and pressure other companies for licensing fees.
Yes, and he mentions them a page further down. You did read the article before posting, didn't you? Of course you did.
While they may have interrupts or timers, you don't have to use them.
But the parent suggested this strategy in the case where all the timers are in use, and I'd imagine those timers using interrupts to implement callbacks (since the alternative is polling, which is very inefficient). And what if the thing you are controlling requires interrupts - I doubt they were put to PICs needlessly, since that would be a waste of resources ?
Hm... yes, he did say that. In such cases, I imagine you'd use the timers for long-term timekeeping/synchronization, and delay loops for short-term delays (we must wait x ns before the A/D presents a valid output, sort of thing). In such cases, I at least, would mask off interrupts during those short intervals and handle the resulting mess. (Wouldn't necessary be that bad, but you might have to worry about latency, etc.)
Or use polling. It may seem inefficient, but since we're single-tasking anyway, why not? Assuming a free-running timer, you wouldn't get very precise timing, but the long-term stability would be there at least. Good enough for periodically updating a display/clock/whatever.
I'm also not convinced that this method of counting time is very accurate - just how accurate is the clock pulse generator anyway ? Seems to me that as long as it generates pulses with a frequency that stays within certain limits, the accuracy is unimportant to the operation [...]
Everything would typically be running off the same clock anyway, including the timers. If we're using an external oscillator/crystal timekeeping would be pretty damn accurate over short times. (And pretty damn useless over longer periods. Just like my desktop computer, in other words...)
Most PICs (all?) can also run from an internal timing source, with just a few external components (resistor/capacitor). This is not very precise at all, but, as you say, is good enough to keep the parts synchronized. I haven't used it though.
Interesting. I notice it says on the page that it is around 700K. That's on the order of half the size of the binary I have, which is nice.
But X is probably overkill for what these laptops do, anyway. You hardly need the networking bits, for example, and you only need to support a single chipset. And with networking gone, there's no need for X's voodoo flavour security either...
Yes, X is hard to use. How many end users do you know who have mastered xauth, for example? I know none. The usual solution is "xhost +localhost", coupled with "SSH will take care of it" for remote access.
I'm an old Amiga user. The Amiga's windowing system fit in a 256 or 512K ROM, together with pretty much all the other parts of a multitasking operating system. X in 700K? I'm not impressed.
OK, so I'm just a whining ex-Amiga user. Fair enough. Let's see what else is on the menu. QNX's Photon is, what, 40-50K? Plan 9's Rio? I don't know the size, but 8 1/2 apparently was around 100K. What about GNUStep/Display Ghostscript? I bet that if we shop around, we can do better than 700K. So let's do it.
My point isn't "X sucks"; it is that Linux, or BSD, or any free/open OS does not necessarily have to run X. And if you think about this laptop not as a standard desktop Unix platform, but as something that has to do just a few specific tasks well, it becomes an embedded system, with a whole new set of priorities. It just doesn't have to be able to run all existing applications. And that gives us (well, them...) freedom to pick whatever lets them minimize the cost of the hardware.
Sorry if this turned into a long rant, it's not meant as an attack on you or your implementation of X. I'm sure it's great for where it's meant to be used.
I tried one of Anonymizers rivals, and it worked just fine. As far as any website could see, I was in California. (Because that's where the exit point was.) Google's main page, ads ("Girls in %s want to meet you"), everything.
Of course, they filtered out Javascript and any plug-in content (for anonymity), so it wouldn't work with these services anyway. See, another reason why embedded players are bad. Just give me the download links...
But I'm sure there's some anonymizing service out there that'll work. Anyone serious about anonymity will block JS and plugins by default, though.
A PIC is not like a modern general-purpose multi-gigahertz CPU. They're too puny to load a huge operating system on top of them, for one thing. So you have to do everything yourself. On the up side, you have total control and there's nothing there to interfere with your program.
So, if you don't want to be disturbed by interrupts, you just don't initialize the hardware that might generate them. Or you mask them off in timing-critical sections of the code. While they may have interrupts or timers, you don't have to use them.
Something odd just occurred to me. Couldn't you CDR RAID an important archive? Similar to hardrive RAIDing, you could save the data bits across multiple CDRs, with an extra CDR for a checksum. Then, if any CDR failed, you could rebuild it from the others.
Or you could just make 2 copies. But where's the fun in that?;)
Exactly, there's no fun in that.
I've thought about it myself, you could use something like parchive to make parity blocks and burn them to a "parity" CD/DVD. Shouldn't be too hard to script something together in Python... It won't be a direct analog to RAID, but who cares. And you wouldn't need to muck around with some big restore/untarring procedure, the original data will still be there as plain files.
Ideally, you should probably automatically generate parity blocks/discs whenever you burn a CD/DVD. Something like, say, 10-15% overhead isn't going to hurt much, at least when you're cleaning out lots of stuff.
Hmm. I should sit down and sketch out something to do this. I'm sure I'll get around to it some day.:-)
What you seem to be talking about is "freesites", websites on Freenet. Freesites is not the only type of content on Freenet. Just as Web sites is not the only thing on the Internet. Even with freesites though, there's not so much of waiting as you seem to be making it out to be. (Yes, a fresh install will suck until it has integrated itself into the network. It does get better. Yes, it's still slow.)
Also similar to Internet vs. Web, you can write your own clients for Freenet (for the Freenet protocol). You're not limited to what the developers have provided for you out of the box.
0.7 will, if I understand things correctly, support almost exactly the scenario you describe (queuing downloads within the main Freenet program itself). The spidering part you'll have to write yourself though. Or...
...use Frost, probably the real killer app for Freenet right now; a message board / filesharing client.
(And parent got modded "Insightful" for what amounts to "Internet == Web"? Yikes.)
The Wikipedia article on Freenet has a screenshot of browsing a freesite, that's probably as close as you'll get. (A "screenshot of Freenet" would probably be about as exciting and obvious as a "screenshot of TCP/IP".)
(This is all from my understanding of the network, which might not be perfect.)
It is different in that you can easily connect to all your friends-of-friends-of-friends-of-friends-of-friend s in Iran, which might be a little too much work to set up "manually".
Or you could make the comparison that your friends and friends-of-friends network (which is a perfectly fine thing to have) is somewhat like being able to dial local BBS'es (like in the good old days), whereas Freenet is more like having access to the Internet. In more ways than one too; there's going to be more to do than "just" exchanging files.
Opera users especially, are aware of the existance of Firefox, so what is keeping the relatively small portion of Opera users from switching to Firefox, and what's still drawing in new users into the Opera club?
Well, I think it might actually be the fact that Opera is good at what it does. (Gasp!)
(Why should Opera users particularly want to switch to Firefox anyway? Is there an offer of 72 virgin brides in Heaven after you die or something that makes it worthwhile?)
Seriously, I'd say it's because Opera just gets the job done. You can pimp your FF installation all you want and it will have a ton of features not present in Opera, but for 90% of what you do, all you'll ever need is already present in the base install (of Opera). And it's not a particularly large download either. Plus it really is fast.
Whether you should migrate... well, you should at least try it for a week, then you'll know enough to decide which was the best for you. You can always switch back.
Good point, but I think I disagree with your definition of "shell".
(Maybe this is what you mean by me breaking the definition of just that. So it would be my very own "command shüll". Or "khell". Whatever, you and I both know what I'm really talking about.):-)
Seriously though, to me a "shell" is just an interface to the operating system. You have a different definition and that's fine with me.
My problem with escape codes is that they are by definition "tacked on afterwards". If you know you're going to use something, why not make it part of the specification? IMHO limiting yourself to what the original TTYs and VT100s could do is rather artificial when we all use terminal emulators that can do ANSI colors, etc, anyway.
Sorry if this is kind of unstructured, I'm really tired right now.
Slashdot Mirror
But then you say: So which is it?
Go ahead and mod me funny, I dare you. :-)
Also, a couple of my comments just got modded up, so I have to agree. :-)
Triple-ROT52? +3 Informative? Shame on you, moderators.
1. Implement legally unexportable cryptographic algorithm in hardware.
2. Lobby for said algorithm/hardware to be mandated for government agencies' equipment. (because it's about homeland security, you know)
3. Profit!!!
Or something like that...
It would seem to me that the highlighted phrases above would set some sort of bound on how sophisticated this encryption can be. Besides, there's probably a bigger incentive to offer encryption that is "good enough for DRM" than there is for offering encryption that is "uncrackable within the nearest 1000 years, even by governments". (Yeah, I realize I'm not adding much useful saying that...)
Besides, it just looks like something to dangle in front of "content providers" (encrypted all the way up to the last step, etc). Paranoid geeks are, I guess, more likely to depend on software tools (PGP/GPG, encrypted partitions) to keep data safe.
My guess: In media center PCs in 3... 2... 1...
Or use polling. It may seem inefficient, but since we're single-tasking anyway, why not? Assuming a free-running timer, you wouldn't get very precise timing, but the long-term stability would be there at least. Good enough for periodically updating a display/clock/whatever.
Everything would typically be running off the same clock anyway, including the timers. If we're using an external oscillator/crystal timekeeping would be pretty damn accurate over short times. (And pretty damn useless over longer periods. Just like my desktop computer, in other words...)Most PICs (all?) can also run from an internal timing source, with just a few external components (resistor/capacitor). This is not very precise at all, but, as you say, is good enough to keep the parts synchronized. I haven't used it though.
"If I can't get any, then neither should anyone else!"
Hmm, suddenly seems like a bad idea to post to Slashdot about "new, affordable" bioweapons...
But X is probably overkill for what these laptops do, anyway. You hardly need the networking bits, for example, and you only need to support a single chipset. And with networking gone, there's no need for X's voodoo flavour security either...
Yes, X is hard to use. How many end users do you know who have mastered xauth, for example? I know none. The usual solution is "xhost +localhost", coupled with "SSH will take care of it" for remote access.
I'm an old Amiga user. The Amiga's windowing system fit in a 256 or 512K ROM, together with pretty much all the other parts of a multitasking operating system. X in 700K? I'm not impressed.
OK, so I'm just a whining ex-Amiga user. Fair enough. Let's see what else is on the menu. QNX's Photon is, what, 40-50K? Plan 9's Rio? I don't know the size, but 8 1/2 apparently was around 100K. What about GNUStep/Display Ghostscript? I bet that if we shop around, we can do better than 700K. So let's do it.
My point isn't "X sucks"; it is that Linux, or BSD, or any free/open OS does not necessarily have to run X. And if you think about this laptop not as a standard desktop Unix platform, but as something that has to do just a few specific tasks well, it becomes an embedded system, with a whole new set of priorities. It just doesn't have to be able to run all existing applications. And that gives us (well, them...) freedom to pick whatever lets them minimize the cost of the hardware.
Sorry if this turned into a long rant, it's not meant as an attack on you or your implementation of X. I'm sure it's great for where it's meant to be used.
Of course, they filtered out Javascript and any plug-in content (for anonymity), so it wouldn't work with these services anyway. See, another reason why embedded players are bad. Just give me the download links...
But I'm sure there's some anonymizing service out there that'll work. Anyone serious about anonymity will block JS and plugins by default, though.
So, if you don't want to be disturbed by interrupts, you just don't initialize the hardware that might generate them. Or you mask them off in timing-critical sections of the code. While they may have interrupts or timers, you don't have to use them.
Oh, I see, it's a noun. Good to see that the capitalize-all-nouns has rubbed off.
I've thought about it myself, you could use something like parchive to make parity blocks and burn them to a "parity" CD/DVD. Shouldn't be too hard to script something together in Python... It won't be a direct analog to RAID, but who cares. And you wouldn't need to muck around with some big restore/untarring procedure, the original data will still be there as plain files.
Ideally, you should probably automatically generate parity blocks/discs whenever you burn a CD/DVD. Something like, say, 10-15% overhead isn't going to hurt much, at least when you're cleaning out lots of stuff.
Hmm. I should sit down and sketch out something to do this. I'm sure I'll get around to it some day. :-)
There's a couple of distros that do this; I believe they're called "NetBSD" and "FreeBSD"...
(NetBSD, at least, has as one of its stated goals to get rid of all GPL-licensed code in the operating system.)
Also similar to Internet vs. Web, you can write your own clients for Freenet (for the Freenet protocol). You're not limited to what the developers have provided for you out of the box.
0.7 will, if I understand things correctly, support almost exactly the scenario you describe (queuing downloads within the main Freenet program itself). The spidering part you'll have to write yourself though. Or...
(And parent got modded "Insightful" for what amounts to "Internet == Web"? Yikes.)
Yes, of course we use Celsius in Sweden. Anders Celsius was Swedish, after all.
Frost's site has a couple of screenshots too.
(This is all from my understanding of the network, which might not be perfect.)
d s in Iran, which might be a little too much work to set up "manually".
It is different in that you can easily connect to all your friends-of-friends-of-friends-of-friends-of-frien
Or you could make the comparison that your friends and friends-of-friends network (which is a perfectly fine thing to have) is somewhat like being able to dial local BBS'es (like in the good old days), whereas Freenet is more like having access to the Internet. In more ways than one too; there's going to be more to do than "just" exchanging files.
(Why should Opera users particularly want to switch to Firefox anyway? Is there an offer of 72 virgin brides in Heaven after you die or something that makes it worthwhile?)
Seriously, I'd say it's because Opera just gets the job done. You can pimp your FF installation all you want and it will have a ton of features not present in Opera, but for 90% of what you do, all you'll ever need is already present in the base install (of Opera). And it's not a particularly large download either. Plus it really is fast.
Whether you should migrate ... well, you should at least try it for a week, then you'll know enough to decide which was the best for you. You can always switch back.
It seems its first application was as a nonsensical statement generator.
Anyway, the Pascal-looking Amiga E (official site) is what I figured would be hiding behind the link. It was even semi-popular for some time.
Good point, but I think I disagree with your definition of "shell".
(Maybe this is what you mean by me breaking the definition of just that. So it would be my very own "command shüll". Or "khell". Whatever, you and I both know what I'm really talking about.) :-)
Seriously though, to me a "shell" is just an interface to the operating system. You have a different definition and that's fine with me.
My problem with escape codes is that they are by definition "tacked on afterwards". If you know you're going to use something, why not make it part of the specification? IMHO limiting yourself to what the original TTYs and VT100s could do is rather artificial when we all use terminal emulators that can do ANSI colors, etc, anyway.
Sorry if this is kind of unstructured, I'm really tired right now.