Doesn't matter. Even if you *could* file and obtain a patent, you'd still be screwed. The day you notice some megacorp abusing your patented technique, one of two things happen, depending on if you actually release product yourself, or if you're just a patent-troll.
If you release product yourself, then something similar to this will happen:
You send a letter, demanding that abuse of your patent stops, or that royalties are paid.
You get a response: We don't agree that what we are doing is covered by your patent. And even if we did, your patent is invalid. And even if it wasn't, did we mention that your product uses 137 techniques that are patented by us ? And even if none of this was true, do you have the time and money to figth a 10-year legal battle to prove it and extract a single cent ?
If you don't, then it's similar, the only exception being that then they cannot claim that you are using their patented techniques, so you're no more likely to actually win, but atleast you're a bit less likely to lose in the inevitable counter-claim suit.
It does. Well, it doesn't need to reboot before you can start using all your new applications, that works perfectly immediately. But one of the components updated is the kernel. And you won't be able to take advantage of the new kernel before you've made a reboot.
Ubuntu doesn't nag about it though. At the end of the upgrade-process it simply informs you that some changes will only take effect after a reboot, and put up (in the menu-bar) a button that you are free to click on (or not!) when you want to reboot.
The contrast to Win-XP is large; there I get an intrusive pop-up that announces: "I'm going to reboot if you don't stop me in the next 30 seconds... 29... 28..." and then, if I *DO* explicitly tell it "do NOT reboot", the very same dialogue pops up again half an hour later, and on and on until I give up and do as it wants. It's pretty clear who wants to be the boss...
With ubuntu, I had my son (3) play gCompris during the update, so I didn't even notice that Ubuntu recommended a reboot until after he was done playing.
Yes. Sure. All the applications where updated too. All the applications that are included in the distro (i.e several thousand) and all the ones from external sources for which newer versions are available.
Sadly, I have to report, that a brilliant new RPG that I started coding in python a few weeks ago was still at version 0.01 (where it's likely to remain, let's be honest) even after I told Ubuntu to upgrade. Very disappointing that.
Apparently, Ubuntu is only capable of automatically updating software for which someone has made a new version available. It doesn't yet manage to automatically upgrade software with *NO* human intervention needed. What a shocker.
Yes there is. And it's very very grandmother-friendly too. My procedure:
Log in, as usual.
There's a message for me, indicated by a slowly pulsing alert-icon. It reads: "There are updates available for Ubuntu, click here if you want to install them. So I do.
I'm met with the familiar update-manager, only this time it has a new button: "There is a new version of Ubuntu available, 7.1 Feisty Fawn, click here to upgrade."
I click, and am informed that this required administrative priviledge, and would I please enter my password to proceed.
I do as told, wait half an hour, and that's it.
I've never seen anything even close to this smooth. It's not just a Linux-best. It's quite simply the best I've ever seen.
Oh, and did I mention I lied above ? You see, all the messages mentioned was nicely localized into my native written Language, nynorsk, the least used variant of Norwegian, which perhaps half a million people in the world write. I'm impressed.
Its funny you should mention geeks meeting up. Is that the best reason to actively recruit women?
IMHO -- no. It's not *the* best reason. But it's decent enough reason. Atleast if you tone it down a notch by noting that gender-balanced studies and work-places tend to be more pleasant for all involved. If you've worked at IT-companies with a very heavy male-dominance, and others with a more balanced workforce, you'll notice the difference.
I don't think it's the most important reason though.
More important is the fact that females and males frequently have different ways of approaching a problem. We desperately need diversity for tackling the hard problems in IT. Someone who thinks completely out of the ordinary solution-space is very valuable, even if his (or her) suggestion is wrong or impractical 80% of the time.
I want a diverse team in general. Gender is just one axis ofcourse. I'd also welcome more variation in age and background. There's way to few 50 and 60-year olds in IT. Hopefully that's a problem that'll correct itself given time. There wasn't all that many CS-graduates 30 years ago afterall.
Ok, so I realize you're joking, but a variant of that is actually true -- and advocated by some of the most respected programmers around.
For example, SICP says that programs should be written for humans to read -- and only as a minor side-effect for machines to execute.
This means programs needs organization, style, and a correctly designed metalinguistic abstraction.
People with no idea about programming, and programming-beginners (along with some long-time-programmers who just never gets it) think that programming is primarily about knowing where to put ; where to put { }, and similar technical details that really don't matter in the larger scheme of things.
It may be easy to claim so, but that doesn't make it true. What is really amazing is that even in the cases where this is very definitely not even close to true, IE-only sites still happen.
For example, in Germany firefox-usage is currently at about 35%, up from about 25% a year ago. And that is according to the browser-id string, the real usage is somewhat higher since some people tell other browsers to lie and claim they're IE. In Europe in general Firefox-usage is at around 25%. And then there's a sprinkling of Safari, Opera, Konqi and the like. IE is still the largest browser with perhaps 70% of the market, but nowhere near 95%. (infact I very much doubt that it *ever* had 95% marketshare)
Making a website in such a way that it blocks the prefered browser of 35% (and growing!) of your visitors *SHOULD* be a severly career-limiting move. But in practice it doesn't work that way, for reasons I'm unable to grasp.
For example, the online web-tv part of the primary Germany tv-station (ARD) is made in such a way that it only works with IE. (without major trickery anyway)
You use one of the larger "reserved" netblocks. Such as 10.0.0.0/8 which has 2^24 or around 16 million available ips in it.
I'm not saying a el-cheapo D-link for $50 would actually manage to maintain NAT for a million users. For one, it doesn't have enough RAM for the NAT-table. I'm just saying, there's nothing technically in NAT that prevents a million or more users from sharing a single IP.
The NAT-table can be dealt with on a larger router. You have 1 million active users and 10 million NAT-entries that aren't expired. (i.e. less than 5 minutes old) (this would mean your users visit a new page every 30 seconds on average -- very active usage) One NAT-table entry is on the order of 16 (minimum) to 64 bytes. So the total space-requirement would be 160 - 640MB for the NAT-table alone. Which a el-cheapo D-link doesn't have. Neither does it have the performance nessecary.
There's other reasons why you'd not actually want to do this. But lack of external ports ain't one of them.
The practical limit for users doing websurfing is huge. 13 ports per user is much much more than you need, infact 1 port pr user may be more than you need.
A single TCP-connection is identified by a quad: ip and port for the two destinations.
So, you only really need a new source-port for every internal user who visits the same site.
NAT is implemented by maintaining an internal table of what external ips/ports should be mapped to which internal ip/port.
An example:
Internal machine X makes a connection from its port Y to external ip Z port W.
Nat machine Z takes note of X:Y - Z:W and uses its external port Q for this.
Later, when a packet from Z:W arrives on port Q, the internal table says to forward this to X:Y internally.
This way, a second request can use the same external port Q, aslong as the new request ain't also directed at Z:W (in which case it'll need to use a new port)
Practical result ?
You can use a single external IP for a group of websurfers, the size of the group has a limit, you run into trouble the moment more than 65000 of your internal users want to visit the same website simultaneously. With simultaneously being defined as within the timeout of the NAT-table (typically 1-5 minutes)
Atleast a million websurfers can easily hide behind a single IP using this technique. 10 million if they're not hugely active, or if they don't visit the same sites all the time. Not that there's any reason to. Ips aren't *that* hard to come by.
You could increase this by another order of magnitude or two by also taking sequence-numbers into the NAT-tables. Two different users connecting to the same service at the same time are likely to get sequence-numbers different enough that the two connections can be recognized based on this. This ain't really a good idea though, because if you did this, you could get unlucky and have two connections accidentaly get sequence-numbers close to oneanother.
Besides, you don't really have a *reason* for hiding a billion websurfers behind a single IP, now do you ?
For example, if you claimed to have copyrigth on a certain text, but the same text could demonstrably be found on Wikipedia at a point *before* you claimed to have written it, thats pretty clear evidence that you're lying.
One shouldn't, offcourse, on the other hand assume something to be true just because Wikipedia says so. But this is true for any other source too, and something courts deal with all the time.
Actually, prostitution is perfectly legal in large parts of Europe. Netherlands, sure, but also say Denmark, Germany, England, Switzerland or Norway.
In some of these its regulated, for example in Norway prostitution as such is legal -- but pimping (as in financially benefitting from the prostitution of others) is outlawed.
Yeah. But CDs are digital. Which means that if you make 3 copies, the 3 are *identical* -- not *similar* as would be the case for analogue media.
So, with a few independent copies, copied over to new media every decade or so, you can preserve digital information perfectly for as long as you care to. There is no way of preventing a analogue audio-recording from degrading over time. Making multiple copies of it, or copying over to new media as time progresses does not help.
I'd even go further. If you want society (laws, courts, police) to assist in protecting your rigths by taking advantage of copyrigth, you need to do your part to ensure that society gets maximum benefit from your work. I suggest that only software that is published with sourcecode (or atleast that has sourcecode in escrow, to be released on expiry of copyrigth) should enjoy protection.
The current "balance" is fully out of whack, the publishers insist on those rigths given to them by copyrigth while at the same time doing their darnedest by technical measures to prevent the public from enjoying the rigths *they* have by copyrigth law. What's mine is mine, and whats yours is mine too. That's not a fair deal.
Aditionally, copyrigth should be turned back to the minimum period nessecary to stimulate the creation of new works. Nobody knows what period this is, but everyone knows its much shorter than currently. Does *anybody* really believe that authors (or musicians, or programmers) exist that say, essentially: "Ok, since I get 100 years copyrigth on this book, I'll write it, if I where to get only 25, I'd not bother." ? That is completely implausible. Even more so for software and music than for books.
It can work sufficiently well that a large portion of the people with playstations, for example, are in practice unable to make a backup of their game and have that backup work.
This is a fact -- despite you being correct: there are ways to break it, and determined people can indeed manage to copy and play playstation-games.
I don't think banning DRM is needed. I would however advocate an either-or approach:
For a work to enjoy copyrigth, it should be published in an open unprotected form. If you have your own technological mechanisms for restricting what people can do with your work, then obviously you don't need copyrigth *aditionally*.
Besides, no current DRM-system has adequate (as in ANY!) mechanisms for ensuring the balances of copyrigth, such as the expiry of protection or fair use.
English spelling is bizarre anyway. The basic idea of our writing (unlikey say chinese writing) is that there's a letter for each *sound*, and so if you know how to pronounce a word, you should in principle also be able to work out how to spell it.
Of the languages I know, english is by far the one where this is least possible. Because you've kept your spelling relatively slow-changing while oral english has developed far more, I've been told.
Still, what is supposed to be the difference between "disc" and "disk" in pronounciation ? I'm guessing at nothing. Certainly you're not supposed to pronounce the c in disc in any way even approaching the c in certainly (which is again the way most people pronounce stuff pretty much indistinguishable from an 's' "summer" "certain" -- ok so some people pronounce these *sligthly* different, but no-one pronounces the second even close to the c in say "cat" (which would be written "kat" in a sane language)
I don't mean to flame or anything. English is a bundle of fun. It's just that it's also a complete mess. Sorta like perl. We foreigners should be forgiven for the occasional slip-up. This wasn't one though -- I deliberatly label round-flat things "discs", and am NOT going to start making a conscious effort to take note of compact-disc but hard-disk. That's just insanity. A disc is a disc. A hard-disc is really a whole stack of discs, in a housing, with electronics. But we all know what is meant. (except those who the entire computer is a "hard-disc")
Dunno about that. The thing is, it's the product of how many you swap with and how much you swap. Disc-sizes are growing exponentially to the point where atleast when it's about music (which it is in this article) you're going to be able to trivially storing huge libraries -- just on the hunch that you may be interested in a small fraction of it.
We already see this. There's a clear trend from sharing a single song by a single artist, then to sharing a complete album by a single artist, and then onwards to sharing complete discographies of artists, aslong as discs keep growing this trend will continue. I can easily see "every-album-that-was-in-the-charts-this-year.zip" and it's not even that much of a stretch to imagine "every-album-that-was-in-the-charts-this-decade"
For that matter, hard-disc-sizes only need to keep growing like they do for a few more years to make "every CD released in USA in the 1990ies" a completely practical thing to store and swap around.
At 200kbps (more or less the needed bandwith for indistinguishalbe-from-cd sound for most people) one hour of music takes up 90MB. An average song perhaphs 7MB. Which means that, for example, the complete content of iTunes (the store, not the program), will take up around 7MB times 3.5 million, which is about 25TB.
Today that's nontrivial, common discs today hold only half a TB or so, so you'd need 50 discs. But discs double in capacity (for the same price) about every 2 years, so that means it's about a dozen years until that entire library, 3.5 million songs, fit on a single standard consumer hard-disc. (yeah yeah, we don't know that the future will play out like that, but it seems a reasonable guess, even if it slowed down it's hard to imagine it'd take more than double that or so)
If RIAA et al think they have a hard time with simple copying now, wait and see what happens when 25TB is a trivial amount of disc-space. They are *so* fucked. I'm not saying its rigth or wrong. I'm just saying it IS so.
No, it's not possible. Because shares at 1 cent fall all the time, they even fall into negative value. Luckily for the shareholders they don't need to pay in -- instead the company is bankrupt and the shares can be used for heating in winther. That's limited liability for you.
Either a bug in the kernel (included loaded modules)
Or a malfunction or bug in one of the components the kernel needs to run (example: flaky memory)
The first are the most common; we somehow got into a state we shouldn't be in. Thus we must have messed up, and the safer choice is to refrain from further actions, since we may be insane in general. These are nevertheless bugs, and should offcourse be acknowledged as such and fixed when possible.
The second isn't really our fault. It's not really possible to write an OS in such a way that it gets the correct results even when the hardware it runs on gets it wrong. If the OS asks the CPU to calculate 2+2 and the CPU comes back with 5, there's not really much the OS can do about it.
That being said, there's cases where Linux kernel-panics in situations where it'd be possible to recover. A failed disc can panic a kernel, yet that shouldn't really be possible, except for if that disc is used as swapspace. These are bugs, or at the very least missing features.
Frankly, I don't know. There's no reason the compositer couldn't tell the program "you are now hidden", so the program will stop wasting time on un-needed redraw. The drawback to this would be if you do something that quickly exposes the program, the compositer will need to do:
Tell program "you're not hidden anymore"
Wait for program to redraw itself
Composite and display
Rather than just the 3rd alternative alone otherwise.
Normally this wouldn't matter, but in some cases it would. If you've run many large programs at once, you've probably noticed that the info needed to redraw is oftentimes swapped out if a program is hidden for a long time. So if you then swap to that program, it'll need to swap back in before it can redraw. The symptoms ? You change program, which happens quickly, and then spend seconds waiting for the program to get around to redrawing itself. I'm sure you've seen it.
If all programs draw all the time, this won't happen. On the other hand that *does* waste resources. I guess "it depends" like always.
The quick explanation (somewhat oversimplified, but you get the idea)
In a traditional (non-compositing) windowing-environment, each application essentially handles its own part of the screen, when, for example, a part of firefox previously hidden behind an xterm get unobscured, firefox is informed of this fact, and is responsible for redrawing that part of its own window.
In a compositing system, instead each program draw on their own private separate area. All these areas are then sent to the compositing manager which makes the overall screen by combining these in various ways.
There's advantages. First, it simplifies things for the programs, since they can pretend they're always alone on the screen. Secondly, it makes it possible to unify visual tricks. Without composition, for example, each and every program that wants to support stuff like being transparent, or animating their appearance, or being transparent only while being dragged or any other of a million visible tricks need to implement this independent of eachother.
With composition, the compositing window-manager can handle all of that, and the programs won't even notice. So it improves consistency by making the same visual options work identically in *all* programs.
Nonsense. Females in IT are on the average as attractive as females in pretty much any other industry, barring the few where there is a strong selection for looks. (music, showbiz, acting, modeling etc)
If you care about brains too, they're more attractive than average. And I ain't talking just knowing nerd-stuff, contrary to cliche, knowledgeable people tend to know lots of interesting stuff across the board, not just on a single narrow topic.
Goes for the males too.
The other clichees fail to match reality too. It-workers are no more likely to be single than say lawyers, doctors or car-mecanics. Among the ~20 programmers/sysadmins at my workplace, 80% or so are in a stable relationship or marriage. This is completely in line with the salesforce, managment, designers and others.
It depends on your usage. For *most* users 15/2 is functionally identical value to 6/6, because either is like an order of magnitude more than they really ever use. This goes for both my parents and all my grandparents for example. For these, anything over say 0.5 0.1 is functionally identical, so they should quite simply go for the cheapest alternative.
For most *heavy* users, downstream is significantly more important than upstream. If your main use is porn, Linux-isos, webradio, streaming TV, shareware or suchlike, then upstream doesn't much matter.
For a few, upstream is just as important. The main group is probably today those that order a lot of digital photos printed. Modern digicams, even consumer-level ones produce 1-10MB of data with each shutter-release, ordering up prints from 100 such files can easily mean uploading a GB. Which is annoying with slow ADSL.
There's a few people actually running servers from home too, but those are the nerdy minority, such as ourselves:-)
Symetrical is uncommon with the traditional (as in TelCo) providers in Norway too. There's it's mostly 1-20Mbps ADSL downstream, and about an order of magnitude less upstream.
There's a few new agile highly competitive isps though, and for those symetrical is the norm. The catch is that they're mostly only available in cities, or require multiple adjacent dwellings to sign up at once (to make installation worthwhile) or both.
Lyse offers 6/6 20/10 and 50/25 (all Mbit/s), so really only their lowest speed is symetrical. But 25 Mbit upstream rocks fairly heavily anyway, despite being trumped by the 50 downstream.
BKK are even more radical. They offer one speed only: 100Mbit full duplex symetrical. Sure, if their customers actually started using 1% of the capacity, they'd go broke or need to up prices drastically. But it's not likely Grandmother Betty will start downloading 30 TB/month anytime soon. (even using 1% of capacity would be 300GB/month, a small fraction of heavy users will actually use this much, most people won't.)
There's not that much overhead. With zero overhead you'd get 6000/8 = 750KBps, in actual practice I tend to get about 700. So, ok 100MB/700K = 142s.
True. 2 minutes something, not 1.5. Big deal.
Slashdot Mirror
If you release product yourself, then something similar to this will happen:
If you don't, then it's similar, the only exception being that then they cannot claim that you are using their patented techniques, so you're no more likely to actually win, but atleast you're a bit less likely to lose in the inevitable counter-claim suit.
Ubuntu doesn't nag about it though. At the end of the upgrade-process it simply informs you that some changes will only take effect after a reboot, and put up (in the menu-bar) a button that you are free to click on (or not!) when you want to reboot.
The contrast to Win-XP is large; there I get an intrusive pop-up that announces: "I'm going to reboot if you don't stop me in the next 30 seconds... 29 ... 28..." and then, if I *DO* explicitly tell it "do NOT reboot", the very same dialogue pops up again half an hour later, and on and on until I give up and do as it wants. It's pretty clear who wants to be the boss...
With ubuntu, I had my son (3) play gCompris during the update, so I didn't even notice that Ubuntu recommended a reboot until after he was done playing.
Sadly, I have to report, that a brilliant new RPG that I started coding in python a few weeks ago was still at version 0.01 (where it's likely to remain, let's be honest) even after I told Ubuntu to upgrade. Very disappointing that.
Apparently, Ubuntu is only capable of automatically updating software for which someone has made a new version available. It doesn't yet manage to automatically upgrade software with *NO* human intervention needed. What a shocker.
I've never seen anything even close to this smooth. It's not just a Linux-best. It's quite simply the best I've ever seen.
Oh, and did I mention I lied above ? You see, all the messages mentioned was nicely localized into my native written Language, nynorsk, the least used variant of Norwegian, which perhaps half a million people in the world write. I'm impressed.
IMHO -- no. It's not *the* best reason. But it's decent enough reason. Atleast if you tone it down a notch by noting that gender-balanced studies and work-places tend to be more pleasant for all involved. If you've worked at IT-companies with a very heavy male-dominance, and others with a more balanced workforce, you'll notice the difference.
I don't think it's the most important reason though.
More important is the fact that females and males frequently have different ways of approaching a problem. We desperately need diversity for tackling the hard problems in IT. Someone who thinks completely out of the ordinary solution-space is very valuable, even if his (or her) suggestion is wrong or impractical 80% of the time.
I want a diverse team in general. Gender is just one axis ofcourse. I'd also welcome more variation in age and background. There's way to few 50 and 60-year olds in IT. Hopefully that's a problem that'll correct itself given time. There wasn't all that many CS-graduates 30 years ago afterall.
For example, SICP says that programs should be written for humans to read -- and only as a minor side-effect for machines to execute.
This means programs needs organization, style, and a correctly designed metalinguistic abstraction.
People with no idea about programming, and programming-beginners (along with some long-time-programmers who just never gets it) think that programming is primarily about knowing where to put ; where to put { }, and similar technical details that really don't matter in the larger scheme of things.
It may be easy to claim so, but that doesn't make it true. What is really amazing is that even in the cases where this is very definitely not even close to true, IE-only sites still happen.
For example, in Germany firefox-usage is currently at about 35%, up from about 25% a year ago. And that is according to the browser-id string, the real usage is somewhat higher since some people tell other browsers to lie and claim they're IE. In Europe in general Firefox-usage is at around 25%. And then there's a sprinkling of Safari, Opera, Konqi and the like. IE is still the largest browser with perhaps 70% of the market, but nowhere near 95%. (infact I very much doubt that it *ever* had 95% marketshare)
Making a website in such a way that it blocks the prefered browser of 35% (and growing!) of your visitors *SHOULD* be a severly career-limiting move. But in practice it doesn't work that way, for reasons I'm unable to grasp.
For example, the online web-tv part of the primary Germany tv-station (ARD) is made in such a way that it only works with IE. (without major trickery anyway)
I'm not saying a el-cheapo D-link for $50 would actually manage to maintain NAT for a million users. For one, it doesn't have enough RAM for the NAT-table. I'm just saying, there's nothing technically in NAT that prevents a million or more users from sharing a single IP.
The NAT-table can be dealt with on a larger router. You have 1 million active users and 10 million NAT-entries that aren't expired. (i.e. less than 5 minutes old) (this would mean your users visit a new page every 30 seconds on average -- very active usage) One NAT-table entry is on the order of 16 (minimum) to 64 bytes. So the total space-requirement would be 160 - 640MB for the NAT-table alone. Which a el-cheapo D-link doesn't have. Neither does it have the performance nessecary.
There's other reasons why you'd not actually want to do this. But lack of external ports ain't one of them.
A single TCP-connection is identified by a quad: ip and port for the two destinations.
So, you only really need a new source-port for every internal user who visits the same site.
NAT is implemented by maintaining an internal table of what external ips/ports should be mapped to which internal ip/port. An example:
Practical result ?
You can use a single external IP for a group of websurfers, the size of the group has a limit, you run into trouble the moment more than 65000 of your internal users want to visit the same website simultaneously. With simultaneously being defined as within the timeout of the NAT-table (typically 1-5 minutes)
Atleast a million websurfers can easily hide behind a single IP using this technique. 10 million if they're not hugely active, or if they don't visit the same sites all the time. Not that there's any reason to. Ips aren't *that* hard to come by.
You could increase this by another order of magnitude or two by also taking sequence-numbers into the NAT-tables. Two different users connecting to the same service at the same time are likely to get sequence-numbers different enough that the two connections can be recognized based on this. This ain't really a good idea though, because if you did this, you could get unlucky and have two connections accidentaly get sequence-numbers close to oneanother.
Besides, you don't really have a *reason* for hiding a billion websurfers behind a single IP, now do you ?
For example, if you claimed to have copyrigth on a certain text, but the same text could demonstrably be found on Wikipedia at a point *before* you claimed to have written it, thats pretty clear evidence that you're lying.
One shouldn't, offcourse, on the other hand assume something to be true just because Wikipedia says so. But this is true for any other source too, and something courts deal with all the time.
In some of these its regulated, for example in Norway prostitution as such is legal -- but pimping (as in financially benefitting from the prostitution of others) is outlawed.
So, with a few independent copies, copied over to new media every decade or so, you can preserve digital information perfectly for as long as you care to. There is no way of preventing a analogue audio-recording from degrading over time. Making multiple copies of it, or copying over to new media as time progresses does not help.
I'd even go further. If you want society (laws, courts, police) to assist in protecting your rigths by taking advantage of copyrigth, you need to do your part to ensure that society gets maximum benefit from your work. I suggest that only software that is published with sourcecode (or atleast that has sourcecode in escrow, to be released on expiry of copyrigth) should enjoy protection.
The current "balance" is fully out of whack, the publishers insist on those rigths given to them by copyrigth while at the same time doing their darnedest by technical measures to prevent the public from enjoying the rigths *they* have by copyrigth law. What's mine is mine, and whats yours is mine too. That's not a fair deal.
Aditionally, copyrigth should be turned back to the minimum period nessecary to stimulate the creation of new works. Nobody knows what period this is, but everyone knows its much shorter than currently. Does *anybody* really believe that authors (or musicians, or programmers) exist that say, essentially: "Ok, since I get 100 years copyrigth on this book, I'll write it, if I where to get only 25, I'd not bother." ? That is completely implausible. Even more so for software and music than for books.
It can work sufficiently well that a large portion of the people with playstations, for example, are in practice unable to make a backup of their game and have that backup work.
This is a fact -- despite you being correct: there are ways to break it, and determined people can indeed manage to copy and play playstation-games.
I don't think banning DRM is needed. I would however advocate an either-or approach:
For a work to enjoy copyrigth, it should be published in an open unprotected form. If you have your own technological mechanisms for restricting what people can do with your work, then obviously you don't need copyrigth *aditionally*.
Besides, no current DRM-system has adequate (as in ANY!) mechanisms for ensuring the balances of copyrigth, such as the expiry of protection or fair use.
Of the languages I know, english is by far the one where this is least possible. Because you've kept your spelling relatively slow-changing while oral english has developed far more, I've been told.
Still, what is supposed to be the difference between "disc" and "disk" in pronounciation ? I'm guessing at nothing. Certainly you're not supposed to pronounce the c in disc in any way even approaching the c in certainly (which is again the way most people pronounce stuff pretty much indistinguishable from an 's' "summer" "certain" -- ok so some people pronounce these *sligthly* different, but no-one pronounces the second even close to the c in say "cat" (which would be written "kat" in a sane language)
I don't mean to flame or anything. English is a bundle of fun. It's just that it's also a complete mess. Sorta like perl. We foreigners should be forgiven for the occasional slip-up. This wasn't one though -- I deliberatly label round-flat things "discs", and am NOT going to start making a conscious effort to take note of compact-disc but hard-disk. That's just insanity. A disc is a disc. A hard-disc is really a whole stack of discs, in a housing, with electronics. But we all know what is meant. (except those who the entire computer is a "hard-disc")
We already see this. There's a clear trend from sharing a single song by a single artist, then to sharing a complete album by a single artist, and then onwards to sharing complete discographies of artists, aslong as discs keep growing this trend will continue. I can easily see "every-album-that-was-in-the-charts-this-year.zip" and it's not even that much of a stretch to imagine "every-album-that-was-in-the-charts-this-decade"
At 200kbps (more or less the needed bandwith for indistinguishalbe-from-cd sound for most people) one hour of music takes up 90MB. An average song perhaphs 7MB. Which means that, for example, the complete content of iTunes (the store, not the program), will take up around 7MB times 3.5 million, which is about 25TB.
Today that's nontrivial, common discs today hold only half a TB or so, so you'd need 50 discs. But discs double in capacity (for the same price) about every 2 years, so that means it's about a dozen years until that entire library, 3.5 million songs, fit on a single standard consumer hard-disc. (yeah yeah, we don't know that the future will play out like that, but it seems a reasonable guess, even if it slowed down it's hard to imagine it'd take more than double that or so)
If RIAA et al think they have a hard time with simple copying now, wait and see what happens when 25TB is a trivial amount of disc-space. They are *so* fucked. I'm not saying its rigth or wrong. I'm just saying it IS so.
No, it's not possible. Because shares at 1 cent fall all the time, they even fall into negative value. Luckily for the shareholders they don't need to pay in -- instead the company is bankrupt and the shares can be used for heating in winther. That's limited liability for you.
Either a bug in the kernel (included loaded modules)
Or a malfunction or bug in one of the components the kernel needs to run (example: flaky memory)
The first are the most common; we somehow got into a state we shouldn't be in. Thus we must have messed up, and the safer choice is to refrain from further actions, since we may be insane in general. These are nevertheless bugs, and should offcourse be acknowledged as such and fixed when possible.
The second isn't really our fault. It's not really possible to write an OS in such a way that it gets the correct results even when the hardware it runs on gets it wrong. If the OS asks the CPU to calculate 2+2 and the CPU comes back with 5, there's not really much the OS can do about it.
That being said, there's cases where Linux kernel-panics in situations where it'd be possible to recover. A failed disc can panic a kernel, yet that shouldn't really be possible, except for if that disc is used as swapspace. These are bugs, or at the very least missing features.
- Tell program "you're not hidden anymore"
- Wait for program to redraw itself
- Composite and display
Rather than just the 3rd alternative alone otherwise.Normally this wouldn't matter, but in some cases it would. If you've run many large programs at once, you've probably noticed that the info needed to redraw is oftentimes swapped out if a program is hidden for a long time. So if you then swap to that program, it'll need to swap back in before it can redraw. The symptoms ? You change program, which happens quickly, and then spend seconds waiting for the program to get around to redrawing itself. I'm sure you've seen it.
If all programs draw all the time, this won't happen. On the other hand that *does* waste resources. I guess "it depends" like always.
In a traditional (non-compositing) windowing-environment, each application essentially handles its own part of the screen, when, for example, a part of firefox previously hidden behind an xterm get unobscured, firefox is informed of this fact, and is responsible for redrawing that part of its own window.
In a compositing system, instead each program draw on their own private separate area. All these areas are then sent to the compositing manager which makes the overall screen by combining these in various ways.
There's advantages. First, it simplifies things for the programs, since they can pretend they're always alone on the screen. Secondly, it makes it possible to unify visual tricks. Without composition, for example, each and every program that wants to support stuff like being transparent, or animating their appearance, or being transparent only while being dragged or any other of a million visible tricks need to implement this independent of eachother.
With composition, the compositing window-manager can handle all of that, and the programs won't even notice. So it improves consistency by making the same visual options work identically in *all* programs.
If you care about brains too, they're more attractive than average. And I ain't talking just knowing nerd-stuff, contrary to cliche, knowledgeable people tend to know lots of interesting stuff across the board, not just on a single narrow topic. Goes for the males too.
The other clichees fail to match reality too. It-workers are no more likely to be single than say lawyers, doctors or car-mecanics. Among the ~20 programmers/sysadmins at my workplace, 80% or so are in a stable relationship or marriage. This is completely in line with the salesforce, managment, designers and others.
For most *heavy* users, downstream is significantly more important than upstream. If your main use is porn, Linux-isos, webradio, streaming TV, shareware or suchlike, then upstream doesn't much matter.
For a few, upstream is just as important. The main group is probably today those that order a lot of digital photos printed. Modern digicams, even consumer-level ones produce 1-10MB of data with each shutter-release, ordering up prints from 100 such files can easily mean uploading a GB. Which is annoying with slow ADSL.
There's a few people actually running servers from home too, but those are the nerdy minority, such as ourselves :-)
There's a few new agile highly competitive isps though, and for those symetrical is the norm. The catch is that they're mostly only available in cities, or require multiple adjacent dwellings to sign up at once (to make installation worthwhile) or both.
Lyse offers 6/6 20/10 and 50/25 (all Mbit/s), so really only their lowest speed is symetrical. But 25 Mbit upstream rocks fairly heavily anyway, despite being trumped by the 50 downstream.
BKK are even more radical. They offer one speed only: 100Mbit full duplex symetrical. Sure, if their customers actually started using 1% of the capacity, they'd go broke or need to up prices drastically. But it's not likely Grandmother Betty will start downloading 30 TB/month anytime soon. (even using 1% of capacity would be 300GB/month, a small fraction of heavy users will actually use this much, most people won't.)
There's not that much overhead. With zero overhead you'd get 6000/8 = 750KBps, in actual practice I tend to get about 700. So, ok 100MB/700K = 142s. True. 2 minutes something, not 1.5. Big deal.