The NHS is meant to be Britain's largest employer. The government does have control over a reasonable proportion of the commercial market. The public sector in the UK is, if I recall correctly, something like 30% to 40% of the total. The NHS, Police, Fire, local council funded projects like leisure centres etc. It all adds up.
That said, I agree with you about everything else. The UK always struck me as having middling politics, perhaps a little to the left but not really left wing or socialist by any measure. But I suppose it depends on where you are sitting. The US looks very right wing from here.
I'd have though quantum theory would stop an infinte use of spectrum. After all, energy is discrete, and since E = hf, wouldn't that mean that f is discrete too? Please correct me if I'm utterly wrong:)
On the other hand, Iraq is a completely different situation. With WMD, Saddam has an extortion tool that he can use - cave to his demands, or we lose Cleveland. He wants money, he wants assets un-frozen, he wants us to abandon our support of Israel, and he will want us to sit on our hands when he and his friends invade their neighbors. Don't like it? Lose Miami, too.
Even if he had WMD capable of that scale of destruction, such a stunt wouldn't work. The US wouldn't cave in to his demands. No developed country would, really. And what's Saddam going to do about it? Sure, he could try nuking the US or something, but he'll only get at most one hit before the US retaliates. So it would be suicide. Saddam isn't the kind of guy to do that; look at all his doubles, his bodyguard constantly by his side. He's not a fanatic. Just a ruthless dictator desperately clinging to power.
A quick check reveals that doesn't compile as perl. And, frankly, it doesn't look like proper perl anyway. It's really irritating when people pick up obfusicated code and use that as an example from a language. Sure, you can write perl programs that are impossible to read, but you can do that with any language.
It's not the language, it's the programmer. A good coder will make beautiful code in whatever language s/he uses. Remember, There's More Than One Way To Do It.
Ok, I know you're a troll, but *gak* Java?! I've just returned from 18 hours of solid Swing coding. *Shudder*. I'd do anything to be allowed to implement this in Perl/Qt.
Well, it's true that it took some smart people to design the infrastructure of the net, but the main protocols behind it; IP, TCP and UDP, aren't difficult to understand. In fact, they're amazingly simple protocols, and hardly something I'd class as rocket science.
The internet is effectively built on very simple premises. And as the report says, it took some very smart people to design it that way.
Does any desktop have to "win"? I suspect that both GNOME and KDE will draw closer together as time passes, but that both projects will be around for quite a time.
It's a mistake to compare GNOME to VHS and KDE to BetaMax. VHS won because you couldn't play BetaMax in VHS recorders and VHS couldn't be played in BetaMax. Obviously, you don't have that problem with GNOME and KDE. Also, develops like Opera already use Qt in commercial software. The licensing fees aren't that large compared to the salary of developing an application. Another advantage is that you don't have to buy the license to use the toolkit, so you can develop an application, see if it works or is feasible, and then shell out for the license cost if everything is ok.
Well, assuming my '70%' figure was correct, which I now realise it isn't, then by information theory you've reduced the probability needed to view the information from 100% to 70%, and so reduced the information needed to store the number. But that point's moot, as I screwed up the math anyway:)
I have an version of Windows 98 on my machine, but I really can't see the point in using it. As hard as it might be for you to understand, some people actually prefer using Linux over Windows for more than idealogical reasons. I use Linux because I like it more than Windows. For me, Linux is better. In fact, whenever I boot back up into Windows, it always annoys me how limited the GUI actually is. Whilst I could use Kazaa, that would mean rebooting and languishing in Windows for several hours, which is a bigger hassle than I really want to go through.
Of course, I realise that from your tone you're either a troll or a moron, so I doubt this reply will make any difference to you either way.
How would you suggest I go about buying.hack? It isn't availiable yet outside Japan. Outlaw Star has been shown on Cartoon Network, and Trigun will be shown on Cartoon Network, since they now have the broadcasting rights. Since I'm subscribed to that channel, and will be for some time, I've already compensated the creators in question through my subscription fees.
Um, I'd say his point is that Kazaa only works on Windows. If a person does not use Windows, then they can't use Kazaa, and therefore it's not a choice for some people. How difficult is this to understand? Therefore some people are doubtless interested in Gnutella. Especially since it's gotten quite a bit better, recently.
Gtk-gnutella works quite nicely, I've found. Downloaded all of Trigun and half of Hack Sign from there so far. I even managed to find some Maaya Sakamoto MP3s. Not that I downloaded them, for that would be illegal and bad. *Ahem*. Anyway, the only Anime I couldn't find reliably was Outlaw Star, so I grabbed that from IRC.
The above argument doesn't appear to make sense now. If you can compact an arbitary sequence of numbers by 70%, then you can keep passing the same data again and again through the system and reduce any string of data to a much smaller length. By information theory, this should be impossible, shouldn't it? Anyone now where I've screwed up?
Point taken. I stand corrected, well, unless some mathematician manages to prove the total randomness of Pi. It seems to be random; considering the statistical analysis here. But I can't find a proof around at the moment.
Interestingly here's a Pi Search page I've found, so you can try out this guy's method of compression. And how did we get so off-topic again?:)
The digits of Pi, are, in essense a random stream of numbers. The chance of finding a specific single digit at any part in the stream, assuming a base system of 10, and ignoring all other considerations is 0.1. The chance of finding it after looking twice is 0.1 + (0.1 * 0.9). Three times is 0.1 + (0.1 * 0.9) + (0.1 * 0.9^2).
Therefore looking for the digit n times is the sum of the series 0.1 * (0.9)^i. Which is a geometric series and equivalent to 0.1(1 - 0.9^n) / (1 - 0.9), which equals (1 - 0.9^n). Thus, the average number of tries to find an single digit is (log 0.5/log 0.9) =~ 6.57881
What about two digits? A similar solution: (log 0.5/log 0.99) =~ 68.9676. Three digits? 692.801. Four? 6931.13.
Thus, the amount of data is approximately 70% of the original (a digit with 10 possible solutions found after an average of approximately 7 searches). So this algorithm would compress data down, on average, by 70%. Not bad, but not great either, especially considering the computing power to achieve this. There are much more efficient algorithms around today in terms of computing power vs. compression.
But Pi is irrational, so is not a repeating pattern. Ergo, all possible sequences will occur, since there are an infinite series of digits to Pi, and can't be represented as a finite function. If you think about it, you'll see that it's self-evident.
Are there any wireless networks about today that don't use the TCP/IP protocol over 802.11b?
Re:640k is more than enough for anybody
on
Slashdot over IPv6
·
· Score: 1
IPv4 has lasted for around 30 to 40 years. Is there any reason to suspect IPv6 won't last just as long, considering the obscene number of IP addresses that are possible under the protocol? It seems to me that worrying whether or not IPv17 will be backward compatible with IPv4 or IPv6 isn't something to immediately be concerned about. Maybe in another 50 to 100 years or so:)
Interestingly, in 1977, a column in Scientific American published a selection of text encrypted by, IIRC, the RSA encryption scheme with a 54bit key. At the time, the most efficient algorithm on the fastest computers in the world would take millions of times the age of Universe to crack the code. Sixteen years later, this was cracked with 8 months of computing time.
So the efficiency of algorithms to break encryption has increased by a considerable amount. In another 16 years, computers are likely to be about a thousand times faster (if Moore's law holds). But if history is anything to go by, the encryption breaking algorithms of tomorrow may reduce code-cracking by a factor of billions.
So yeh I'd say OSS can be shit, patches don't get picked up, however obviously corret they are.
And, of course, Microsoft has a far better record at fixing flaws in its software than open source projects like Apache do. </sarcasm>
Your problems with open source seem to apply just as much, if not more so, to commercial software. Except for the whole, knee-jerk 'communism' reaction. I'm guessing you're a troll, rather than merely an idiot.
In some case CGI can be seen to be a bad thing. It is, as you pointed out, notoriously insecure. However, it doesn't have to be, especially with Perl's variable tainting, and, to my mind, CGI scripts are a useful tool for users to have. Of course, I'm biased; the hosting company which serves up my webpage doesn't have mod_perl on its system, so I've done it all through CGI instead. When done correctly, Perl/CGI scripts can be very neat, even if they do give you more than enough rope to hang yourself with.
Slashdot Mirror
Name me a government owned company...
The NHS is meant to be Britain's largest employer. The government does have control over a reasonable proportion of the commercial market. The public sector in the UK is, if I recall correctly, something like 30% to 40% of the total. The NHS, Police, Fire, local council funded projects like leisure centres etc. It all adds up.
That said, I agree with you about everything else. The UK always struck me as having middling politics, perhaps a little to the left but not really left wing or socialist by any measure. But I suppose it depends on where you are sitting. The US looks very right wing from here.
I'd have though quantum theory would stop an infinte use of spectrum. After all, energy is discrete, and since E = hf, wouldn't that mean that f is discrete too? Please correct me if I'm utterly wrong :)
On the other hand, Iraq is a completely different situation. With WMD, Saddam has an extortion tool that he can use - cave to his demands, or we lose Cleveland. He wants money, he wants assets un-frozen, he wants us to abandon our support of Israel, and he will want us to sit on our hands when he and his friends invade their neighbors. Don't like it? Lose Miami, too.
Even if he had WMD capable of that scale of destruction, such a stunt wouldn't work. The US wouldn't cave in to his demands. No developed country would, really. And what's Saddam going to do about it? Sure, he could try nuking the US or something, but he'll only get at most one hit before the US retaliates. So it would be suicide. Saddam isn't the kind of guy to do that; look at all his doubles, his bodyguard constantly by his side. He's not a fanatic. Just a ruthless dictator desperately clinging to power.
Really? My mistake then. Still, the rest of my point still holds.
:)
As much as I like perl, reading all through the apocalypse after a marathon of computer science coursework seems a little much
A quick check reveals that doesn't compile as perl. And, frankly, it doesn't look like proper perl anyway. It's really irritating when people pick up obfusicated code and use that as an example from a language. Sure, you can write perl programs that are impossible to read, but you can do that with any language.
It's not the language, it's the programmer. A good coder will make beautiful code in whatever language s/he uses. Remember, There's More Than One Way To Do It.
Ok, I know you're a troll, but *gak* Java?! I've just returned from 18 hours of solid Swing coding. *Shudder*. I'd do anything to be allowed to implement this in Perl/Qt.
Select text in Mozilla, go to OpenOffice, middle click. There. That wasn't so hard, was it?
Well, it's true that it took some smart people to design the infrastructure of the net, but the main protocols behind it; IP, TCP and UDP, aren't difficult to understand. In fact, they're amazingly simple protocols, and hardly something I'd class as rocket science.
The internet is effectively built on very simple premises. And as the report says, it took some very smart people to design it that way.
Does any desktop have to "win"? I suspect that both GNOME and KDE will draw closer together as time passes, but that both projects will be around for quite a time.
It's a mistake to compare GNOME to VHS and KDE to BetaMax. VHS won because you couldn't play BetaMax in VHS recorders and VHS couldn't be played in BetaMax. Obviously, you don't have that problem with GNOME and KDE. Also, develops like Opera already use Qt in commercial software. The licensing fees aren't that large compared to the salary of developing an application. Another advantage is that you don't have to buy the license to use the toolkit, so you can develop an application, see if it works or is feasible, and then shell out for the license cost if everything is ok.
You're not meant to admit it! :)
Well, assuming my '70%' figure was correct, which I now realise it isn't, then by information theory you've reduced the probability needed to view the information from 100% to 70%, and so reduced the information needed to store the number. But that point's moot, as I screwed up the math anyway :)
:)
Ah well. It was a nice idea whilst it lasted
I have an version of Windows 98 on my machine, but I really can't see the point in using it. As hard as it might be for you to understand, some people actually prefer using Linux over Windows for more than idealogical reasons. I use Linux because I like it more than Windows. For me, Linux is better. In fact, whenever I boot back up into Windows, it always annoys me how limited the GUI actually is. Whilst I could use Kazaa, that would mean rebooting and languishing in Windows for several hours, which is a bigger hassle than I really want to go through.
Of course, I realise that from your tone you're either a troll or a moron, so I doubt this reply will make any difference to you either way.
How would you suggest I go about buying .hack? It isn't availiable yet outside Japan. Outlaw Star has been shown on Cartoon Network, and Trigun will be shown on Cartoon Network, since they now have the broadcasting rights. Since I'm subscribed to that channel, and will be for some time, I've already compensated the creators in question through my subscription fees.
Um, I'd say his point is that Kazaa only works on Windows. If a person does not use Windows, then they can't use Kazaa, and therefore it's not a choice for some people. How difficult is this to understand? Therefore some people are doubtless interested in Gnutella. Especially since it's gotten quite a bit better, recently.
Gtk-gnutella works quite nicely, I've found. Downloaded all of Trigun and half of Hack Sign from there so far. I even managed to find some Maaya Sakamoto MP3s. Not that I downloaded them, for that would be illegal and bad. *Ahem*. Anyway, the only Anime I couldn't find reliably was Outlaw Star, so I grabbed that from IRC.
The above argument doesn't appear to make sense now. If you can compact an arbitary sequence of numbers by 70%, then you can keep passing the same data again and again through the system and reduce any string of data to a much smaller length. By information theory, this should be impossible, shouldn't it? Anyone now where I've screwed up?
Make that: Assuming the digits of Pi are a stream of random numbers.
Point taken. I stand corrected, well, unless some mathematician manages to prove the total randomness of Pi. It seems to be random; considering the statistical analysis here. But I can't find a proof around at the moment.
:)
Interestingly here's a Pi Search page I've found, so you can try out this guy's method of compression. And how did we get so off-topic again?
The digits of Pi, are, in essense a random stream of numbers. The chance of finding a specific single digit at any part in the stream, assuming a base system of 10, and ignoring all other considerations is 0.1. The chance of finding it after looking twice is 0.1 + (0.1 * 0.9). Three times is 0.1 + (0.1 * 0.9) + (0.1 * 0.9^2).
Therefore looking for the digit n times is the sum of the series 0.1 * (0.9)^i. Which is a geometric series and equivalent to 0.1(1 - 0.9^n) / (1 - 0.9), which equals (1 - 0.9^n). Thus, the average number of tries to find an single digit is (log 0.5/log 0.9) =~ 6.57881
What about two digits? A similar solution: (log 0.5/log 0.99) =~ 68.9676. Three digits? 692.801. Four? 6931.13.
Thus, the amount of data is approximately 70% of the original (a digit with 10 possible solutions found after an average of approximately 7 searches). So this algorithm would compress data down, on average, by 70%. Not bad, but not great either, especially considering the computing power to achieve this. There are much more efficient algorithms around today in terms of computing power vs. compression.
But Pi is irrational, so is not a repeating pattern. Ergo, all possible sequences will occur, since there are an infinite series of digits to Pi, and can't be represented as a finite function. If you think about it, you'll see that it's self-evident.
Are there any wireless networks about today that don't use the TCP/IP protocol over 802.11b?
IPv4 has lasted for around 30 to 40 years. Is there any reason to suspect IPv6 won't last just as long, considering the obscene number of IP addresses that are possible under the protocol? It seems to me that worrying whether or not IPv17 will be backward compatible with IPv4 or IPv6 isn't something to immediately be concerned about. Maybe in another 50 to 100 years or so :)
Interestingly, in 1977, a column in Scientific American published a selection of text encrypted by, IIRC, the RSA encryption scheme with a 54bit key. At the time, the most efficient algorithm on the fastest computers in the world would take millions of times the age of Universe to crack the code. Sixteen years later, this was cracked with 8 months of computing time.
So the efficiency of algorithms to break encryption has increased by a considerable amount. In another 16 years, computers are likely to be about a thousand times faster (if Moore's law holds). But if history is anything to go by, the encryption breaking algorithms of tomorrow may reduce code-cracking by a factor of billions.
So yeh I'd say OSS can be shit, patches don't get picked up, however obviously corret they are.
And, of course, Microsoft has a far better record at fixing flaws in its software than open source projects like Apache do. </sarcasm>
Your problems with open source seem to apply just as much, if not more so, to commercial software. Except for the whole, knee-jerk 'communism' reaction. I'm guessing you're a troll, rather than merely an idiot.
I call troll!
In some case CGI can be seen to be a bad thing. It is, as you pointed out, notoriously insecure. However, it doesn't have to be, especially with Perl's variable tainting, and, to my mind, CGI scripts are a useful tool for users to have. Of course, I'm biased; the hosting company which serves up my webpage doesn't have mod_perl on its system, so I've done it all through CGI instead. When done correctly, Perl/CGI scripts can be very neat, even if they do give you more than enough rope to hang yourself with.