I think the difference doesn't actually look good for Microsoft really. Yes they say
"we're here and responsible for our stuff"
but phrased a little differently, what they're really saying is that in all the world there's only one company that has sufficient faith in Microsoft OS software that they're willing to be responsible for it (and if you read the EULA they're not responsible anyway). In contrast Linux has many companies who are all sufficiently confident in Linux that they're willing to stand up and actually take responsibility for it. Why are they so confident? Because they know that even if a problem is found they can fix it themselves and provide that fix to their customers.
Personally I'd be more willing to trust the system that has lots of companies wanting to step up and offer to be responsible. If I wanted accountability I'd pay one those comanies to be responsible for any issues, rather than Microsoft, standing alone, claiming they are responsible "sort of, in a way, maybe".
Fact: Much of what winders suffers from is incompetent users. Nothing is really stopping the developers from writing spam bots for windows because idiot users on Linux could run bad code just as easily as idiot users on windows.
For now, yes, but as SELinux, or RSBAC, or any of the Mandatory Access Control, role based systems gain popularity in mainstream Linux (and SELinux, for now, seems to be the best candidate on the popularity front), the ability for idiot users to run bad code goes down massively.
Yes, in theory an idiot user could run bad code, but under a well implemented SELinux policy, while the code may run, it wouldn't actually have rights to do much of anything. At worst it might be able to fill up the home partition with useless data, or something along those lines, but spam bots and zombies and mass mailing viruses would be a far more difficult task to write indeed. A sufficiently smart idiot could grant the process the rights to do what it wants, but really...
Yes, such a system is not a cure all. People can still do bad things to themselves, and no matter how well you build it, there's always an idiot who can break it. It does, however, significantly raise the security bar on what it is easy to trick a user into doing.
Well duh. But that is hardly time developing a better villain, and actually Palpatine doesn't get that much either. That's a pretty sorry excuse as far as the lack of credible interesting villains goes. You may as well say "But Darth Vader (as Annakin) gets lots and lots of screen time!" Yes, he does, but it hardly makes up for the lack of a villain of any note.
Dooku is evil, and he is more than likely the current apprentice to sidious. Anakin takes his place after Dooku is disposed of.
Which, I expect, will occur in the very first act of the movie. Part of the problem of Ep I and II (and from the look of "General Grievous" Ep III as well) has been Lucas' "rent a villain" approach - we get the villain of the week for each new episode, and that means much less time actually constructing credible villains. Honestly, compare the screen time either villain has gone in either of the current episodes to the screen time Vader gets in any of IV, V or VI. Poorly developed completely disposable villains don't tend to make for very interesting stories, especially when new evil villains crop up out of nowhere all the time (they just get mentioned in the scroller at the beginning as the new villain - no other explanation required).
Yes, yes, Darth Sidious is consistent through all 3, but he hardly has any screen time, and isn't exactly a menacing villain that our heroes must battle with, given that none of them have even seen him yet.
Slashdot has preferences, and they are helpful. I can't be bothered reading everything from people with high Karma (because there's a lot of them these days), but some people will. If they don't want to read this they can:
(1) Set their threshold to +3 (2) Make me a foe, and give foes a -1 (3) Set the Karma posting bonus to +0
If they're anonymous cowards... they get whatever they're fed, and that's not my problem.
The problem with LaTeX is that it's impossible to Google for document classes or other documents about it, because of the porn that comes up. Vanilla TeX has the same problem, except that you get derogatory pages about the President, instead.
You could start by actually looking in the obvious places first, and save yourself the trouble. There are a suprising amount of prepared packages and documentation for LaTeX available, you just actually have to look for it.
Jedidiah.
Re:Please hack open office's SIZE
on
Hacking OpenOffice
·
· Score: 3, Interesting
I had to remove OOo from my home box last night. I needed the disk space back. Why does a office suite on a Linux box have to take up a gig of disk?!
I've had similar issues. It's the i18n module that is most annoying to me, in that it is listed as a dependency (and so gets downloaded and installed) but isn't really required in a lot of cases. Yes the i18n support is a great thing... but could the packagers at the distros make it a little more optional?
With regard to your sig: I agree, it is an issue. I set my preferences to remove karma posting bonuses, which helps a bit, but it still doesn't really cover it.
I suggest a more useful scheme would be to have an uncapped logarithmic scale for score based on mods. That is, you need 1 mod point to get to +1, 2 more for +2, 4 more for +3, etc. or some similar scheme.
Odds that Slashdot implements such a thing: very very low.
It's okay though because the lawyers are suing MGM for a little over a million dollars to cover legal fees. I'm glad that the big money is going where it's supposed to. I mean, the important thing here is that the lawyers get paid right? My understanding was that that was the point of US society these days: to pay Lawyers.
Google could also roll out a thin client service in which you do everything within any browser window connected to Google. Google could host user accounts that go beyond email and search. A person could browse through the google browser, manage their googlefiles, run googleoffice, send gmail, buy stuff through froogle, etc. It would be a totally portable thin client service.
Which all sounds fantastic, but they could do all of that with either XUL, or XAML (if it ever shows up). In the end they may simply be getting in some people to do some XUL applications. If they want to include a browser in that suite of applications then rebranding Firefox seems the obvious way to go. As for having an office suite - that one is a surpringly large amount of work.
In general the concept of Google using, say, XUL to create a suite of basic applications (a browser, a mail reader, a chat client, some calendaring and groupware etc. makes some sense. I wonder how much of that would simply be a matter of rebranding all the mozilla sub-projects?
It's possible, but I still rate it all as highly unlikely.
I've done some cryptology too - enough to know that your analogy is rather unfair. Here's a better one:
You have a new unknown crypto algorithm to analyse. You know some basic facts about the structures of the algorithm (stream or block, etc.) and a fairly large sample of encrypted text, but nothing else. Based on what you know about the algorithm you try to make logical guesses about roughly how the whole should work, and construct a broad model algorithm. The goal now is not to match bit by bit the output of your sample ciphertext (you'd need the keys and IV for that anyway) but simply to try and make qualitive predictions about the characteristics of the resulting cipher. For instance you might, based on your model, say that you would expect a higher than normal occurrence of the bit sequence 11010. If when you go through your sample data you find that to be true you would probably suspect you may be on the right track in the reconstruction of the algorithm.
That's basically what the Climatologists are doing. They aren't predicting what the next bit will be (the temperature tomorrow), but they are trying to predict what broad patterns will be apparent in the system. They are making the model not just from prior data, but from knowledge and understanding of how the atmosphere and the climate ought to work based on what we do know about air masses, heating, cooling, and all the associated dynamics (equivalent to knowing a little about the crypto algorithm, and having to fill in the details to make your model).
The problem here is that you're answering the wrong question.
Mostly I was trying to correct the glaring misuse of "Chaos Theory" which, being a mathematician by trade, I found to be quite irritating.
The question is, can climatological models predict weather ten years in advance?
Indeed, that is the important question.
My answer is "I don't know. Let's get the best climatologists together and sponsor a year of research. At the end of that, they give us a ten-year projection. We keep on going. Ten years later, if their projection is accurate to within ten percent, we say their model has merit and go on from there."
You seem to presume we haven't been doing that for the last 10 years. Okay, the model at the level of detail used in this compute hasn't been tried before, but it is, I presume, simply a refinement of previous climatic models which have been in use for the last 20 years or so. Personally I don't know how well climatic modelling has gone for the last 10 years, nor how accurate any predictions were. I expect it is all in the literature from 10 years ago though, and you can probably look it up.
I'm pretty neutral on this debate for now, but I think the dismissals of "we can't predict the climate" are awfully premature. If they showed that predictions of global warming from 10 years ago have proved to be complete crap... well, then I'd start listening. I've never heard that particular argument trotted out though. Instead I've mostly seen a lot of climatologists who seem pretty confident that their models (up to a reasonable margin of error) are reasonably good. Apparently this particular model was used to predict the weather for the last 50 years based on data from 50 years ago and did a decent job. That's no guarantee that the model works, but its not a bad pointer that maybe they're onto something.
Does anybody remember how Chaos Theory was first postulated?
Yes, I seem to recall reading about Henri Poincare finding a "homoclinic tangle" while trying to solve the problem of the stability of the Solar System (to win a prize put up by the King of Sweden). It amounted to a strange attractor, and a chaotic system. That probably wasn't "the first" being only around 1890, but it was one of the earlier points. Why what did you have in mind?
The crux of Chaos Theory is that some systems will NEVER be predictable because there are so many variables that it is impossible to know all the starting conditions.
Not really. Chaos Theory generally has more to say about what you can predict/say about such systems, and the fact that your predictions will have to be formed differently than those of nice classical linear systems.
Or were you talking about "Popular Chaos Theory" where people who don't know what they're talking about make vague generalisations about what they think "Chaos Theory" probably means, largely based on a few half assed descriptions from MIchael Crichton books and Hollywood films?
If a computer model can't even predict weather more than a few days out, how is it that it can predict weather a hundred years from now?
Really? I can make quite a few fairly accurate predictions about the weather over the coming year: It will be (in the northen hemisphere) warmer over June July and August, but will cold come the end of the year. On average Florida will be warmer than Maine this year. Seattle will see a lot of days with rain this year.
You see, despite it being a chaotic system, it's still possible to discuss some of the more qualitive aspects with some accuracy. I can't predict exactly what the weather will be like on July 23rd, but I can make a fairly accurate guess that it will be warmer than the weather tomorrow (unless you're in the southern hemisphere). They can't tell you exactly what the weather will be like 100 years from now, but they can make qualitive broad statements about it.
Chaos Theory has to be the single most misunderstood and misrepresented theory next to Quantum Physics. Could you please refrain from further spreading this bizarre contaminated view of what is, actually, an interesting field of mathematics.
bin Laden does have a stated intent to kill Americans yes. The fundamental goal they are striving for, however, is the for muslim states of the middle east to become Islamic states. As to killing Americans... yes, he will try, and is limited only by his capability. I am not questioning intent, I am questioning capability.
Aum Shinrikyo was a japanese cult that had billions at their disposal, and were interested in making chemical and nuclear weapons. With all that money, and recruiting intelligent young grad students from major Japanese universities they managed a single sarin gas attack in 1995 killing 12 people.
al Qaeda has used coventional explosives in all their attacks, and have, aside from 9/11, failed to show anything resembling global reach. In fact there is much evidence that al Qaeda is more of a venture capital firm for anyone wanting to attack Americans, and don't have any extant network at all - and never did.
Which raises the point that really the issue is constraining capability, and capability mostly takes the form of money. In theory the US has a vast and powerful foreign intelligence agency (NSA) that is supposed to be good at tracing and shutting down money flowing into terrorist causes. That, it would seem to me, would be the most effective way to fight the war on terror: quietly and efficiently in the background, not drawing attention and inspiring other random terrorist groups to act in sympathy. I'm sure that's probably happening. All the rest - the tromping around of military, the random security measures applied piecemeal to random points of infrastructure in the US, the arrests of terrorist cells (usually innocent) - that's all for show. To be honest, it's probably more counterproductive than anything.
Sad to tell you this, but if Iraq gets a taste of democracy and it catches on in the middle east
Yes, that's one possible outcome for Iraq. Another possible outcome is that out of all the chaos Iraq manages to form itself into an Islamic state - what Zawahiri and bin Laden have been trying (and repeatedly failing) to do for the last 15 years or so. Who knows, Zawahiri and bin Laden believe that, sould that actually happen it will cause the muslim masses to rise up, overthrow their leaders and create a slew of Islamic states throughout the middle east. That was, is, and will be their goal. For the most part the state "jihad against America" is a way to try and rally support - a lesson they learned when their attempted efforts in, for instance, Algeria failed to attract the support of the masses (oddly the general population was rather repelled, rather than attracted by, their violence).
So, we have 2 competing theories:
(1) Install a democracy in the Iraq and watch democracy then sweep the middle east.
(2) Rally support by encouraging people to rise up against the Americans that interfere in middle east politics and institute an Islamic state in Iraq. The Islamic Jihad movement can then sweep the middle east.
To be honest, no matter what happens in Iraq, I don't really expect anything to "sweep the middle east". In the meantime though the two theories seem to be fairly well in balance. Iraq is in chaos, there's ill will by the common people toward the US, and Islamic clerics (like al Sadr) are polling very well leading up the elections. In the meantime Iraq is actually having free and open elections so democracy will arrive. It looks to me if things could go either way - which means I'm not so sure this whole "introduce democracy and watch it spread through the middle east" idea was quite all it was cracked up to be.
Ya, because the distruction of the twin towers on 911 was just a hoax.
Sure, it happened. What we don't really have is much evidence that it was anything more than a freak occurence. How many terrorist attacks have occured on US soil? How many attacks in the US have been foiled? Have you actually looked at any of those cases of "terrorist cells" in the US? They pretty much all either got quietly dismissed, or otherwise shuffled down and effectively dropped.
Exactly how big an organisation is al Qaeda really? I mean, besides what the administration has told you about a globe spanning terrorist network with amazing resources, what do you actually know about al Qaeda? Try digging around a wee bit and read material from people that were following Zawahiri and bin Laden from prior to all this. This article might be a good place to start.
If you look at the statistics, global terrorism has been in decline since the mid to late 80's. Aside from the occasional anomaly (9/11), the only growth area of terrorism in the last 10 years has been... Iraq in the last 2 years.
Countries in the Indian Ocean just suffered from an earthquake and tsunami that literally killed 100 times as many people as 9/11. Most people seem willing to accept that that was just a freak occurence. Sometimes bad things happen. It is possible that 9/11 was just a freak occurence, and in no way indicative of a vast and powerful global terrorist network. How can we know? The evidence for the vast network is surprisingly lacking if you actually look at it.
Granted. Sorry to springboard from your post to make a different point.
The "system" only requires such because "investors" (which is to say all the slashdot geeks busy making day trades) only care about short term quarterly profits rather than true long term earning potential. Thus getting profits yet cutting staff (boosting quarterly earnings, though potentially damaging earnings down the line, be it from having to rehire on to refill those positions more expensively later, or simply garnering ill will from prospective employees).
The "system" being at fault is not so much the corporations, but the internet day trading weenies.
The reality, however, is that while hard decisions like this might be very good for the investors in the short term (and hence help boost investment) because of the boost in quarterly profits, in the long term EA is aquiring quite the reputation as a slave driver with no loyalty to people it employs. If EA develops too much of a reputation for that they won't get anywhere near the same employee pool to pick and choose from - the smart people will be staying away. Long term it is potentially gutting the company if they push it too far.
And that, right there, is the big problem that causes so many people to complain about big corporations: They have come to favour short term quarterly profits over long term sustainable profits. If you look at most complaints, from environmental, to labour, to political, when you pare it down it is occuring because companies are considering their short term future but not bothering to look at the long term results.
I seem to recall some Debian developers doing a computation to find the "centre of mass" of Debian developers - the closest you could get to a "centre" for open source. If I remember rightly it was somewhere north of Iceland. You can go there if you want.
He does work for a Finnish company, so that explains alot. Did you ever notice that there is no body more british than someone who USED to live in Britain?
So you're saying that someone with an obviously (to anyone who knows any Finnish) Finnish name Mikko Hyppönnen, who works for a Finnish company {is British/used to live in Britain} because he says "aeroplane"?
I mean, I know it was a troll, but try to do it with a little more class in future.
Imagine the effect that it would have if Firefox forked, its current maintainers left the project b/c all the devs went to the forked project. You would have a bunch of people still using Firefox that would never switch to the new one (hell it took them long enough to trust something without the little "e" already). Firefox would go to shit because no one would maintain it. The fork would grow in popularity among the educated. Once Firefox broke the people that switched would slowly migrate back to IE.
You would think that that's how it would happen, but that's the point - forks often don't pan out like you would expect. ECGS forked from GCC and eventually the GCC people accepted ECGS as the new GCC. XEmacs forked from Emacs, but Emacs development continued, and these days it is Emacs that has the more vibrant and creative dev team (AFAIAC - I used to be an XEmacs fan, but comparatively it has stagnated). Often enough things haven't gone as expected and people sticking with the "original line" get the benefits (eventually), whether it be from the fork being folded back in (GCC) or from the spurred competition arising from the fork (Emacs). Don't expect to be able to predict what will happen when a fork occurs. It's a lot more complicated than you might imagine.
One of the problems I see with high levels of security without a superuser-style account is the possibility of someone leaving, dying, or forgetting his password, and not being able to get to critical business data.
In SELinux, which I am more familiar with, and which also gets rid of the "superuser" account, everything is handled by context or role. That means you can isolate a process that wants "root" access to certain files by restricting its role to one that has access to only those files. Thus there is no "root" account that has access to everything. At the same time, it's possible to create a role that allows suitable access to make changes and/or recover lost data if necessary.
I presume Coyotos, with its "capabilities" will work similarly - ie. there is no "root" account that has access to everything, but instead various capabilities that bound access to various resources.
The Emacs/XEmacs fork is given passing mention in the article, but is actually one of the more interesting ones. At the time XEmacs really did represent a step forward, mostly in its embrace of an X based GUI using modern toolkits. Consequently XEmacs tended to romp along and be the feature leader. Most recently, however, the situation has reversed. It is now XEmacs that is unwilling to use modern toolkits, and GNU Emacs is starting to push back.
Let's be frank, these days when we say "modern GUI toolkit" for X we mean wither GTK or QT. XEmacs does have GTK support, but the developers are not interested in it, and mostly it is just slow, and bug ridden, even in CVS. Compare that to Emacs, which has finally decided that GTK might not be such a bad idea. The current CVS versions of Emacs have excellent GTK support, making full use of the latest versions of GTK. It looks and behaves very nicely indeed, and integrates quite well into a GNOME desktop. The new GNU Emacs will also sport excellent Unicode support. It will be interesting to see how the GNU Emacs/XEmacs debate stands once XEmacs 22 and Emacs 22 come out. I expect to see GNU Emacs get a real boost in popularity.
Well, I'm not using Gentoo, but it's not just %10. I get _huge_ performance gains on my 3.2GHz Northwood if I use proper compiler optimizations. I'm using this workstation for audio related tasks. Ardour (Digital Audio Workstation software) and several plugin effects / audio encoders are 30-40% faster compared to precompiled i386 binaries.
See, for my money I much prefer to let the distribution prepackage binaries for the 98% of stuff that makes up the base system and doesn't really require speed gains (and won't really get any) because that's just easier. Then you can compile your own versions of whatever few applications and libraries you're using that desperately need that extra performance gain. So, for example, in your case I'd just use any old distro and then compile my own version of Ardour, and more importantly compile all the plugins as well, using every trick I can to squeeze the extra performance.
Really, do you get 40% speed gains out of KDE and GNOME? No, not really. So why not let them be prepackaged. How about Python, perl, libXML, Firefox, gettext and all those other bits and pieces? Again, for all it matters they may as well be prepackaged.
The truth is, for speed gains from compiler flags it varies from application to application. For some -ffast-math is a gain, for others not. Likewise with every other flag. Trying to set a bunch of flags for everything is as likely to provide inefficiencies in various bits as gains. Besides, for the majority of applications the speed just isn't that important. KDE spends most of its time idle waiting for input..
The other problem with switching over to Dvorak is most common keyboard shortcuts aren't so convenient anymore.
Just use Emacs. Sure some keyboard shortcuts will be harder to reach, but some will be easier to reach. By using every possible key combination for something Emacs ensures that you're no worse off no matter what keyboard layout you choose.
Slashdot Mirror
I think the difference doesn't actually look good for Microsoft really. Yes they say
"we're here and responsible for our stuff"
but phrased a little differently, what they're really saying is that in all the world there's only one company that has sufficient faith in Microsoft OS software that they're willing to be responsible for it (and if you read the EULA they're not responsible anyway). In contrast Linux has many companies who are all sufficiently confident in Linux that they're willing to stand up and actually take responsibility for it. Why are they so confident? Because they know that even if a problem is found they can fix it themselves and provide that fix to their customers.
Personally I'd be more willing to trust the system that has lots of companies wanting to step up and offer to be responsible. If I wanted accountability I'd pay one those comanies to be responsible for any issues, rather than Microsoft, standing alone, claiming they are responsible "sort of, in a way, maybe".
Jedidiah.
Fact: Much of what winders suffers from is incompetent users. Nothing is really stopping the developers from writing spam bots for windows because idiot users on Linux could run bad code just as easily as idiot users on windows.
For now, yes, but as SELinux, or RSBAC, or any of the Mandatory Access Control, role based systems gain popularity in mainstream Linux (and SELinux, for now, seems to be the best candidate on the popularity front), the ability for idiot users to run bad code goes down massively.
Yes, in theory an idiot user could run bad code, but under a well implemented SELinux policy, while the code may run, it wouldn't actually have rights to do much of anything. At worst it might be able to fill up the home partition with useless data, or something along those lines, but spam bots and zombies and mass mailing viruses would be a far more difficult task to write indeed. A sufficiently smart idiot could grant the process the rights to do what it wants, but really...
Yes, such a system is not a cure all. People can still do bad things to themselves, and no matter how well you build it, there's always an idiot who can break it. It does, however, significantly raise the security bar on what it is easy to trick a user into doing.
Jedidiah
Well duh. But that is hardly time developing a better villain, and actually Palpatine doesn't get that much either. That's a pretty sorry excuse as far as the lack of credible interesting villains goes. You may as well say "But Darth Vader (as Annakin) gets lots and lots of screen time!" Yes, he does, but it hardly makes up for the lack of a villain of any note.
Jedidiah.
Dooku is evil, and he is more than likely the current apprentice to sidious. Anakin takes his place after Dooku is disposed of.
Which, I expect, will occur in the very first act of the movie. Part of the problem of Ep I and II (and from the look of "General Grievous" Ep III as well) has been Lucas' "rent a villain" approach - we get the villain of the week for each new episode, and that means much less time actually constructing credible villains. Honestly, compare the screen time either villain has gone in either of the current episodes to the screen time Vader gets in any of IV, V or VI. Poorly developed completely disposable villains don't tend to make for very interesting stories, especially when new evil villains crop up out of nowhere all the time (they just get mentioned in the scroller at the beginning as the new villain - no other explanation required).
Yes, yes, Darth Sidious is consistent through all 3, but he hardly has any screen time, and isn't exactly a menacing villain that our heroes must battle with, given that none of them have even seen him yet.
Jedidiah.
Slashdot has preferences, and they are helpful. I can't be bothered reading everything from people with high Karma (because there's a lot of them these days), but some people will. If they don't want to read this they can:
(1) Set their threshold to +3
(2) Make me a foe, and give foes a -1
(3) Set the Karma posting bonus to +0
If they're anonymous cowards... they get whatever they're fed, and that's not my problem.
Jedidiah.
The problem with LaTeX is that it's impossible to Google for document classes or other documents about it, because of the porn that comes up. Vanilla TeX has the same problem, except that you get derogatory pages about the President, instead.
You could start by actually looking in the obvious places first, and save yourself the trouble. There are a suprising amount of prepared packages and documentation for LaTeX available, you just actually have to look for it.
Jedidiah.
I had to remove OOo from my home box last night. I needed the disk space back. Why does a office suite on a Linux box have to take up a gig of disk?!
I've had similar issues. It's the i18n module that is most annoying to me, in that it is listed as a dependency (and so gets downloaded and installed) but isn't really required in a lot of cases. Yes the i18n support is a great thing... but could the packagers at the distros make it a little more optional?
Jedidiah.
With regard to your sig: I agree, it is an issue. I set my preferences to remove karma posting bonuses, which helps a bit, but it still doesn't really cover it.
I suggest a more useful scheme would be to have an uncapped logarithmic scale for score based on mods. That is, you need 1 mod point to get to +1, 2 more for +2, 4 more for +3, etc. or some similar scheme.
Odds that Slashdot implements such a thing: very very low.
Jedidiah.
It's okay though because the lawyers are suing MGM for a little over a million dollars to cover legal fees. I'm glad that the big money is going where it's supposed to. I mean, the important thing here is that the lawyers get paid right? My understanding was that that was the point of US society these days: to pay Lawyers.
Jedidiah.
Google could also roll out a thin client service in which you do everything within any browser window connected to Google. Google could host user accounts that go beyond email and search. A person could browse through the google browser, manage their googlefiles, run googleoffice, send gmail, buy stuff through froogle, etc. It would be a totally portable thin client service.
Which all sounds fantastic, but they could do all of that with either XUL, or XAML (if it ever shows up). In the end they may simply be getting in some people to do some XUL applications. If they want to include a browser in that suite of applications then rebranding Firefox seems the obvious way to go. As for having an office suite - that one is a surpringly large amount of work.
In general the concept of Google using, say, XUL to create a suite of basic applications (a browser, a mail reader, a chat client, some calendaring and groupware etc. makes some sense. I wonder how much of that would simply be a matter of rebranding all the mozilla sub-projects?
It's possible, but I still rate it all as highly unlikely.
Jedidiah.
I've done some cryptology too - enough to know that your analogy is rather unfair. Here's a better one:
You have a new unknown crypto algorithm to analyse. You know some basic facts about the structures of the algorithm (stream or block, etc.) and a fairly large sample of encrypted text, but nothing else. Based on what you know about the algorithm you try to make logical guesses about roughly how the whole should work, and construct a broad model algorithm. The goal now is not to match bit by bit the output of your sample ciphertext (you'd need the keys and IV for that anyway) but simply to try and make qualitive predictions about the characteristics of the resulting cipher. For instance you might, based on your model, say that you would expect a higher than normal occurrence of the bit sequence 11010. If when you go through your sample data you find that to be true you would probably suspect you may be on the right track in the reconstruction of the algorithm.
That's basically what the Climatologists are doing. They aren't predicting what the next bit will be (the temperature tomorrow), but they are trying to predict what broad patterns will be apparent in the system. They are making the model not just from prior data, but from knowledge and understanding of how the atmosphere and the climate ought to work based on what we do know about air masses, heating, cooling, and all the associated dynamics (equivalent to knowing a little about the crypto algorithm, and having to fill in the details to make your model).
Jedidiah.
The problem here is that you're answering the wrong question.
Mostly I was trying to correct the glaring misuse of "Chaos Theory" which, being a mathematician by trade, I found to be quite irritating.
The question is, can climatological models predict weather ten years in advance?
Indeed, that is the important question.
My answer is "I don't know. Let's get the best climatologists together and sponsor a year of research. At the end of that, they give us a ten-year projection. We keep on going. Ten years later, if their projection is accurate to within ten percent, we say their model has merit and go on from there."
You seem to presume we haven't been doing that for the last 10 years. Okay, the model at the level of detail used in this compute hasn't been tried before, but it is, I presume, simply a refinement of previous climatic models which have been in use for the last 20 years or so. Personally I don't know how well climatic modelling has gone for the last 10 years, nor how accurate any predictions were. I expect it is all in the literature from 10 years ago though, and you can probably look it up.
I'm pretty neutral on this debate for now, but I think the dismissals of "we can't predict the climate" are awfully premature. If they showed that predictions of global warming from 10 years ago have proved to be complete crap... well, then I'd start listening. I've never heard that particular argument trotted out though. Instead I've mostly seen a lot of climatologists who seem pretty confident that their models (up to a reasonable margin of error) are reasonably good. Apparently this particular model was used to predict the weather for the last 50 years based on data from 50 years ago and did a decent job. That's no guarantee that the model works, but its not a bad pointer that maybe they're onto something.
Jedidiah
Does anybody remember how Chaos Theory was first postulated?
Yes, I seem to recall reading about Henri Poincare finding a "homoclinic tangle" while trying to solve the problem of the stability of the Solar System (to win a prize put up by the King of Sweden). It amounted to a strange attractor, and a chaotic system. That probably wasn't "the first" being only around 1890, but it was one of the earlier points. Why what did you have in mind?
The crux of Chaos Theory is that some systems will NEVER be predictable because there are so many variables that it is impossible to know all the starting conditions.
Not really. Chaos Theory generally has more to say about what you can predict/say about such systems, and the fact that your predictions will have to be formed differently than those of nice classical linear systems.
Or were you talking about "Popular Chaos Theory" where people who don't know what they're talking about make vague generalisations about what they think "Chaos Theory" probably means, largely based on a few half assed descriptions from MIchael Crichton books and Hollywood films?
If a computer model can't even predict weather more than a few days out, how is it that it can predict weather a hundred years from now?
Really? I can make quite a few fairly accurate predictions about the weather over the coming year: It will be (in the northen hemisphere) warmer over June July and August, but will cold come the end of the year. On average Florida will be warmer than Maine this year. Seattle will see a lot of days with rain this year.
You see, despite it being a chaotic system, it's still possible to discuss some of the more qualitive aspects with some accuracy. I can't predict exactly what the weather will be like on July 23rd, but I can make a fairly accurate guess that it will be warmer than the weather tomorrow (unless you're in the southern hemisphere). They can't tell you exactly what the weather will be like 100 years from now, but they can make qualitive broad statements about it.
Chaos Theory has to be the single most misunderstood and misrepresented theory next to Quantum Physics. Could you please refrain from further spreading this bizarre contaminated view of what is, actually, an interesting field of mathematics.
Jedidiah.
bin Laden does have a stated intent to kill Americans yes. The fundamental goal they are striving for, however, is the for muslim states of the middle east to become Islamic states. As to killing Americans... yes, he will try, and is limited only by his capability. I am not questioning intent, I am questioning capability.
Aum Shinrikyo was a japanese cult that had billions at their disposal, and were interested in making chemical and nuclear weapons. With all that money, and recruiting intelligent young grad students from major Japanese universities they managed a single sarin gas attack in 1995 killing 12 people.
al Qaeda has used coventional explosives in all their attacks, and have, aside from 9/11, failed to show anything resembling global reach. In fact there is much evidence that al Qaeda is more of a venture capital firm for anyone wanting to attack Americans, and don't have any extant network at all - and never did.
Which raises the point that really the issue is constraining capability, and capability mostly takes the form of money. In theory the US has a vast and powerful foreign intelligence agency (NSA) that is supposed to be good at tracing and shutting down money flowing into terrorist causes. That, it would seem to me, would be the most effective way to fight the war on terror: quietly and efficiently in the background, not drawing attention and inspiring other random terrorist groups to act in sympathy. I'm sure that's probably happening. All the rest - the tromping around of military, the random security measures applied piecemeal to random points of infrastructure in the US, the arrests of terrorist cells (usually innocent) - that's all for show. To be honest, it's probably more counterproductive than anything.
Jedidiah.
Sad to tell you this, but if Iraq gets a taste of democracy and it catches on in the middle east
Yes, that's one possible outcome for Iraq. Another possible outcome is that out of all the chaos Iraq manages to form itself into an Islamic state - what Zawahiri and bin Laden have been trying (and repeatedly failing) to do for the last 15 years or so. Who knows, Zawahiri and bin Laden believe that, sould that actually happen it will cause the muslim masses to rise up, overthrow their leaders and create a slew of Islamic states throughout the middle east. That was, is, and will be their goal. For the most part the state "jihad against America" is a way to try and rally support - a lesson they learned when their attempted efforts in, for instance, Algeria failed to attract the support of the masses (oddly the general population was rather repelled, rather than attracted by, their violence).
So, we have 2 competing theories:
(1) Install a democracy in the Iraq and watch democracy then sweep the middle east.
(2) Rally support by encouraging people to rise up against the Americans that interfere in middle east politics and institute an Islamic state in Iraq. The Islamic Jihad movement can then sweep the middle east.
To be honest, no matter what happens in Iraq, I don't really expect anything to "sweep the middle east". In the meantime though the two theories seem to be fairly well in balance. Iraq is in chaos, there's ill will by the common people toward the US, and Islamic clerics (like al Sadr) are polling very well leading up the elections. In the meantime Iraq is actually having free and open elections so democracy will arrive. It looks to me if things could go either way - which means I'm not so sure this whole "introduce democracy and watch it spread through the middle east" idea was quite all it was cracked up to be.
Jedidiah.
Ya, because the distruction of the twin towers on 911 was just a hoax.
Sure, it happened. What we don't really have is much evidence that it was anything more than a freak occurence. How many terrorist attacks have occured on US soil? How many attacks in the US have been foiled? Have you actually looked at any of those cases of "terrorist cells" in the US? They pretty much all either got quietly dismissed, or otherwise shuffled down and effectively dropped.
Exactly how big an organisation is al Qaeda really? I mean, besides what the administration has told you about a globe spanning terrorist network with amazing resources, what do you actually know about al Qaeda? Try digging around a wee bit and read material from people that were following Zawahiri and bin Laden from prior to all this. This article might be a good place to start.
If you look at the statistics, global terrorism has been in decline since the mid to late 80's. Aside from the occasional anomaly (9/11), the only growth area of terrorism in the last 10 years has been... Iraq in the last 2 years.
Countries in the Indian Ocean just suffered from an earthquake and tsunami that literally killed 100 times as many people as 9/11. Most people seem willing to accept that that was just a freak occurence. Sometimes bad things happen. It is possible that 9/11 was just a freak occurence, and in no way indicative of a vast and powerful global terrorist network. How can we know? The evidence for the vast network is surprisingly lacking if you actually look at it.
Jedidiah.
Granted. Sorry to springboard from your post to make a different point.
The "system" only requires such because "investors" (which is to say all the slashdot geeks busy making day trades) only care about short term quarterly profits rather than true long term earning potential. Thus getting profits yet cutting staff (boosting quarterly earnings, though potentially damaging earnings down the line, be it from having to rehire on to refill those positions more expensively later, or simply garnering ill will from prospective employees).
The "system" being at fault is not so much the corporations, but the internet day trading weenies.
Jedidiah.
The reality, however, is that while hard decisions like this might be very good for the investors in the short term (and hence help boost investment) because of the boost in quarterly profits, in the long term EA is aquiring quite the reputation as a slave driver with no loyalty to people it employs. If EA develops too much of a reputation for that they won't get anywhere near the same employee pool to pick and choose from - the smart people will be staying away. Long term it is potentially gutting the company if they push it too far.
And that, right there, is the big problem that causes so many people to complain about big corporations: They have come to favour short term quarterly profits over long term sustainable profits. If you look at most complaints, from environmental, to labour, to political, when you pare it down it is occuring because companies are considering their short term future but not bothering to look at the long term results.
Jedidiah.
I seem to recall some Debian developers doing a computation to find the "centre of mass" of Debian developers - the closest you could get to a "centre" for open source. If I remember rightly it was somewhere north of Iceland. You can go there if you want.
Jedidiah.
He does work for a Finnish company, so that explains alot. Did you ever notice that there is no body more british than someone who USED to live in Britain?
So you're saying that someone with an obviously (to anyone who knows any Finnish) Finnish name Mikko Hyppönnen, who works for a Finnish company {is British/used to live in Britain} because he says "aeroplane"?
I mean, I know it was a troll, but try to do it with a little more class in future.
Jedidiah.
Imagine the effect that it would have if Firefox forked, its current maintainers left the project b/c all the devs went to the forked project. You would have a bunch of people still using Firefox that would never switch to the new one (hell it took them long enough to trust something without the little "e" already). Firefox would go to shit because no one would maintain it. The fork would grow in popularity among the educated. Once Firefox broke the people that switched would slowly migrate back to IE.
You would think that that's how it would happen, but that's the point - forks often don't pan out like you would expect. ECGS forked from GCC and eventually the GCC people accepted ECGS as the new GCC. XEmacs forked from Emacs, but Emacs development continued, and these days it is Emacs that has the more vibrant and creative dev team (AFAIAC - I used to be an XEmacs fan, but comparatively it has stagnated). Often enough things haven't gone as expected and people sticking with the "original line" get the benefits (eventually), whether it be from the fork being folded back in (GCC) or from the spurred competition arising from the fork (Emacs). Don't expect to be able to predict what will happen when a fork occurs. It's a lot more complicated than you might imagine.
Jedidiah.
One of the problems I see with high levels of security without a superuser-style account is the possibility of someone leaving, dying, or forgetting his password, and not being able to get to critical business data.
In SELinux, which I am more familiar with, and which also gets rid of the "superuser" account, everything is handled by context or role. That means you can isolate a process that wants "root" access to certain files by restricting its role to one that has access to only those files. Thus there is no "root" account that has access to everything. At the same time, it's possible to create a role that allows suitable access to make changes and/or recover lost data if necessary.
I presume Coyotos, with its "capabilities" will work similarly - ie. there is no "root" account that has access to everything, but instead various capabilities that bound access to various resources.
Jedidiah
The Emacs/XEmacs fork is given passing mention in the article, but is actually one of the more interesting ones. At the time XEmacs really did represent a step forward, mostly in its embrace of an X based GUI using modern toolkits. Consequently XEmacs tended to romp along and be the feature leader. Most recently, however, the situation has reversed. It is now XEmacs that is unwilling to use modern toolkits, and GNU Emacs is starting to push back.
Let's be frank, these days when we say "modern GUI toolkit" for X we mean wither GTK or QT. XEmacs does have GTK support, but the developers are not interested in it, and mostly it is just slow, and bug ridden, even in CVS. Compare that to Emacs, which has finally decided that GTK might not be such a bad idea. The current CVS versions of Emacs have excellent GTK support, making full use of the latest versions of GTK. It looks and behaves very nicely indeed, and integrates quite well into a GNOME desktop. The new GNU Emacs will also sport excellent Unicode support. It will be interesting to see how the GNU Emacs/XEmacs debate stands once XEmacs 22 and Emacs 22 come out. I expect to see GNU Emacs get a real boost in popularity.
Jedidiah.
Well, I'm not using Gentoo, but it's not just %10. I get _huge_ performance gains on my 3.2GHz Northwood if I use proper compiler optimizations. I'm using this workstation for audio related tasks. Ardour (Digital Audio Workstation software) and several plugin effects / audio encoders are 30-40% faster compared to precompiled i386 binaries.
See, for my money I much prefer to let the distribution prepackage binaries for the 98% of stuff that makes up the base system and doesn't really require speed gains (and won't really get any) because that's just easier. Then you can compile your own versions of whatever few applications and libraries you're using that desperately need that extra performance gain. So, for example, in your case I'd just use any old distro and then compile my own version of Ardour, and more importantly compile all the plugins as well, using every trick I can to squeeze the extra performance.
Really, do you get 40% speed gains out of KDE and GNOME? No, not really. So why not let them be prepackaged. How about Python, perl, libXML, Firefox, gettext and all those other bits and pieces? Again, for all it matters they may as well be prepackaged.
The truth is, for speed gains from compiler flags it varies from application to application. For some -ffast-math is a gain, for others not. Likewise with every other flag. Trying to set a bunch of flags for everything is as likely to provide inefficiencies in various bits as gains. Besides, for the majority of applications the speed just isn't that important. KDE spends most of its time idle waiting for input..
Jedidiah.
The other problem with switching over to Dvorak is most common keyboard shortcuts aren't so convenient anymore.
Just use Emacs. Sure some keyboard shortcuts will be harder to reach, but some will be easier to reach. By using every possible key combination for something Emacs ensures that you're no worse off no matter what keyboard layout you choose.
Jedidiah.