Not sure if anyone's still reading this thread, but I'm happy to talk more about it. Reply with questions if you like!
Then, let me take you up on that offer!
I don't know half the physics I'd like to know, but I have at least been reading some popular science books about string theory, and in my understanding, even M-theory suffers from the fact that you can choose an almost arbitrary geometry of the new dimensions (from a certain class of geometries) and receive a theory about a certain universe; the hope being that finding the right geometry will yield our universe (though I'm sure there are other parameters as well). Even so, the foremost touted advantage of string theory seems to be that it eliminates the necessity of arbitrarily chosen constants for a variety of phenomena, as is present in quantum field theories.
Maybe I've misunderstood something fundamentally, but I don't understand how the choice of a geometry of the dimensions can be considered less arbitrary than the choice of particle masses. I know you said that you aren't a string theorist, but you seem to have a greater idea of the theory than I have, at least. I'm really hoping you can answer that question!
I Ain't a String Physicist, but as far as I understand it, these different state vectors of which you speak, are what originally gave rise to the five different, predominant string theories. If I'm not mistaken, the goal of M-theory is to unite the five into one theory without arbitrary variants.
CDs have as much bandwidth as you can hear (in stereo). There is absolutely no point in a higher sample rate. That's not true, though. At 20 kHz, "CD quality", being sampled in 44.1 kHz, cannot represent any difference between a sine, square, triangle or, for that matter, any other waveform. It also encounters serious problems when representing the combination of 2 or more frequencies of sine waves being played simultaneously. Perhaps needlessly to say, it cannot properly represent the much more complex combinations that make out music.
Generally speaking, if you want to represent a sound with component frequencies up to f Hz, you should sample at at least 10f Hz (though I don't have a readily available source to cite for that). Therefore, the Audio DVDs sampled at 192 kHz makes a much better match for audible frequencies.
The first "fact" the GP references is something he pulled out of his ass, and is unable to substantiate it with a reference -- why should he respond to anything else? Maybe because just responding to one weak argument while ignoring the more fundamental issues would be what you call setting up a strawman.
Not that it doesn't matter whether MSOOXML encodes VBA data into files meant for interoperation, but that isn't the fundamental issue at hand. The fundamental issue (or, at least, one of them) is the fact that there already is an ISO standard for interoperable office documents, and that that fact alone makes it stupid and evil by Microsoft to try and push another one down our throats -- whether it is a good standard or not. Instead of making up a completely new format, they could just have standardized a few additions to ODF while reusing the rest of it.
It may or may not be true that OpenOffice.org (and maybe also Abiword) have to support both ODF and MSOOXML, so that from Miguel's point of view, it "doesn't matter much" if MSOOXML is adopted as an ISO standard, but that's just missing the point. The point is that by having one common format for office documents, it isn't just possible to interoperate between well known office suits such as MSO and OOo, but that ideally it should be much easier to write new ones as well, rather than keep being locked into the ones that already exist. Having to support both ODF and MSOOXML is not something that makes that task any easier.
something which the whole Unix world[2] did years ago.
[2]. All non-toy distros do this by default, and if not for few whiners, non-UTF8 locales would probably be dropped by now.
Unfortunately, that isn't quite true. As far as I know, none of the BSDs use UTF8 by default. I have verified it on FreeBSD 6.2, but I cannot imagine the {Net,Open}BSD would use it either. Internationalization is definitely an area where Linux is above and ahead of BSD.
I've heard rumors that that's one of the things being improve for FreeBSD 7.0, but I don't know just how improved it is.
Have I missed something? Not that I use Windows all that often, but last I tried, running the task manager with M-C-Del requires none of Explorer's cooperation, since M-C-Del isn't handled by Explorer (and, in fact, cannot be, since M-C-Del is a "secure key combination" in NT, and always goes through the kernel to a privileged handler). I've used it several times to kill Explorer when it has hung.
While 'Activation' and WGA are ostensibly an anti-pirating measure, in my eyes Microsoft is trying to steer the desktop PC market back to the old mainframe model of paying a yearly (or perhaps monthly) tithe to keep your computer working. Get the market used to phone-home features, and slowly close the net. They've been interested in subscription models for quite awhile, now.
That is truly interesting -- it wouldn't surprise me at all if that is correct. One of Microsoft's greatest problems for quite a while does seem to have been to compete with their own products, so they are often having problems making money by selling upgrades to their older products (like Office).
If they could switch the world over to a rental paradigm, that wouldn't be a problem for them anymore. They can just sit on the current version of Windows and Office and just watch the money roll in. Instead of having to be somewhat innovative every now and then in order to outdo their own work, they would only need to produce upgrades to match the competition.
It is also consistent with some of their patents recently posted on Slashdot. Unfortunately, I don't remember the patent numbers (and I am too lazy to Google them), but one of them was about an ad-driven operating system, and the other was about an operating system where the "core" is free of charge, but the user rents various add-on components (like IP networking). I have no doubt that if there is any company that wouldn't have the morals to stop themselves from pushing something so sinister onto the public, that company would be Microsoft.
It will be much interesting to see how it develops.
It is true that these kinds of things don't exactly increase the respect for the standards organizations like SIS, but when looking at it from their side, I don't think that it's very easy to come up with something that they could have done to prevent it. After all, the companies that were hired to vote for Microsoft were still independent companies in their own right. I think it's hard to think of a reason why they should not have been allowed to vote.
Vista is USING the memory that is unused. What do you pay for your memory for? To have it unused? If nobody is using it, Vista will just use it damn it!. Don't worry, if some application will need it, Vista's memory manager will give it back. What you are talking about in that paragraph, however, is the block device cache. Indeed, if some process needs more memory, Vista will (hopefully, at least) take some block cache page and give it.
However, a lot of the memory used by Vista is not block caches, but private memory allocated to the heap of some processes, and that won't be given to other processes if they need it. It may, of course, be swapped out and then having the backing page given away, but that still requires the pagefile write latency to pass.
See, it's fine if an operating system uses up your entire RAM for caches, but if it uses it up with heap memory, you're going to get performance problems when the page faults become too frequent, and that, in my experience, is too frequent a problem with Vista.
Alternatively, it could be dealt with on a case-by-case basis. Seriously, the ISO must be seeing what is going on, since this isn't exactly the first story on this subject. Weren't at least the U.S. and Portuguese standards bodies reported on Slashdot to have experienced the same thing? Even if the national standards bodies themselves for some weird reason can't see what's going on, then at least the ISO itself should be clear-sighted enough to launch an investigation into the matter. Surely, they don't want to become an organization where just about anyone with enough money can buy a "standard"?
I, too, do not agree with the GP's idea of actually taxing high-power computers, but I do think he might well have a point in just rating the computers after their power usage. If people buying computers see some real statistics of how much it is going to cost them in electricity to run their new computer, it is very likely that they are going to choose after that criterion, which will drive manufacturers to make more power-efficient computers. Which is good, because if they make computers that draw less power, then I, too, could get one of them and pay less for electricity (and having them run longer on the UPS:).
I certainly understand that point, and very much experience it myself. If I don't turn up the volume on my own car stereo, the engine and road noise just drowns out too many important sounds in the music leaving me unable to appreciate the music to its fullest. However, the simple remedy should just be to crank up the volume. What I don't understand is why the record companies would want to do that for me while at the same time decreasing the sound quality when there is a perfectly good volume knob I can crank myself.
I mean, surely the reason isn't that people just appreciate the laziness of just not having to turn the volume knob too far. But then what is the reason?
I remember seeing that video linked from the last time a loudness-related article was posted on Slashdot, and I found it quite shocking at the time. Having had some time to reflect upon it, though, I still don't get why the record companies would even want to be doing anything like that. I mean, it is obvious that a lot of sound quality is being lost, and I just don't get what advantages they gain from it. Do people actually, thoughtfully, choose to buy loudness-wrecked records because they... want them to sound louder without having to turn up the volume? It doesn't make sense to me.
Gentoo is like going to a restaurant, ordering your dinner, and having the chef take you back into the kitchen and put you to work making your own meal. I like an OS with a little LESS configurability than Gentoo. Some like it though. Using Gentoo, I'd argue it's a lot more like having the chef take you back into the kitchen and let you watch while he's preparing your meal, not like having put to work on it. You get to see how he works, and if you like, you can make small suggestions, like using a particular spice, leaving the peas out, or put on a bit more sauce.
There are those who argue that Gentoo gives teh performanze by allowing you to tweak the compile flags, but I can't say I've ever noticed any difference. If anything, my Gentoo systems run slower than precompiled versions like Fedora or Ubuntu. What I like with Gentoo is getting to choose, for example, if I want Qt installed or not, or being able to apply small patches to various programs.
However, the last year or so, I've been growing less and less fond of Gentoo, I have to admit. It feels as if the quality of the distro has deteriorated, and I have also come to like the more "integrated" feeling of Fedora and Ubuntu (and, to an even greater degree, FreeBSD). I have to find out either 1) a way to dist-upgrade Fedora without reinstalling and without using the default botched upgrade procedure or 2) a way to patch individual packets in Debian/Ubuntu before I'll be able to switch, though.
I'm pretty sure it's supposed to be my, as in "not an answerable question". If you assume the GP's premise that only Mac users listen to Coldplay, the very question is invalid, since it asks what would happen in an impossible situation. Therefore, it has no answer. The classical example is the question "have you stopped beating your wife?", to which one normally would not want to answer either "yes" or "no".
a) Reasonable. The fees required are in line with whatever it is. It's not a "Oh you want a license for that video codec? Ok $1,000,000 per player, no cap." That's clearly unreasonable and designed to keep people from licensing it. One has to wonder what that really does entail. Leaving Free Software aside for a second, what about ordinary people who just like to DIY, such as myself? When licensing patents for a MPEG implementation to a company like Microsoft, Sun or Apple, $1,000,000 doesn't seem at all unreasonable if it is a perpetual license (not a "per player" license). If I want to license it just for my own purposes, however, it is clearly an unreasonable amount that I couldn't afford in a lifetime. Then again, surely they don't have the right to choose a price arbitrarily depending on the licensee, right?
I know you were kidding, but what bothers me very much is that top-posting is, as far as I'm aware, yet another problem initiated by Microsoft.
When replying to something in MS Outlook (at least the last time I looked), what you got was the standard "Original Message:" header, followed by the unquoted original message, more or less forcing you to top-post unless you want to quote the original message manually. I don't think that Outlook Express does the same though, although I'm not sure.
It may well be that I'm uninformed and that some other program began this vile malpractice before Outlook (cc:Mail comes to mind as a potential candidate), but at the very least it is being encouraged by Microsoft.
Heh, speaking of having done stupid things... Not too long ago, a local user account got compromised on my home network because it had a weak password. The attacker hadn't managed to do any local privilege raising, so the attack was isolated to only that user's account. I, wondering what the attacker may be doing, was snooping around a bit in the directories he had created, and since they were 700, I had to be root. While snooping around -- as root -- I found a program he had left behind and wondered what it did, so I ran it. Yeah, that was stupid.
Well, it gave me a great reason to finally replace the aging FC2 installation with Gentoo anyway...
[...]I suspect that the first several implementations (at least) will leave the resulting systems vulnerable to subversion, and this subversion would be difficult, at best, to detect.
personally I have to say I quite like apaches configuration file format. It has an easy to follow struture that will be easy for anyone who is familiar with html to follow without the over-verboseness of xml.
In my mind, the problem with httpd.conv isn't the file format itself (although I do find it terribly ugly), but rather the actual contents. It is so large! Every time I install a new Apache web server (fortunately, I've only had to do it around five times in the last few years) it takes me hours to walk through all the options and try to get everything right, from hostnames and admin addresses to MIME type settings and extension modules. Of course, I never get it right on the first go either, leading to a number of problems until it has finally been tweaked correctly. Going through a Linux kernel configuration is a breeze in comparison.
Of course, Apache isn't alone in that. Similar programs include at least Squid, Dovecot and FreeRADIUS (some would probably mention sendmail, but I wouldn't, because the sendmail.mc format is very easy to deal with in my mind). It's great that they are very customizable, but when you have to read through thousand of lines of comments in the example config files just to get a basic idea what the setup is up to, I think something has gone wrong along the way.
Just as an aside -- On the other side of the spectrum lies programs like UW IMAP or the MIT Kerberos ftpd, which have no configuration files at all. They may be slightly configurable with simple command line arguments, but they are very much plug and play as long as they do what you want.
(Aero-like feats don't count as major contributions to operating systems...)
I would disagree with that, depending on exactly what you mean. It is true that the Aero theme and shell functions aren't very impressive at all, but the underlying compositing technology actually is pretty cool, if you ask me (of course, Microsoft was last to join that party). With the amount of memory today's computers have, I think it is a rather good call to give each window an actual backing store, rather than having to rely on expose events to redraw it every once in a while. It both makes the system more responsive with regard to a number of common windowing functions, and it most likely saves bandwidth, too (bus bandwidth locally, or network bandwidth if you run X11 over a network).
Add a compositing window manager, and you can actually do rather useful things with it. I'd say OSX has been the champion at actually doing useful things with compositing. If you care for such things, Beryl is probably the champion at doing eyecandy. I'm not sure where that leaves Vista, of course, but that's not my problem.
Before I had read the summary, I thought the article was about the desktop metaphor for GUI shells, and I though "if only". The actual article turned out to be much less exciting, unfortunately.
To be honest, I don't know if there is some actual GUI grafted on it, but if you're using the internal open/save dialog boxes (as opposed to the operating environment's native ones!), and you have a WebDAV-enabled URL, then you can just enter it right into the filename text box and it'll do the right thing. If the URL is a DAV collection, it will browse it as were it a local directory.
MOST Slashdot people would reflexively blame Microsoft for that, but neither the client nor the drivers were written by Microsoft.
The client and the drivers may not be written by Microsoft, but it wouldn't be extremely illogical to blame Microsoft for not designing their kernel to isolate the fault to only the sound system. After all, they have more than a decade of experience with bad third party drivers, so one would think someone over in Redmond would have gotten the idea to run them in their own address space.
Do I or anyone else blame Red Hat when I have trouble getting third party screen savers to build and work right on the newest iteration of Fedora Core?
Indeed, no, but isn't that kind of a different thing? I wouldn't blame Microsoft either if some third party driver didn't install or do its work properly on Windows. It's another thing if it regularly allows the system to crash because of it. I would be rather critical of the affected component on a Linux system as well if it allowed a third party screen saver to take the entire system down.
Slashdot Mirror
I don't know half the physics I'd like to know, but I have at least been reading some popular science books about string theory, and in my understanding, even M-theory suffers from the fact that you can choose an almost arbitrary geometry of the new dimensions (from a certain class of geometries) and receive a theory about a certain universe; the hope being that finding the right geometry will yield our universe (though I'm sure there are other parameters as well). Even so, the foremost touted advantage of string theory seems to be that it eliminates the necessity of arbitrarily chosen constants for a variety of phenomena, as is present in quantum field theories.
Maybe I've misunderstood something fundamentally, but I don't understand how the choice of a geometry of the dimensions can be considered less arbitrary than the choice of particle masses. I know you said that you aren't a string theorist, but you seem to have a greater idea of the theory than I have, at least. I'm really hoping you can answer that question!
I Ain't a String Physicist, but as far as I understand it, these different state vectors of which you speak, are what originally gave rise to the five different, predominant string theories. If I'm not mistaken, the goal of M-theory is to unite the five into one theory without arbitrary variants.
Generally speaking, if you want to represent a sound with component frequencies up to f Hz, you should sample at at least 10f Hz (though I don't have a readily available source to cite for that). Therefore, the Audio DVDs sampled at 192 kHz makes a much better match for audible frequencies.
Not that it doesn't matter whether MSOOXML encodes VBA data into files meant for interoperation, but that isn't the fundamental issue at hand. The fundamental issue (or, at least, one of them) is the fact that there already is an ISO standard for interoperable office documents, and that that fact alone makes it stupid and evil by Microsoft to try and push another one down our throats -- whether it is a good standard or not. Instead of making up a completely new format, they could just have standardized a few additions to ODF while reusing the rest of it.
It may or may not be true that OpenOffice.org (and maybe also Abiword) have to support both ODF and MSOOXML, so that from Miguel's point of view, it "doesn't matter much" if MSOOXML is adopted as an ISO standard, but that's just missing the point. The point is that by having one common format for office documents, it isn't just possible to interoperate between well known office suits such as MSO and OOo, but that ideally it should be much easier to write new ones as well, rather than keep being locked into the ones that already exist. Having to support both ODF and MSOOXML is not something that makes that task any easier.
something which the whole Unix world[2] did years ago.
[2]. All non-toy distros do this by default, and if not for few whiners, non-UTF8 locales would probably be dropped by now.
Unfortunately, that isn't quite true. As far as I know, none of the BSDs use UTF8 by default. I have verified it on FreeBSD 6.2, but I cannot imagine the {Net,Open}BSD would use it either. Internationalization is definitely an area where Linux is above and ahead of BSD.I've heard rumors that that's one of the things being improve for FreeBSD 7.0, but I don't know just how improved it is.
Have I missed something? Not that I use Windows all that often, but last I tried, running the task manager with M-C-Del requires none of Explorer's cooperation, since M-C-Del isn't handled by Explorer (and, in fact, cannot be, since M-C-Del is a "secure key combination" in NT, and always goes through the kernel to a privileged handler). I've used it several times to kill Explorer when it has hung.
While 'Activation' and WGA are ostensibly an anti-pirating measure, in my eyes Microsoft is trying to steer the desktop PC market back to the old mainframe model of paying a yearly (or perhaps monthly) tithe to keep your computer working. Get the market used to phone-home features, and slowly close the net. They've been interested in subscription models for quite awhile, now.
That is truly interesting -- it wouldn't surprise me at all if that is correct. One of Microsoft's greatest problems for quite a while does seem to have been to compete with their own products, so they are often having problems making money by selling upgrades to their older products (like Office).If they could switch the world over to a rental paradigm, that wouldn't be a problem for them anymore. They can just sit on the current version of Windows and Office and just watch the money roll in. Instead of having to be somewhat innovative every now and then in order to outdo their own work, they would only need to produce upgrades to match the competition.
It is also consistent with some of their patents recently posted on Slashdot. Unfortunately, I don't remember the patent numbers (and I am too lazy to Google them), but one of them was about an ad-driven operating system, and the other was about an operating system where the "core" is free of charge, but the user rents various add-on components (like IP networking). I have no doubt that if there is any company that wouldn't have the morals to stop themselves from pushing something so sinister onto the public, that company would be Microsoft.
It will be much interesting to see how it develops.
It is true that these kinds of things don't exactly increase the respect for the standards organizations like SIS, but when looking at it from their side, I don't think that it's very easy to come up with something that they could have done to prevent it. After all, the companies that were hired to vote for Microsoft were still independent companies in their own right. I think it's hard to think of a reason why they should not have been allowed to vote.
However, a lot of the memory used by Vista is not block caches, but private memory allocated to the heap of some processes, and that won't be given to other processes if they need it. It may, of course, be swapped out and then having the backing page given away, but that still requires the pagefile write latency to pass.
See, it's fine if an operating system uses up your entire RAM for caches, but if it uses it up with heap memory, you're going to get performance problems when the page faults become too frequent, and that, in my experience, is too frequent a problem with Vista.
Alternatively, it could be dealt with on a case-by-case basis. Seriously, the ISO must be seeing what is going on, since this isn't exactly the first story on this subject. Weren't at least the U.S. and Portuguese standards bodies reported on Slashdot to have experienced the same thing? Even if the national standards bodies themselves for some weird reason can't see what's going on, then at least the ISO itself should be clear-sighted enough to launch an investigation into the matter. Surely, they don't want to become an organization where just about anyone with enough money can buy a "standard"?
I, too, do not agree with the GP's idea of actually taxing high-power computers, but I do think he might well have a point in just rating the computers after their power usage. If people buying computers see some real statistics of how much it is going to cost them in electricity to run their new computer, it is very likely that they are going to choose after that criterion, which will drive manufacturers to make more power-efficient computers. Which is good, because if they make computers that draw less power, then I, too, could get one of them and pay less for electricity (and having them run longer on the UPS :).
I mean, surely the reason isn't that people just appreciate the laziness of just not having to turn the volume knob too far. But then what is the reason?
I remember seeing that video linked from the last time a loudness-related article was posted on Slashdot, and I found it quite shocking at the time. Having had some time to reflect upon it, though, I still don't get why the record companies would even want to be doing anything like that. I mean, it is obvious that a lot of sound quality is being lost, and I just don't get what advantages they gain from it. Do people actually, thoughtfully, choose to buy loudness-wrecked records because they... want them to sound louder without having to turn up the volume? It doesn't make sense to me.
There are those who argue that Gentoo gives teh performanze by allowing you to tweak the compile flags, but I can't say I've ever noticed any difference. If anything, my Gentoo systems run slower than precompiled versions like Fedora or Ubuntu. What I like with Gentoo is getting to choose, for example, if I want Qt installed or not, or being able to apply small patches to various programs.
However, the last year or so, I've been growing less and less fond of Gentoo, I have to admit. It feels as if the quality of the distro has deteriorated, and I have also come to like the more "integrated" feeling of Fedora and Ubuntu (and, to an even greater degree, FreeBSD). I have to find out either 1) a way to dist-upgrade Fedora without reinstalling and without using the default botched upgrade procedure or 2) a way to patch individual packets in Debian/Ubuntu before I'll be able to switch, though.
I'm pretty sure it's supposed to be my, as in "not an answerable question". If you assume the GP's premise that only Mac users listen to Coldplay, the very question is invalid, since it asks what would happen in an impossible situation. Therefore, it has no answer. The classical example is the question "have you stopped beating your wife?", to which one normally would not want to answer either "yes" or "no".
When replying to something in MS Outlook (at least the last time I looked), what you got was the standard "Original Message:" header, followed by the unquoted original message, more or less forcing you to top-post unless you want to quote the original message manually. I don't think that Outlook Express does the same though, although I'm not sure.
It may well be that I'm uninformed and that some other program began this vile malpractice before Outlook (cc:Mail comes to mind as a potential candidate), but at the very least it is being encouraged by Microsoft.
Well, it gave me a great reason to finally replace the aging FC2 installation with Gentoo anyway...
Of course, Apache isn't alone in that. Similar programs include at least Squid, Dovecot and FreeRADIUS (some would probably mention sendmail, but I wouldn't, because the sendmail.mc format is very easy to deal with in my mind). It's great that they are very customizable, but when you have to read through thousand of lines of comments in the example config files just to get a basic idea what the setup is up to, I think something has gone wrong along the way.
Just as an aside -- On the other side of the spectrum lies programs like UW IMAP or the MIT Kerberos ftpd, which have no configuration files at all. They may be slightly configurable with simple command line arguments, but they are very much plug and play as long as they do what you want.
Add a compositing window manager, and you can actually do rather useful things with it. I'd say OSX has been the champion at actually doing useful things with compositing. If you care for such things, Beryl is probably the champion at doing eyecandy. I'm not sure where that leaves Vista, of course, but that's not my problem.
Before I had read the summary, I thought the article was about the desktop metaphor for GUI shells, and I though "if only". The actual article turned out to be much less exciting, unfortunately.
To be honest, I don't know if there is some actual GUI grafted on it, but if you're using the internal open/save dialog boxes (as opposed to the operating environment's native ones!), and you have a WebDAV-enabled URL, then you can just enter it right into the filename text box and it'll do the right thing. If the URL is a DAV collection, it will browse it as were it a local directory.