Free software is most often licensed under copyright, but in order for free software to be truly free, as defined by FSF, its use has to be free of any exclusive privilege owned by a third party: "The freedom to run the program, for any purpose (freedom 0)."
I don't believe I ever made the point of AAC (or FairPlay) being a 'Free' (FSF-Free) codec; but that it was an 'Open' codec. OpenGL is an 'Open' specification for a 3D graphics API. But OpenGL also has patented/non-Free extensions and components.
Of course, this argument comes down to whether an independant implementation of AAC is "Open Source" versus "Free Software". "Free Software" is tightly defined. "Open Source" is considerably less exact; just look at the early history of KDE/QT being 'Open Source' and not 'Free Software.' The OSI standars for being "Open Source" have yet to be accepted in the mainstream as the definitive definition; for now, 'Open' is typically understood to mean that 'the hood isn't welded shut.'
In any case; I view even a non-free standard (but open, published) standard like AAC/FairPlay as being considerably better than the entirely proprietary/closed WMA. AAC is in my view vastly better than WMA, if only because time is all that is required for it to become truly 'Free'.
You bring up Mongolia, but doesn't Dolby own counterpart patents in most developed countries that are markets for MPEG-4 audio technology (i.e. USA, Canada, EU members, and Japan)? For all I know, Mongolia has a similar counterpart patent.;) But that doesn't mean you'd be violating any copyright laws in implementing AAC/FairPlay. Which means the implementer would still own the copyright; he just couldn't make use of it...
Perhaps for the encoder, but I'd imagine that licensees get decoder source code so that they can port it to whatever DSP a device uses (as you hint later). Remember that pocket-size players don't have an x86 monoculture the way consumer PCs do.
Not necessarily; just because most of Microsoft's work is x86/Windows, that doesn't mean they are above writing the DSP code for the licensee. Since I don't have the money to even find out what the terms are, it's really just conjecture at this point. But in any event, there would be some serious NDA verbage to ensure that you cannot share the information with others. (Since the point of being 'open' is that an NDA becomes pointless -- it's all published publically anyway; what's the point of keeping the secret?)
AAC implementations aren't as independent as you think; they're licensed by a division of Dolby on a royalty basis, which is incompatible with free software. Your argument may be valid in a decade and a half when the essential AAC patents expire, but it isn't now.
You seem to be confusing a patent with a copyright. It's quite possible to write an implementation of AAC from the spec, without violating any of Dolby's copyrights. On the other hand, you cannot use/sell/distribute that implementation (in the US, at least), without violating Dolby's patent. The FAAC group has already made a free implementation of AAC; it doesn't violate Dolby's copyrights at all. It may well violate Dolby's patents without a proper license, as you pointed out -- but that only matters within the borders of the United States.
And Dolby's patents are only valid in the countries that recognize them as valid. I doubt a coder in Mongolia needs to worry about violating Dolby's patents if s/he chooses to implement an AAC codec from the specs; at least if the coder doesn't intend to sell it in the United States.
In fact, Dolby's patents can't keep anyone from making an independant implementation of AAC (even American coders). All the patents do is make it illegal for someone to use or sell the implementation within the United States.
Although it's of dubious merit to have your own implementation of AAC if you can't use it... but that has nothing to do with it being possible.
WMA can be licensed by any compay that wants to use it.
Really? And they get source code? What specific techniques are used to encode audio into WMA? Oh yeah... they're not available. Microsoft just points you at their pre-compiled library and says 'use that.'
All Microsoft licenses is the ability to use the precompiled WMA libraries for commercial uses (and even then only on Windows and to a smaller extent Mac). You don't get the source code at all. The actual spec on how to create an implementation of WMA isn't available except under an expensive license and NDA (and even then only for making a hardware decoder). Everything that plays WMA in Linux is reusing the microsoft-compiled WMA libraries (for Windows).
AAC is a published standard, with hundreds of implementations of the codec. WMA is an unpublished standard, and Microsoft's implementation or bust.
FairPlay is a layering of AES (an open encryption standard) inside AAC, and is documented; in spite of the leagalities involved, non-Apple implementations exist. WMA's DRM is still not fully understood outside an NDA.
If the reverse engineer it, Apple sues them. Um... and Microsoft won't sue if you reverse-engineer WMA (and its DRM) and use an unlicensed implementation (and therefore not pay Microsoft for it)? Just because you reverse engineer it doesn't really give Apple too much to sue over. It's that these compnaies were then using FairPlay without a license that spun the lawyers into action.
No company can license fairplay to use it. Fair enough; but Apple is free to make the licensing decisions it chooses to. But recall that there are more than a few Microsoft technologies that Microsoft refuses to license.
If this is your idea of an open format, then I'll take closed formats. An open format is one that is disclosed and/or published openly. There is nothing about FairPlay that is not openly published on the web. The internals of AAC, internals of AES, internals of how Fairplay joins the two. However, try to find out enough about WMA to implement it yourself. Or the DRM that WMA uses. Go ahead and fail. There is nothing about the WMA codec that is freely available. It's not an open codec, not an open standard.
I don't know of any independant software implementations of WMA (with or without DRM). FairPlay (without its DRM) is an AAC file -- a format that has more than a few independant implementations. And the actual wrapping of the AAC into a FairPlay formatted file is also easily handled.
An "open format" in this case is much like the OpenGL vs. Direct3D battle:
There is exactly one vendor of Direct3D: Microsoft. There are no independant implementations of D3D. The various versions of Wine that run D3D games on Linux simply convert the D3D call to an OpenGL call; it is not an implementation of D3D by any means.
However, there are literally hundreds (if not thousands) of independant implementations of OpenGL. Its spec is so complete that the Mesa project was able to start with the OpenGL spec and implement from scratch a conformant version of OpenGL. Apple has its own implementation. So does ATI, nVidia, SGI, Sun, Matrox, and Microsoft, among others. However (and this is the catch): Mesa can't call itself OpenGL due to OpenGL not licensing and/or certifying it. (Apple won't license FairPlay). That doesn't make Mesa (or OpenGL) any less functional, or any less 'open'.
The bottom line is this: I already have all the information I need to implement the FairPlay/AAC format; the spec is freely available. I can write a compliant FairPlay encoder, a decoder, and DRM protection facilities. And as long as I don't try to sell music using my implementation, Apple will be hard-pressed to touch me.
I do not have the ability or resources to even get the WMA specs to make a non-DRM'd implementation. I cannot write a compliant encoder, decoder, or DRM facilit
The biggest complaint I have is that I really don't own a physical copy of my music. I downloaded an album from iTMS. After listening to it quite a bit, I realized that I really don't like it. But there's nothing I can do about it. If I had purchased the CD, I could take it down to my local independent music store and sell them the CD and get a few bucks back. Thats the biggest problem I have with purchasing online. So if you are going to purchase an album online, better make damn sure you are going to like it.
Of course, the argument Apple would make here is that they offer the 30-second preview. Which is fine in many cases, but in others, the preview is from part of the song that is little more than dead air or is not at all representative of the song's overall feel.
That being said, yeah, it's not too wonderful that you can't take it to a CD reseller and trade it in/get cash back. But, the upside is that you aren't required to buy the entire album for one song. In the end, I think it comes out as a wash...
And the difference in quality between iTMS music and CD's? Well, none of the equipment I have is able to produce a noticible difference in quality; and this is with THX/Klipsch multimedia speakers and a THX sound card, or SennHeiser headphones; not audiophile grade, but well above consumer grade... I guess I'm just 'lucky' in that my ears are quite unable to tell the difference between compressed music and a CD. (Unless the codec *really* has issues, or is at a very low bitrate)
That being said, Steve Jobs also said that one thing the consumers surveyed insisted on is that they did not want a 'subscription' music service; but that they wanted to buy their music outright.
Frankly, as far as I'm concerned, it's just another subscription service, which is something consumers have repeatedly told the industry they would not accept. (Both with their pocketbooks and in surveys).
I honestly believe one of iTunes's greatest strengths is that they do not have a 'subscription' service-- you buy the song, or don't. You never have to keep paying to hear that song.
Not that I would use a WMA-based service even if it didn't try to force a subscription model on me; for all the ranting about FairPlay being a closed format, WMA is an even more closed format. Fairplay is the open AAC format, then uses AES encryption (another standard) for DRM; all the details are documented and freely available on the web (albeit not from Apple). WMA? It's not even published; the exact nature of it's non-DRM compression isn't available; in fact, more is known about its DRM than about its codec.
*If* it's actually fusion (and not just a weird chemical reaction) and *if* we can make it regularly reproducible, then Cold Fusion could essentially change the world.
Yeah. The real shame in my view is that the particular kind of reaction was largely dismissed by scientists. Those that did not dismiss it immediately found that it is reproducible; just not 100% of the time. And I still have yet to hear a reasonable explanation of what is actually happening in the Pons-Fleishmann cold fusion experiments. I have, on the other hand read from several sources that, while it may not be fusion, there is something happening that can't currently be explained to a decent degree of satisfaction.
If Pons & Fleishmann (and others involved) had not chosen to call it 'fusion', I wonder if the phenomenon would have been so thoroughly rejected as 'impossible'.
(You know -- like a high-energy particle bouncing back at its source as Rutherford discovered; I think his quote was something like it 'being as likely as shooting a bullet at tissue paper, and having the bullet bounce back.')
Pons & Fleishmann's 'cold fusion' is actually a rather dark chapter in the history of science -- where an observation was nearly universally deemed as 'impossible', even with independant groups reproducing nearly identical results (albeit only small a fraction of attempts were successful). But it had been given an unfortunate name, and was never really given a chance.
Hopefully this will further fuel investigation into exactly what is happening in the Pons-Fleishmann's experiment.
You gotta think back further. To circa 1992/3 or so.
Back when the PowerPC 615 was where it was at: A PowerPC core with a 486/66 bolted on. It would run both natively, so Apple would be able to run anything...
Yeah; parked illegally. That's a serious crime. It's the greatest threat to our country since flouridated water!
We should require all cars to be parked by a licenced and bonded valet, who is familiar with all state and federal parking regulations. It's simply too great a danger to the general public to be left to amateurs.
Same goes for shopping near a competitor's office. No more buying clothes next to Wall-Mart! You do not have the right to go shopping within 300 yards of a competitor's interests. Wake up and smell the restraining order!
Seriously, flying off the handle like that over a parking offense is really quite silly. And suggesting that any citizen of the United States can't go shopping in a store next to a lobbying organization's office is absolutely absurd.
Senator Frist is pretty far from a decent human being; but by your own admission the only illegal thing he did was park his vechile incorrectly. There isn't a law on the books saying where anybody can or cannot shop for shoes.
So, boot/tow the Senator's car, give him a parking ticket, and get on with life.
Honestly, if I knew beyond any doubt that the worst thing an individual would do while in public office is park illegally, he'd win my vote hands down.
Not that the Senator in question would actually qualify; by all accounts he is of most questionable character. But claiming he should be punished severely for parking illegally is rather astonishing. It was his only crime, and one that rarely even results in anything more than a modest fine (actually quite trivial at a Senator's salary).
And for "Free Speech Zones" -- I've seen first hand what happens when 'free speech' gets out of hand; the result is frequently bloody, with neither side being particularly innocent. A simple fact is that there are few things that incite more strong feelings than politics or religion; it's also a simple fact that not everybody will see things the same way.
And in spite of what many would like to believe, humanity hasn't grown past the intolerance of millenia long past. It would be nice in a civilized society if people would agree to disagree; but society is rarely so civil. Both 'liberal' and 'conservative' are typically mutually immature, mutually arrogant, and mutually foolish. The observation that each cannot imagine the other as having honorable goals speaks volumes.
So, when such an event happens, the local city governments have a choice: Either try to organize how people express their viewpoints, or to allow chaos to reign. In an organized structure, it's easier to physically protect those of opposing viewpoints from assaulting one another, given the limited resources of said city's police force. In the chaotic structure, it's virtually assured that there will be some form of assault over differing points of view.
Which leads to why the cities make these 'free speech zones:' If a disagreement does come to blows, the city can stand in a court and say that they did everything they reasonably could to protect the physical being of an individual, while providing the right of free speech to said citizens. You cannot claim that freedom of speech is infringed upon because you can't say what you want, where you want, when you want to. It's already been ruled upon that freedom of speech does not extend to yelling "Fire!" in a crowded theatre, and that doctrine is applied to 'free speech zones' -- that the government can place minimal restrictions on the right to free speech in order to protecect the physical well-being of its citizens.
There's also a difference between the right to speak your mind, and the right to force someone to hear it.
The courts have not (and probably will never) rule against the cities for protecting its citizens from each other at a political rally. However, there are several counts of when a court has ruled against a city for not physically protecting its citizens from each other at a political rally.
I agree completely with the arguments against using Gentoo for performance reasons.
It certainly isn't why I run Gentoo.
I run it because I like the way it works. I've recently taken a look at some of the other distros (SuSE, Fedora,... I still run debian on a fileserver, though). Distros like SuSE, and Fedora (even when using things like APT-RPM, yum, etc.) drove me nuts. There were dependancies that were *completely* unnecessary being installed. And the 'RPM Hell'; drove me INSANE. Debian is a serious step up from SuSE and Fedora in that respect, but Gentoo puts Debian to shame.
And I can honestly say -- I don't see where people have fits about having to compile everything. After the initial install, you can have any updates compiling at a well-niced priority. Unless I'm trying to play a CPU-intensive game (like Doom3, UT2004, etc.), I never even notice a performance difference. (I will admit this: my development work is done completely over SSH to another box, so emerging isn't slowing my 'real' compile jobs).
Even stranger is that most of the time, things don't even take long to compile. Huge apps/groups of apps like KDE, OpenOffice, and to a lesser extent, GNOME are an exception. And for those, I start the build when I leave work for the day, and it's done when I come back.
I don't use Gentoo for performance reasons. I use it because I hate the dependancy hell involved with pre-compiled binaries. I don't have to deal with it when using Gentoo; it's that simple.
And people whining about compile times -- seriously... unless you want to play Doom 3 while compiling KDE-- you've got issues, because compiling the apps isn't an issue at all.
Light bulbs, incandescent or fluorescent, running off of house current "flash" 120 times per second.
Yeah, but incandescent don't have as much as an impulse to the flash. This is mostly due to the fact that they produce light as side-effect of their heat, and the wire doesn't cool down anywhere near as quickly as the next peak in the current. As a result, incandescent bulbs have a much smaller delta between the 'bright' and 'dark' parts of the cycle. Turn the power off on an incandescent bulb, and it has a perceptible dimming after the power is gone. Flourescent bulbs just go dark instantly.
I guess I should have waited until there was an editorial summarizing the study then linked to that and write a summary which mis-states what the study says.
Well, obviously you're journalistic credentials haven't fallen to the levels of David Coursey. A columnist whose perception of reality is solely based on whom gives him the most advertising dollars / freebies.
Somewhat similar to John Dvorak: I used to respect his opinion as a knowledgeble industry columnist.
Then I turned 15.
That was years before Windows 95 was inflicted on an unsuspecting populace; before the company that employs both men decided to endorse the Microsoft product, in spite of every other competitor having major technical and interface advantages over the MS offering.
When asked why the publisher chose Microsoft as being 'superior:' It's what people are going to buy anyway.
I've long since relegated their rags to the shredder. (I don't just recycle it, because I don't want to run the risk that some innocent kid will open up one of their magazines. A child's innocence is a precious thing; they shoulnd't be viewing Microsoft propaganda at their age!)
So many cores. So many variations. So many significant architectural differences.
Seriously... when it it enough to be the Pentium5? I seriously doubt there is as much difference between the Pentium-3 and the IV (original P4) as there is between ANY other P4 cpu and this one.
Seriously... what's the deal?
Other than the 5-for-$5 jokes (Pentium 5 being a rather redundant name, after all...)
The concept of a leagal precedent is both wholly supported by the US Constitution (or at least by the Supreme Court, whom has the final say anyway), and is... suprise, suprise... based on 'common law.'
What intel needs to do is license IBM's Power5 technology and Sony/Toshiba/IBM's Cell to come up with a super Cell architecture.
Yeah, that makes sense. A company that is fully capable of designing a chip to compete with the Power5 and Cell should just give up and fab somebody else's design.
Might as well have GM, Ford, and Chrysler stop designing their own cars and just build the designs made by Toyota, Honda, and Nissan.
Nevermind the fact that Intel already suffered the pains of making a hybrid super-chip. Ah... the Itanium. A PA-RISC chip, a Pentium, and its own VLIW design all balled up into one wonderful package. A package so wonderful that it has garnered the title "Itanic."
I remember hearing about the PowerPC 615 -- It was supposed to be a PowerPC chip with a 486-66 core bolted on as well, so it would run everything at lightning speed (IBM actually had both the design rights and skill to do this.) It goes without saying that the 615 never saw the light of day.
No, blending the Power5 with the Cell would be doomed to failure from the start -technologically. first, economically second. History has already doomed such attempts in a horribly poetic way.
Remember, this is Sony, not Sega; they're good at the marketing thing.
Oddly enough, I seem to remember Sega being the company to beat before the days of the Saturn; in fact, I remember quite clearly when the Game Gear was impossible to get, because they were completely sold out.
And I've seen far more people playing games on a Nokia system (not n-gauge) than on any gameboy; I've seen people who wouldn't be caught dead playing on a console playing games on a Nokia phone. Nokia isn't inexperienced in the game market -- they honestly have higher penetration than Nintendo has with the GameBoy Advance. It's just that their purpose-built game-phone makes its owner look like an IDIOT when he is using it as a phone; this alone kept many from buying it.
I look at it this way: I've seen many portable systems by many talented companies.
Only one has ever been sucessful: Nintendo.
I'm no fanboy of either; the PSP has a large screen of supposedly great quality. Nintendo's answer to such a large screen was to have two smaller screens (one of them a touch screen). Both have wireless networking/multiplayer. I would give the durability nod to Nintendo -- both for its games and for the system itself. And durability is rather important for a portable system. The disc the PSP uses will definately reduce the playtime per charge. I've seen this show before, and it always goes very badly for Nintendo's competition.
But not impossible. possible means you can't do it under any circumstances. Very very hard to do means lots of time and money.
Yup; so much time and money I doubt even Microsoft would have the resources to toss at it. Espescially since they do have executives smart enough (and/or arrogant enough) to figure out that the time and money can be invested much more wisely on other projects. Yeah; they do seem to like destroying enemies -- but only when they can't beat them outright, and I don't see a reason for Microsoft can't do so. (I still prefer Linux; but Linux lacks many features that Windows has no problems with. Until that gap is bridged, Microsoft still has a niche.)
Well, with some Linux distros, you still have to answer a bunch of stupid questions before you can use it...
Just not with a phone.
I remember it being called 'The Debian APTitude test'. (Back before apt-rpm or yum).
Sadly, it's been so long since I've installed Debian from scratch (haven't had to; just too reliable/stable...) that I can't comment on if a Debian install requires fewer questions than it used to...
Patently false. It's just much harder to melt down with pebble beds. And pebble-bed is among the next-gen US reactors planned.
The 'impossiblity' of meltdown argument ignores entirely the problems with pebble bed reactors -- Namely that the 'pebbles' are made of, or contain high concentrations of graphite.
The reactors use noble gases (usually helium or argon) as the coolant. The reason: If the graphite pebbles are exposed to oxygen, they burst into flame; a well-known property of graphite that 'watchdog' groups pounce on when pebble-bed reactors are proposed. It's also a property that is downplayed by proponents of pebble-beds. Both sides are right: Graphite does and will burn quite easily. It's downplayable because the risk of combustion is small compared to the probability of failure in current reactors.
If the pebble-bed combustion goes on long enough (which isn't very long), the thermal expansion 'safety' feature of pebble bed reactors is lost -- the graphite has combusted to Carbon Dioxide, nixing the effect of thermal expansion, and the reactor still melts down. As a bonus, you get measurable amounts of highly-radioactive soot with sizeable levels of high-level nuclear waste in it as well.
So, like ALL fission reactors, the coolant system (which is used to provide the usable power) still must meet certain criteria to avoid meltdown. If there is a break in a coolant line (or any other situation) allowing for combustion, all bets are completely off. The primary advantage of pebble beds is this: as long as there are not conditions to invite combustion, it won't melt down, and in fact thermal expansion stops the reaction. But the caveat remains: As long as there are not conditions that invite combustion of the graphite the pebbles are made of.
There have been many (albeit small, 'experimental') pebble bed reactors being run by power companies, generating REAL power sent across consumer lines in the US for years; there have not been large pebble bed reactors in the US; they get around the 'moratorium' on new reactors because they are primarily experimental research reactors. (either Popular Mechanics or Popular Science had an extensive article on pebble beds in the past 24 months or so.)
China's distinction is they are going to be the first to make a large-scale pebble-bed. This is fairly reasonable, since most of the western world is highly allergic to building any fission power plants at all. (Thank you very much, Montgomery Burns...)
I'm yet to see a benchmark that proves me that the price per (volume/speed) - whatever unit of comparison you wanna use - justifies me buying a Mac.
This argument, while certainly valid from a point of view, is similar to saying that the price per horsepower is the only important criteria for buying a car. Fred Flintstone has you beat there, pal.
In computing terms, your brain has a better price per volume/speed than any computer, so why buy a computer?
I've yet to see a benchmark that proves that price per (volume/battery) justifies buying an iPod. iPods are more expensive than most portable music players. iPods also dominate the market because there are virtues other than cost per song stored on the device, and iPod customers sing their praises. iPods certainly weren't the first, they aren't the cheapest, nor are they the smallest, batteries aren't their strongest suit either... They are popular because they are easy to use, have stellar sound quality, and cost only a little more than the competition.
People who buy a Mac aren't buying it because it's the fastest ship in the fleet; they buy it because it's more luxurious than a Wintel box, or because it's able to do things that a PC currently doesn't do well, if at all. They choose macs because they are still more intuitive and easy than a Wintel box.
The majority of users I know of who complain about Macs are really only complaining about two things: Games, and 'upgradeability'. If it doesn't play their newest AMOR (Amusing Misuse of Resources -- apologies to the KDE team), the computer therefore 'sucks'. Then they complain about 'upgradeability.' That's an interesting argument, seeing that I can't 'upgrade' my PC without replacing the at least the Motherboard, CPU, and RAM. Yet PC's are more upgradeable? If I want a longer 'upgrade path' than sticking with AGP gives me, I'd have to also get an entirely new case and power supply for PCI-Express. Somehow this strikes me as little different than having to buy a whole new computer.
I don't own a Mac; but I've actually used them for real work(gasp). Once you get past the fact it isn't a Wintendo Entertainment System, Macs really are excellent machines, and I'll be glad to shell out the cash for a Mac the next time I 'Upgrade' my computer.
It's unimaginably trite to talk about code development time when that isn't the critical issue for scientific computing. It's a typically weak CS response who want to get their part of the work done as lazily as possible. There are cases where development time is inconsequential to the runtime, and scientific computing is one of them.
For serious numeric computation, you're going to be using a supercomputer; not some wimply workstation. You generally can't access them when you want, as long as you want, or as often as you want. If you have an hour on a supercomputer, you want to get the most work done in that hour as possible; you wait weeks or months just to get a timeslice; you don't waste ANY time doing development on a supercomputer -- development has to be finished before you even sit down at one of its terminals.
You carefully develop the code for weeks to get maximum throughput; then you finally get a timeslice scheduled when you are finished with development-- and then you have to wait until your timeslice is available, during which you can pore over the code again and again. Development time isn't even a factor; runtime is the thing that the developer is limited to. Many of these operations are so long and intensive you actually DO save time by spending weeks and/or months programming it in Fortran, rather than using something like MatLab, Mathematica, or Java.
Slashdot Mirror
Free software is most often licensed under copyright, but in order for free software to be truly free, as defined by FSF, its use has to be free of any exclusive privilege owned by a third party: "The freedom to run the program, for any purpose (freedom 0)."
;) But that doesn't mean you'd be violating any copyright laws in implementing AAC/FairPlay. Which means the implementer would still own the copyright; he just couldn't make use of it...
I don't believe I ever made the point of AAC (or FairPlay) being a 'Free' (FSF-Free) codec; but that it was an 'Open' codec. OpenGL is an 'Open' specification for a 3D graphics API. But OpenGL also has patented/non-Free extensions and components.
Of course, this argument comes down to whether an independant implementation of AAC is "Open Source" versus "Free Software". "Free Software" is tightly defined. "Open Source" is considerably less exact; just look at the early history of KDE/QT being 'Open Source' and not 'Free Software.' The OSI standars for being "Open Source" have yet to be accepted in the mainstream as the definitive definition; for now, 'Open' is typically understood to mean that 'the hood isn't welded shut.'
In any case; I view even a non-free standard (but open, published) standard like AAC/FairPlay as being considerably better than the entirely proprietary/closed WMA. AAC is in my view vastly better than WMA, if only because time is all that is required for it to become truly 'Free'.
You bring up Mongolia, but doesn't Dolby own counterpart patents in most developed countries that are markets for MPEG-4 audio technology (i.e. USA, Canada, EU members, and Japan)?
For all I know, Mongolia has a similar counterpart patent.
Perhaps for the encoder, but I'd imagine that licensees get decoder source code so that they can port it to whatever DSP a device uses (as you hint later). Remember that pocket-size players don't have an x86 monoculture the way consumer PCs do.
Not necessarily; just because most of Microsoft's work is x86/Windows, that doesn't mean they are above writing the DSP code for the licensee. Since I don't have the money to even find out what the terms are, it's really just conjecture at this point. But in any event, there would be some serious NDA verbage to ensure that you cannot share the information with others. (Since the point of being 'open' is that an NDA becomes pointless -- it's all published publically anyway; what's the point of keeping the secret?)
AAC implementations aren't as independent as you think; they're licensed by a division of Dolby on a royalty basis, which is incompatible with free software. Your argument may be valid in a decade and a half when the essential AAC patents expire, but it isn't now.
You seem to be confusing a patent with a copyright. It's quite possible to write an implementation of AAC from the spec, without violating any of Dolby's copyrights. On the other hand, you cannot use/sell/distribute that implementation (in the US, at least), without violating Dolby's patent. The FAAC group has already made a free implementation of AAC; it doesn't violate Dolby's copyrights at all. It may well violate Dolby's patents without a proper license, as you pointed out -- but that only matters within the borders of the United States.
And Dolby's patents are only valid in the countries that recognize them as valid. I doubt a coder in Mongolia needs to worry about violating Dolby's patents if s/he chooses to implement an AAC codec from the specs; at least if the coder doesn't intend to sell it in the United States.
In fact, Dolby's patents can't keep anyone from making an independant implementation of AAC (even American coders). All the patents do is make it illegal for someone to use or sell the implementation within the United States.
Although it's of dubious merit to have your own implementation of AAC if you can't use it... but that has nothing to do with it being possible.
WMA can be licensed by any compay that wants to use it.
Really? And they get source code? What specific techniques are used to encode audio into WMA? Oh yeah... they're not available. Microsoft just points you at their pre-compiled library and says 'use that.'
All Microsoft licenses is the ability to use the precompiled WMA libraries for commercial uses (and even then only on Windows and to a smaller extent Mac). You don't get the source code at all. The actual spec on how to create an implementation of WMA isn't available except under an expensive license and NDA (and even then only for making a hardware decoder). Everything that plays WMA in Linux is reusing the microsoft-compiled WMA libraries (for Windows).
AAC is a published standard, with hundreds of implementations of the codec.
WMA is an unpublished standard, and Microsoft's implementation or bust.
FairPlay is a layering of AES (an open encryption standard) inside AAC, and is documented; in spite of the leagalities involved, non-Apple implementations exist.
WMA's DRM is still not fully understood outside an NDA.
If the reverse engineer it, Apple sues them.
Um... and Microsoft won't sue if you reverse-engineer WMA (and its DRM) and use an unlicensed implementation (and therefore not pay Microsoft for it)? Just because you reverse engineer it doesn't really give Apple too much to sue over. It's that these compnaies were then using FairPlay without a license that spun the lawyers into action.
No company can license fairplay to use it.
Fair enough; but Apple is free to make the licensing decisions it chooses to.
But recall that there are more than a few Microsoft technologies that Microsoft refuses to license.
If this is your idea of an open format, then I'll take closed formats.
An open format is one that is disclosed and/or published openly. There is nothing about FairPlay that is not openly published on the web. The internals of AAC, internals of AES, internals of how Fairplay joins the two. However, try to find out enough about WMA to implement it yourself. Or the DRM that WMA uses. Go ahead and fail. There is nothing about the WMA codec that is freely available. It's not an open codec, not an open standard.
I don't know of any independant software implementations of WMA (with or without DRM).
FairPlay (without its DRM) is an AAC file -- a format that has more than a few independant implementations. And the actual wrapping of the AAC into a FairPlay formatted file is also easily handled.
An "open format" in this case is much like the OpenGL vs. Direct3D battle:
There is exactly one vendor of Direct3D: Microsoft. There are no independant implementations of D3D. The various versions of Wine that run D3D games on Linux simply convert the D3D call to an OpenGL call; it is not an implementation of D3D by any means.
However, there are literally hundreds (if not thousands) of independant implementations of OpenGL. Its spec is so complete that the Mesa project was able to start with the OpenGL spec and implement from scratch a conformant version of OpenGL. Apple has its own implementation. So does ATI, nVidia, SGI, Sun, Matrox, and Microsoft, among others. However (and this is the catch): Mesa can't call itself OpenGL due to OpenGL not licensing and/or certifying it. (Apple won't license FairPlay). That doesn't make Mesa (or OpenGL) any less functional, or any less 'open'.
The bottom line is this: I already have all the information I need to implement the FairPlay/AAC format; the spec is freely available. I can write a compliant FairPlay encoder, a decoder, and DRM protection facilities. And as long as I don't try to sell music using my implementation, Apple will be hard-pressed to touch me.
I do not have the ability or resources to even get the WMA specs to make a non-DRM'd implementation. I cannot write a compliant encoder, decoder, or DRM facilit
The biggest complaint I have is that I really don't own a physical copy of my music. I downloaded an album from iTMS. After listening to it quite a bit, I realized that I really don't like it. But there's nothing I can do about it. If I had purchased the CD, I could take it down to my local independent music store and sell them the CD and get a few bucks back. Thats the biggest problem I have with purchasing online. So if you are going to purchase an album online, better make damn sure you are going to like it.
Of course, the argument Apple would make here is that they offer the 30-second preview. Which is fine in many cases, but in others, the preview is from part of the song that is little more than dead air or is not at all representative of the song's overall feel.
That being said, yeah, it's not too wonderful that you can't take it to a CD reseller and trade it in/get cash back. But, the upside is that you aren't required to buy the entire album for one song. In the end, I think it comes out as a wash...
And the difference in quality between iTMS music and CD's? Well, none of the equipment I have is able to produce a noticible difference in quality; and this is with THX/Klipsch multimedia speakers and a THX sound card, or SennHeiser headphones; not audiophile grade, but well above consumer grade... I guess I'm just 'lucky' in that my ears are quite unable to tell the difference between compressed music and a CD. (Unless the codec *really* has issues, or is at a very low bitrate)
That being said, Steve Jobs also said that one thing the consumers surveyed insisted on is that they did not want a 'subscription' music service; but that they wanted to buy their music outright.
Frankly, as far as I'm concerned, it's just another subscription service, which is something consumers have repeatedly told the industry they would not accept. (Both with their pocketbooks and in surveys).
I honestly believe one of iTunes's greatest strengths is that they do not have a 'subscription' service-- you buy the song, or don't. You never have to keep paying to hear that song.
Not that I would use a WMA-based service even if it didn't try to force a subscription model on me; for all the ranting about FairPlay being a closed format, WMA is an even more closed format. Fairplay is the open AAC format, then uses AES encryption (another standard) for DRM; all the details are documented and freely available on the web (albeit not from Apple). WMA? It's not even published; the exact nature of it's non-DRM compression isn't available; in fact, more is known about its DRM than about its codec.
Except that it's unexplained and is generating more energy than is put into it. Using water.
More importantly, it hasn't been proven that it isn't a nuclear reaction. Nobody knows what is really happening, which is my entire point.
As 'fusion' yes.
That doesn't explain, however, what is happening, and where the 'extra' energy is coming from.
*If* it's actually fusion (and not just a weird chemical reaction) and *if* we can make it regularly reproducible, then Cold Fusion could essentially change the world.
Yeah. The real shame in my view is that the particular kind of reaction was largely dismissed by scientists. Those that did not dismiss it immediately found that it is reproducible; just not 100% of the time. And I still have yet to hear a reasonable explanation of what is actually happening in the Pons-Fleishmann cold fusion experiments. I have, on the other hand read from several sources that, while it may not be fusion, there is something happening that can't currently be explained to a decent degree of satisfaction.
If Pons & Fleishmann (and others involved) had not chosen to call it 'fusion', I wonder if the phenomenon would have been so thoroughly rejected as 'impossible'.
(You know -- like a high-energy particle bouncing back at its source as Rutherford discovered; I think his quote was something like it 'being as likely as shooting a bullet at tissue paper, and having the bullet bounce back.')
Pons & Fleishmann's 'cold fusion' is actually a rather dark chapter in the history of science -- where an observation was nearly universally deemed as 'impossible', even with independant groups reproducing nearly identical results (albeit only small a fraction of attempts were successful). But it had been given an unfortunate name, and was never really given a chance.
Hopefully this will further fuel investigation into exactly what is happening in the Pons-Fleishmann's experiment.
You gotta think back further. To circa 1992/3 or so.
Back when the PowerPC 615 was where it was at: A PowerPC core with a 486/66 bolted on. It would run both natively, so Apple would be able to run anything...
An argument I never really understood...
Yeah; parked illegally. That's a serious crime. It's the greatest threat to our country since flouridated water!
We should require all cars to be parked by a licenced and bonded valet, who is familiar with all state and federal parking regulations. It's simply too great a danger to the general public to be left to amateurs.
Same goes for shopping near a competitor's office. No more buying clothes next to Wall-Mart! You do not have the right to go shopping within 300 yards of a competitor's interests. Wake up and smell the restraining order!
Seriously, flying off the handle like that over a parking offense is really quite silly. And suggesting that any citizen of the United States can't go shopping in a store next to a lobbying organization's office is absolutely absurd.
Senator Frist is pretty far from a decent human being; but by your own admission the only illegal thing he did was park his vechile incorrectly. There isn't a law on the books saying where anybody can or cannot shop for shoes.
So, boot/tow the Senator's car, give him a parking ticket, and get on with life.
Honestly, if I knew beyond any doubt that the worst thing an individual would do while in public office is park illegally, he'd win my vote hands down.
Not that the Senator in question would actually qualify; by all accounts he is of most questionable character. But claiming he should be punished severely for parking illegally is rather astonishing. It was his only crime, and one that rarely even results in anything more than a modest fine (actually quite trivial at a Senator's salary).
And for "Free Speech Zones" -- I've seen first hand what happens when 'free speech' gets out of hand; the result is frequently bloody, with neither side being particularly innocent. A simple fact is that there are few things that incite more strong feelings than politics or religion; it's also a simple fact that not everybody will see things the same way.
And in spite of what many would like to believe, humanity hasn't grown past the intolerance of millenia long past. It would be nice in a civilized society if people would agree to disagree; but society is rarely so civil. Both 'liberal' and 'conservative' are typically mutually immature, mutually arrogant, and mutually foolish. The observation that each cannot imagine the other as having honorable goals speaks volumes.
So, when such an event happens, the local city governments have a choice: Either try to organize how people express their viewpoints, or to allow chaos to reign. In an organized structure, it's easier to physically protect those of opposing viewpoints from assaulting one another, given the limited resources of said city's police force. In the chaotic structure, it's virtually assured that there will be some form of assault over differing points of view.
Which leads to why the cities make these 'free speech zones:' If a disagreement does come to blows, the city can stand in a court and say that they did everything they reasonably could to protect the physical being of an individual, while providing the right of free speech to said citizens. You cannot claim that freedom of speech is infringed upon because you can't say what you want, where you want, when you want to. It's already been ruled upon that freedom of speech does not extend to yelling "Fire!" in a crowded theatre, and that doctrine is applied to 'free speech zones' -- that the government can place minimal restrictions on the right to free speech in order to protecect the physical well-being of its citizens.
There's also a difference between the right to speak your mind, and the right to force someone to hear it.
The courts have not (and probably will never) rule against the cities for protecting its citizens from each other at a political rally. However, there are several counts of when a court has ruled against a city for not physically protecting its citizens from each other at a political rally.
I agree completely with the arguments against using Gentoo for performance reasons.
... I still run debian on a fileserver, though). Distros like SuSE, and Fedora (even when using things like APT-RPM, yum, etc.) drove me nuts. There were dependancies that were *completely* unnecessary being installed. And the 'RPM Hell'; drove me INSANE. Debian is a serious step up from SuSE and Fedora in that respect, but Gentoo puts Debian to shame.
It certainly isn't why I run Gentoo.
I run it because I like the way it works. I've recently taken a look at some of the other distros (SuSE, Fedora,
And I can honestly say -- I don't see where people have fits about having to compile everything. After the initial install, you can have any updates compiling at a well-niced priority. Unless I'm trying to play a CPU-intensive game (like Doom3, UT2004, etc.), I never even notice a performance difference. (I will admit this: my development work is done completely over SSH to another box, so emerging isn't slowing my 'real' compile jobs).
Even stranger is that most of the time, things don't even take long to compile. Huge apps/groups of apps like KDE, OpenOffice, and to a lesser extent, GNOME are an exception. And for those, I start the build when I leave work for the day, and it's done when I come back.
I don't use Gentoo for performance reasons. I use it because I hate the dependancy hell involved with pre-compiled binaries. I don't have to deal with it when using Gentoo; it's that simple.
And people whining about compile times -- seriously... unless you want to play Doom 3 while compiling KDE-- you've got issues, because compiling the apps isn't an issue at all.
Light bulbs, incandescent or fluorescent, running off of house current "flash" 120 times per second.
Yeah, but incandescent don't have as much as an impulse to the flash. This is mostly due to the fact that they produce light as side-effect of their heat, and the wire doesn't cool down anywhere near as quickly as the next peak in the current. As a result, incandescent bulbs have a much smaller delta between the 'bright' and 'dark' parts of the cycle. Turn the power off on an incandescent bulb, and it has a perceptible dimming after the power is gone. Flourescent bulbs just go dark instantly.
I guess I should have waited until there was an editorial summarizing the study then linked to that and write a summary which mis-states what the study says.
Well, obviously you're journalistic credentials haven't fallen to the levels of David Coursey. A columnist whose perception of reality is solely based on whom gives him the most advertising dollars / freebies.
Somewhat similar to John Dvorak: I used to respect his opinion as a knowledgeble industry columnist.
Then I turned 15.
That was years before Windows 95 was inflicted on an unsuspecting populace; before the company that employs both men decided to endorse the Microsoft product, in spite of every other competitor having major technical and interface advantages over the MS offering.
When asked why the publisher chose Microsoft as being 'superior:' It's what people are going to buy anyway.
I've long since relegated their rags to the shredder. (I don't just recycle it, because I don't want to run the risk that some innocent kid will open up one of their magazines. A child's innocence is a precious thing; they shoulnd't be viewing Microsoft propaganda at their age!)
The devil you know is better than the devil you don't know.
That's one such (American) English equivalent. (No idea if it's used much outside North America...)
Just a question...
So many versions of the Pentium4.
So many cores. So many variations. So many significant architectural differences.
Seriously... when it it enough to be the Pentium5? I seriously doubt there is as much difference between the Pentium-3 and the IV (original P4) as there is between ANY other P4 cpu and this one.
Seriously... what's the deal?
Other than the 5-for-$5 jokes (Pentium 5 being a rather redundant name, after all...)
Don't forget, most cab drivers speak more than one language (no matter how badly).
Try cracking a book.
... suprise, suprise... based on 'common law.'
The concept of a leagal precedent is both wholly supported by the US Constitution (or at least by the Supreme Court, whom has the final say anyway), and is
What intel needs to do is license IBM's Power5 technology and Sony/Toshiba/IBM's Cell to come up with a super Cell architecture.
Yeah, that makes sense. A company that is fully capable of designing a chip to compete with the Power5 and Cell should just give up and fab somebody else's design.
Might as well have GM, Ford, and Chrysler stop designing their own cars and just build the designs made by Toyota, Honda, and Nissan.
Nevermind the fact that Intel already suffered the pains of making a hybrid super-chip. Ah... the Itanium. A PA-RISC chip, a Pentium, and its own VLIW design all balled up into one wonderful package. A package so wonderful that it has garnered the title "Itanic."
I remember hearing about the PowerPC 615 -- It was supposed to be a PowerPC chip with a 486-66 core bolted on as well, so it would run everything at lightning speed (IBM actually had both the design rights and skill to do this.) It goes without saying that the 615 never saw the light of day.
No, blending the Power5 with the Cell would be doomed to failure from the start -technologically. first, economically second. History has already doomed such attempts in a horribly poetic way.
Remember, this is Sony, not Sega; they're good at the marketing thing.
Oddly enough, I seem to remember Sega being the company to beat before the days of the Saturn; in fact, I remember quite clearly when the Game Gear was impossible to get, because they were completely sold out.
And I've seen far more people playing games on a Nokia system (not n-gauge) than on any gameboy; I've seen people who wouldn't be caught dead playing on a console playing games on a Nokia phone. Nokia isn't inexperienced in the game market -- they honestly have higher penetration than Nintendo has with the GameBoy Advance. It's just that their purpose-built game-phone makes its owner look like an IDIOT when he is using it as a phone; this alone kept many from buying it.
I look at it this way: I've seen many portable systems by many talented companies.
Only one has ever been sucessful: Nintendo.
I'm no fanboy of either; the PSP has a large screen of supposedly great quality. Nintendo's answer to such a large screen was to have two smaller screens (one of them a touch screen). Both have wireless networking/multiplayer. I would give the durability nod to Nintendo -- both for its games and for the system itself. And durability is rather important for a portable system. The disc the PSP uses will definately reduce the playtime per charge. I've seen this show before, and it always goes very badly for Nintendo's competition.
But not impossible. possible means you can't do it under any circumstances. Very very hard to do means lots of time and money.
Yup; so much time and money I doubt even Microsoft would have the resources to toss at it. Espescially since they do have executives smart enough (and/or arrogant enough) to figure out that the time and money can be invested much more wisely on other projects. Yeah; they do seem to like destroying enemies -- but only when they can't beat them outright, and I don't see a reason for Microsoft can't do so. (I still prefer Linux; but Linux lacks many features that Windows has no problems with. Until that gap is bridged, Microsoft still has a niche.)
Well, with some Linux distros, you still have to answer a bunch of stupid questions before you can use it...
Just not with a phone.
I remember it being called 'The Debian APTitude test'. (Back before apt-rpm or yum).
Sadly, it's been so long since I've installed Debian from scratch (haven't had to; just too reliable/stable...) that I can't comment on if a Debian install requires fewer questions than it used to...
A pebble bed reactor cannot melt down.
Patently false. It's just much harder to melt down with pebble beds. And pebble-bed is among the next-gen US reactors planned.
The 'impossiblity' of meltdown argument ignores entirely the problems with pebble bed reactors -- Namely that the 'pebbles' are made of, or contain high concentrations of graphite.
The reactors use noble gases (usually helium or argon) as the coolant. The reason: If the graphite pebbles are exposed to oxygen, they burst into flame; a well-known property of graphite that 'watchdog' groups pounce on when pebble-bed reactors are proposed. It's also a property that is downplayed by proponents of pebble-beds. Both sides are right: Graphite does and will burn quite easily. It's downplayable because the risk of combustion is small compared to the probability of failure in current reactors.
If the pebble-bed combustion goes on long enough (which isn't very long), the thermal expansion 'safety' feature of pebble bed reactors is lost -- the graphite has combusted to Carbon Dioxide, nixing the effect of thermal expansion, and the reactor still melts down. As a bonus, you get measurable amounts of highly-radioactive soot with sizeable levels of high-level nuclear waste in it as well.
So, like ALL fission reactors, the coolant system (which is used to provide the usable power) still must meet certain criteria to avoid meltdown. If there is a break in a coolant line (or any other situation) allowing for combustion, all bets are completely off. The primary advantage of pebble beds is this: as long as there are not conditions to invite combustion, it won't melt down, and in fact thermal expansion stops the reaction. But the caveat remains: As long as there are not conditions that invite combustion of the graphite the pebbles are made of.
There have been many (albeit small, 'experimental') pebble bed reactors being run by power companies, generating REAL power sent across consumer lines in the US for years; there have not been large pebble bed reactors in the US; they get around the 'moratorium' on new reactors because they are primarily experimental research reactors. (either Popular Mechanics or Popular Science had an extensive article on pebble beds in the past 24 months or so.)
China's distinction is they are going to be the first to make a large-scale pebble-bed. This is fairly reasonable, since most of the western world is highly allergic to building any fission power plants at all. (Thank you very much, Montgomery Burns...)
I'm yet to see a benchmark that proves me that the price per (volume/speed) - whatever unit of comparison you wanna use - justifies me buying a Mac.
This argument, while certainly valid from a point of view, is similar to saying that the price per horsepower is the only important criteria for buying a car. Fred Flintstone has you beat there, pal.
In computing terms, your brain has a better price per volume/speed than any computer, so why buy a computer?
I've yet to see a benchmark that proves that price per (volume/battery) justifies buying an iPod. iPods are more expensive than most portable music players. iPods also dominate the market because there are virtues other than cost per song stored on the device, and iPod customers sing their praises. iPods certainly weren't the first, they aren't the cheapest, nor are they the smallest, batteries aren't their strongest suit either... They are popular because they are easy to use, have stellar sound quality, and cost only a little more than the competition.
People who buy a Mac aren't buying it because it's the fastest ship in the fleet; they buy it because it's more luxurious than a Wintel box, or because it's able to do things that a PC currently doesn't do well, if at all. They choose macs because they are still more intuitive and easy than a Wintel box.
The majority of users I know of who complain about Macs are really only complaining about two things: Games, and 'upgradeability'. If it doesn't play their newest AMOR (Amusing Misuse of Resources -- apologies to the KDE team), the computer therefore 'sucks'. Then they complain about 'upgradeability.' That's an interesting argument, seeing that I can't 'upgrade' my PC without replacing the at least the Motherboard, CPU, and RAM. Yet PC's are more upgradeable? If I want a longer 'upgrade path' than sticking with AGP gives me, I'd have to also get an entirely new case and power supply for PCI-Express. Somehow this strikes me as little different than having to buy a whole new computer.
I don't own a Mac; but I've actually used them for real work(gasp). Once you get past the fact it isn't a Wintendo Entertainment System, Macs really are excellent machines, and I'll be glad to shell out the cash for a Mac the next time I 'Upgrade' my computer.
It's unimaginably trite to talk about code development time when that isn't the critical issue for scientific computing. It's a typically weak CS response who want to get their part of the work done as lazily as possible. There are cases where development time is inconsequential to the runtime, and scientific computing is one of them.
For serious numeric computation, you're going to be using a supercomputer; not some wimply workstation. You generally can't access them when you want, as long as you want, or as often as you want. If you have an hour on a supercomputer, you want to get the most work done in that hour as possible; you wait weeks or months just to get a timeslice; you don't waste ANY time doing development on a supercomputer -- development has to be finished before you even sit down at one of its terminals.
You carefully develop the code for weeks to get maximum throughput; then you finally get a timeslice scheduled when you are finished with development-- and then you have to wait until your timeslice is available, during which you can pore over the code again and again. Development time isn't even a factor; runtime is the thing that the developer is limited to. Many of these operations are so long and intensive you actually DO save time by spending weeks and/or months programming it in Fortran, rather than using something like MatLab, Mathematica, or Java.