What the heck, Apple (HDCP; it has begun)

[Dominus Atheos]

Alright, conceded. Maybe Apple not suing isn't an indicator this isn't RIAA pressure, but why does EMI let them distribute music without DRM?

EMI suffered the worst of slumping music sales and hence were more open to breaking out of their traditional business model than the other labels.

Is your tin-foil hat on too tight? Your explanation for why Apple still releases songs with proprietary drm is that it's an industry wide conspiracy to drive Apple out of business? You must really be chugging that kool-aid if you honestly believe that the only way Apple would behave in any way but as a perfect angel that has only the concerns of it's customers at the fore-front of every decision. Instead of, you know, money.

Because Apple put's that music in a non-standard format to try to make sure it will playback on as few 3rd party devices as possible. Have you even been paying attention to the bottom half of the debate?

Oooh yeah, they're really killing interoperability. Making programmers deal with a file extension. What a burden!

Once again, iTunes will not play a .gobbledygook file even though it's just an mp3 file, and portable audio devices that can play .aac or .mp4 will not play .m4a.

Thanks for the lesson, but what does that have to do with the debate? The point is that there are mp3 players out there that have hardware capable of decoding and playing back AAC or mp4 files that won't playback these files because they don't recognize the extension, and I suspect Apple did that deliberately.

My point is that a filename extension is not part of a file format, and Apple has used the .m4a extension since well before it was offering DRM-free music on iTunes. If it was a move to kill interoperability, it wasn't very effective. Your argument that changing a filename extension constitutes changing a file format is nonsensical.

Non-sequitur. These won't playback on non-apple devices without either the manufacturer or user going through an extra step, and most software and hardware will only look at the extension when trying to open a file.

Also, I'm confused. Here you seem to be explaining why file extensions don't matter, but below you are explaining why the are very important. Aren't theses two arguments you're making self-contradictory?

Don't give me that bullshit, we both know any media container will have file headers that will have information about the streams in it. If you actually had to parse the entire file before playback, it would take several minutes to load a several gig hi-def mp4, and I don't even want to think about how long it would take to load a blu-ray disk.

And what parses those file headers? Magic gnomes?

By declaring "There is only audio in here", you enable a decoding program to skip at least one phase of header parsing, the phase that determines what streams are contained in the file and (presumably) lazily loads the needed decoders. There could be tons of different performance scenarios where such a hint would be beneficial, especially at the implementation level. Since you know there's only one stream type in the header, the parser can allocate space on the stack rather than hitting the heap, which may be required for an arbitrary number of streams.

You can also keep the audio decoder cached in memory and not have to hit the disk for, say, a decoder plug-in for video, which might happen when QuickTime hits something with a .mp4 extension. It may do some work upfront to make the common case faster, but this would be a detriment to the uncommon case.

At the end of the day, the existence of this filename extension enables programs to make more assumptions. And being able to make more assumptions means less code complexity and more opportunity for performance optimization. If you had any experience writing software, you'd know this. But you don't, so instead, you just look like an idiot.

What the hell are you babbling about? File headers are parsed the exact same way regardless of what it says, just like AAC is decoded the exact same way regardless of whether or not the audio is just speech and no music, or just piano sounds and nothing else. And even if that weren't true, an mp4 container can contain any audio format in the world, with the standard specifying support for AAC, MP3, Apple Lossless, MP2 MPEG-1 Audio Layer 1, CELP (speech), TwinVQ (very low bitrates), and SAOL (MIDI). And if you really want an extension that absolutely for sure contains AAC, just use the .aac extension which is part of the standard, at least according to the Library of Congress.

[Sarevok]

What the hell are you babbling about? File headers are parsed the exact same way regardless of what it says, just like AAC is decoded the exact same way regardless of whether or not the audio is just speech and no music, or just piano sounds and nothing else. And even if that weren't true, an mp4 container can contain any audio format in the world, with the standard specifying support for AAC, MP3, Apple Lossless, MP2 MPEG-1 Audio Layer 1, CELP (speech), TwinVQ (very low bitrates), and SAOL (MIDI). And if you really want an extension that absolutely for sure contains AAC, just use the .aac extension which is part of the standard, at least according to the Library of Congress.

You should read what Durandal wrote. When a file is being read a program does not magically know what the sequence bytes represent. Some formats are kind enough of to include obvious hints that are apparent once first few bytes had been read. Others are not so. Either way it saves both the programmer and the computer a lot of time to have a handy file extension so it knows for sure what it is dealing with.

[Durandal]

Is your tin-foil hat on too tight? Your explanation for why Apple still releases songs with proprietary drm is that it's an industry wide conspiracy to drive Apple out of business?

My explanation was that the labels (which act as a cartel on many occasions) were unhappy with Apple's power to dictate song prices to them due to the success of the iTunes Music Store. So they tried to give competitors a leg up on the iTunes store by licensing DRM-free music to competitors to cut down on Apple's bargaining power. EMI was struggling, so they decided to try their luck with DRM-free music, and it looks to have worked out for them.

Is this really that hard to believe?

You must really be chugging that kool-aid if you honestly believe that the only way Apple would behave in any way but as a perfect angel that has only the concerns of it's customers at the fore-front of every decision. Instead of, you know, money.

Once again, iTunes will not play a .gobbledygook file even though it's just an mp3 file, and portable audio devices that can play .aac or .mp4 will not play .m4a.

Oh please. The Zune supports .m4a, as do recent Creative players. This is not some gigantic burden on player manufacturers.

Non-sequitur. These won't playback on non-apple devices without either the manufacturer or user going through an extra step, and most software and hardware will only look at the extension when trying to open a file.

That's a failing on the software's part. There isn't some sort of deep reverse-engineering required here. It's a god damn file extension.

Also, I'm confused. Here you seem to be explaining why file extensions don't matter, but below you are explaining why the are very important. Aren't theses two arguments you're making self-contradictory?

Filename extensions do not have any effect on the bits of the file. The entire filename is a filesystem construct. But many programs use them as a guide when determining how to open a file.

What the hell are you babbling about? File headers are parsed the exact same way regardless of what it says, just like AAC is decoded the exact same way regardless of whether or not the audio is just speech and no music, or just piano sounds and nothing else. And even if that weren't true, an mp4 container can contain any audio format in the world, with the standard specifying support for AAC, MP3, Apple Lossless, MP2 MPEG-1 Audio Layer 1, CELP (speech), TwinVQ (very low bitrates), and SAOL (MIDI). And if you really want an extension that absolutely for sure contains AAC, just use the .aac extension which is part of the standard, at least according to the Library of Congress.

You're trying to comment authoritatively on things you don't understand. Go back and re-read what I wrote, this time without thinking about it as something you must argue against because Apple is evil and that I'm an Apple fanboy who thinks Apple can do no wrong.

[Dominus Atheos]

Is your tin-foil hat on too tight? Your explanation for why Apple still releases songs with proprietary drm is that it's an industry wide conspiracy to drive Apple out of business?

My explanation was that the labels (which act as a cartel on many occasions) were unhappy with Apple's power to dictate song prices to them due to the success of the iTunes Music Store. So they tried to give competitors a leg up on the iTunes store by licensing DRM-free music to competitors to cut down on Apple's bargaining power. EMI was struggling, so they decided to try their luck with DRM-free music, and it looks to have worked out for them.

Is this really that hard to believe?

It's defiantly plausible, but if the two explanations are "industry-wide conspiracy" or "money", money is much more likely. Pretend this is a different company that's having this problem. Let's say Bridgestone convinces Ford to make a policy that from now on you will only be able to buy Bridgestone tires as replacements for your Ford, and Ford get's a cut of every sale Bridgestone makes as part of this new policy. Would you say this deal was made mostly because Bridgestone want to put Goodyear out of business, or because Bridgestone thinks they can make more money due to this partnership? Obviously it's possible Bridgestone did it to put Goodyear out of business, but isn't a much more likely explanation that they just think they can make more money then if they didn't screw over Ford's customers?

Once again, iTunes will not play a .gobbledygook file even though it's just an mp3 file, and portable audio devices that can play .aac or .mp4 will not play .m4a.

Oh please. The Zune supports .m4a, as do recent Creative players. This is not some gigantic burden on player manufacturers.

Recent being the key word there. All players made before Apple announced the iTunes Plus store would not be able to play these.

Non-sequitur. These won't playback on non-apple devices without either the manufacturer or user going through an extra step, and most software and hardware will only look at the extension when trying to open a file.

That's a failing on the software's part. There isn't some sort of deep reverse-engineering required here. It's a god damn file extension.

Do you have a goddamn reading impairment? As I've explained many times, iTunes won't play files with extensions it isn't programed to know it supports! Regardless of how simple it would be to implement, it's not an industry standard to try to parse an unknown file by using the profiles of every format the player knows it supports. And Apple knew this then it decided to release it's Plus tracks in that extension.

Also, I'm confused. Here you seem to be explaining why file extensions don't matter, but below you are explaining why the are very important. Aren't theses two arguments you're making self-contradictory?

Filename extensions do not have any effect on the bits of the file. The entire filename is a filesystem construct. But many programs use them as a guide when determining how to open a file.

So should good programs pay attention to file extensions, or ignore them?

What the hell are you babbling about? File headers are parsed the exact same way regardless of what it says, just like AAC is decoded the exact same way regardless of whether or not the audio is just speech and no music, or just piano sounds and nothing else. And even if that weren't true, an mp4 container can contain any audio format in the world, with the standard specifying support for AAC, MP3, Apple Lossless, MP2 MPEG-1 Audio Layer 1, CELP (speech), TwinVQ (very low bitrates), and SAOL (MIDI). And if you really want an extension that absolutely for sure contains AAC, just use the .aac extension which is part of the standard, at least according to the Library of Congress.

You're trying to comment authoritatively on things you don't understand. Go back and re-read what I wrote, this time without thinking about it as something you must argue against because Apple is evil and that I'm an Apple fanboy who thinks Apple can do no wrong.

Okay, I've done that, now what? You are trying to argue that Apple's decision to release all of it's Plus store's music as .m4a files was due to technical considerations and that breaking compatibility with other players was an unfortunate and unwanted side effect. I'm arguing that it was due to money. At least for a little while Apple was the only game in town if you wanted to play those files, and this probably helped to drive iPod sales just like Fairplay. "Technical Considerations" like the ones you've described are a lot less crazy then "industry wide conspiracy", and I probably would have conceeded if it wasn't for one other thing that, looking over the debate, I haven't brought up (and probably contributes to why you think I'm being an idiot for not getting this): .m4v. I meant to mention it a while ago, but it never came up. If this were an isolated one-time incident, I'd be willing to accept your explanation for why Apple had to make up a file extension instead of following the standard. But it's not, it's part of a string of decisions at Apple that all seem to have the effect of making sure that any media you get from Apple can only be played on Apple's hardware, and any hardware you get from Apple won't play media you get from places besides Apple. Like I said, those reasons you gave for why Apple was justified in making up the m4a file extension may be good ones, but even you can't possibly argue that they had a good reason for making up the m4v extension. Those files are mp4 files exactly adhering to the mpeg standard and exactly what ISO had in mind when the created the mp4 container. The only thing making up that extension did was break compatibility with every existing mp4 player.

[Durandal]

It's defiantly plausible, but if the two explanations are "industry-wide conspiracy" or "money", money is much more likely.

Of course it has to do with money. The labels' money. And my argument is that it probably costs Apple more to maintain a DRM infrastructure for its music. After all, that DRM system costs time and money in terms of support calls, engineering effort, server maintenance, etc. Steve Jobs pretty much said it's more trouble than it's worth for music.

Pretend this is a different company that's having this problem. Let's say Bridgestone convinces Ford to make a policy that from now on you will only be able to buy Bridgestone tires as replacements for your Ford, and Ford get's a cut of every sale Bridgestone makes as part of this new policy. Would you say this deal was made mostly because Bridgestone want to put Goodyear out of business, or because Bridgestone thinks they can make more money due to this partnership? Obviously it's possible Bridgestone did it to put Goodyear out of business, but isn't a much more likely explanation that they just think they can make more money then if they didn't screw over Ford's customers?

This isn't about putting Apple out of business. It's about diminishing their ability to dictate price. Your analogy is missing several parallels and is just plain out of whack in some places. The record company wants to keep iTunes around because it makes them a shitload of money. But they're pissed off that Apple has the power to tell them "No" when they want to raise the price above $0.99 per song. The friction between Apple and the music labels on this matter is very well documented, as is Apple's insistence to control price in general. Remember when NBC walked away from iTunes and took all their content with them?

Recent being the key word there. All players made before Apple announced the iTunes Plus store would not be able to play these.

Ever since the days of QuickTime 6 (which was released in 2002), Apple has used .m4a in QuickTime.

Do you have a goddamn reading impairment? As I've explained many times, iTunes won't play files with extensions it isn't programed to know it supports! Regardless of how simple it would be to implement, it's not an industry standard to try to parse an unknown file by using the profiles of every format the player knows it supports. And Apple knew this then it decided to release it's Plus tracks in that extension.

See above. That extension isn't something new. Again, if it was an attempt to prohibit interoperability, it made no sense at all.

So should good programs pay attention to file extensions, or ignore them?

Many can (and do) check file headers for magic numbers in the file header if they encounter a file that they don't immediately recognize by an extension. And Mac OS X has mechanisms for identifying file types aside from the extension, such as UTIs.

Okay, I've done that, now what? You are trying to argue that Apple's decision to release all of it's Plus store's music as .m4a files was due to technical considerations and that breaking compatibility with other players was an unfortunate and unwanted side effect. I'm arguing that it was due to money.

Which is ridiculous. There's no money to be made. Why would they bother releasing un-DRMed music in the first place? Because they had the dastardly clever scheme of using a different file extension? Wow, the Microsoft guys will really be thrown by that! Put another way, that extension didn't stop their biggest competitor from supporting their files.

At least for a little while Apple was the only game in town if you wanted to play those files, and this probably helped to drive iPod sales just like Fairplay. "Technical Considerations" like the ones you've described are a lot less crazy then "industry wide conspiracy", and I probably would have conceeded if it wasn't for one other thing that, looking over the debate,

That's exactly what I'm arguing, and I'm not about to hand-hold you through the intricacies of software design and optimization. I've detailed the benefits that come from that type of optimization, and you haven't addressed any of them specifically -- because you don't know what you're talking about. You've just thrown up your hands and said, "THAT DURN SOUNDS COMPLICATED!!!! YER WRONG!!"

What you are probably right about is that this extension was used for the benefit of the iPod. Embedded devices need all the help they can get. Not having to drag an entire MPEG-4 parser into memory would be nice under embedded systems constraints. In fact, cell phone manufacturers developed a subset of the MPEG-4 file format for their cell phones with the .3gp extension. No one screamed bloody murder at them. In fact, just about every MPEG-4 player out there supports playing back .3gp.

Also, just for grins...

The existence of two different file extensions for naming audio-only MP4 files has been a source of confusion among users and multimedia playback software. Since MPEG-4 Part 14 is a container format, MPEG-4 files may contain any number of audio, video, and even subtitle streams, making it impossible to determine the type of streams in an MPEG-4 file based on its filename extension alone. In response, Apple Inc. started using and popularizing the .m4a file extension.

So depending on how reliable you think Wikipedia is, you can at least walk away knowing that I'm not the only person with this crazy idea that being able to make more assumptions about input is faster and less complex. At the end of the day, being able to assume exactly one stream in the file may very well save the parser a few trips to the heap and just keep certain data structures on the stack.

I haven't brought up (and probably contributes to why you think I'm being an idiot for not getting this): .m4v. I meant to mention it a while ago, but it never came up. If this were an isolated one-time incident, I'd be willing to accept your explanation for why Apple had to make up a file extension instead of following the standard.

.m4v files downloaded from the iTunes Music Store are wrapped in DRM. Hence they are not MPEG-4 files. In the case that they're not wrapped in DRM, the same arguments apply as before.

Hell, I also wouldn't be surprised if the use of custom file extensions was (at least in part) so that the icons generated for the file on Windows (which relies exclusively on filename extensions) could reflect the content of the file and use the same icons that Mac OS X does.