Amazon is a private company, what the heck you care what platform they use to implement their software. They are not directly obliged to use any particular techonologi but one that suites their needs/interests/tastes etc .... 

You sound in your article like they must choose java .... which is buggy as hell to develop on ... but you probably never do any software so you never got to deal with buggy File implementation, JDialogs etc .... Java is filled with bugs, and any non-trivial bug is never solved; still after 10 years they didn't fixed filefilters on awt ... 

[You're banned Anon. Ed.]

I suspect the reason that Amazon is going the secret route is that the Amazon music servers are trivially simple to hack, and that they don't want the cat to get out of the bag early.

At any rate, if Amazon's security model holds up, there is no downside to releasing the code. It should be trivial to translate their efforts into Java, Qt, Perl, or anything else you can think of (my personal preference is Commodore 64 BASIC, but oh well).

Maybe it's the whole "if we reveal our intellectual property, we won't be able to get a patent on it" boogeyman. Well, bollocks to Amazon. A fanatical Amazon MP3-buying fanboi will be able to reverse the code by wiresharking the process.

I just don't see this as "worth it". It certainly is not as burning an issue as the BBC iPlayer, for instance (doesn't Apple have a copyright on every instance of i[Word]?).

That's a lot of punctuation, at the end.

Fernando, common, there are so many alternatives better than Java (e.g. QT), from which you could find good apps much more stable and diverse than the few buggy ones that you mentioned (Azureus, argh). And, unlike Java, they're fully open-sourced and you as a developer wouldn't need to struggle with something as crappy as Eclipse.

Thanks for that comment, Stephen Brooks.

I dislike sites that need "special software" to download files. 

It's perfectly possible to have a web-based login system for secure/paid-for downloads, so why don't these sites do that, obviating the need for any additional development on any platform that has an operational browser.

In fact I dislike such sites so much I input fake user information to them if they ask for it as part of the registration process... Information like address that they don't need if I'm just downloading a file to my computer.