Not everyone listens to their CDs 100s of times.
I don't think i have have even come close for any CD, i might make it to 20 - 40 for a good album. Think how long 4500 min (100 plays of the album with 3min tracks) is if you believe that is the norm. No average artist will be listened to the required amount to achieve the breakeven price.

Streaming eliminates the required investment to listen to something else so unless you get advertised on the front page and your whole album is worth listening to you wont anywhere near the same number of plays as a CD does.

do you not see the repartition there, that method produces terrible passwords. how is pressing the same key exactly 4 times in a row at speed?

They are getting strong enough not to low fruit and fall to a mass hash cracking but someone only has to observe you typing that in from a distance once, observe your fingers not move, and will rearrange word list to favor small character spaces.

PS when you put this method on the internet you can no longer use it unless you never reuse you user-name.

??
There are password managers that use ssl apis to do all this for your now, keyrings will be no different.

Just the one ever, your servers (fingerprint i think its called, the one you get the unknown host warnings for). That only if you want to do syncing yourself though.

What are you on about, you would sync it over an ssh/ssl direct connection and make sure the (cert) hash matches so you know you connected to the right device. Just get them on the same subnet and click sync on both devices (with a password you remember the first time). If you trust a third party it's a non issue to begin with.

If there is any every push to remove the need for remembering passwords there will have software to handle these problems. I am trying not to say how to do it because someone will create a system that the general public can make scene of and it will probably be better anyway.

Once you get over the need to have a password you can do far more tricky things.
Ideally each device would have its own key so it got compromised you could just revoke the key.

https://en.wikipedia.org/wiki/Public-key_cryptography
Understand that and reread my post (the authentication does not rely on stored hashes on the server). I am not saying that there is anything wrong with keepass's software.

Keepass can't fix this, its not their problem (though password manager evangelists are going to hold back changing anything). This is about if your computer is going to remember something for you a password is terrible compared to some alternatives.
I guess the post title was misleading.

Its not really about the length of the key about how the key is used as authentication.
Have a look at ssh logins.

Its not that having you computer generate passwords for you is bad idea by its self (you can do things such as regenerate a password without user interaction), it's that you may as well do it properly.

They are a stopgap with backwards compatibility while everyone moves to something else, if a computer is storing your password and not a private key someone somewhere is doing something wrong.
Your computer is equally as good at "remembering" a key as a password and the key is considerably stronger security. You can still password you private keys (with a keyring).
A key is stronger and if the server is compromised they can't use your public key to login.

This is New Zealand its not going to be the maximum fine or probably even an order of magnitude less. They will likely not recoup the costs to get it there.
Last fines RIANZ tried to push were based on figures based on the damage it cause. The only way for to them get 3 MP3 downloads into the 3 figure mark was to try to argue that they were shared 90 times (would like to know how they got this figure) and then triple it (at least this is what they tried to do to the last distort person before it was thrown out).

As long as it stays tied to real damages NZs fines will not make it to the 500 dollar mark, covering the 250 or so in court fees to get it there in the first place. Internet is too expensive here to seed.

The "Broken Window Fallacy" is about the overall economy, having windows to repair is good for the window repairer.

This only true in a very limited micro processor. Working with 16 bit you might have to be careful, I guess. Once you are working with 32bit integers this is not an issue and if it is use 64. There is not really any benefit to using floating point here just pick a good "unit", say 1/32", to work from.
Most modern processors are as fast at multiplication and addition and some are faster in cases.

Its not the desktop environment its just that browsers even with a few tabs open would like about 300Mb to function properly.
Chose a lighter weight browser or get aggressive with cutting back your browsers memory usage.
Both Debian and Arch's minimum requirements are 64Mb and you should be able to get it using less.

It was rushed but my point was the people who wrote your "general purpose machine learning code" accounted for different various types of input by using what i would guess to be an adaptive filter or you accidentally did by using the GPML code. I guess you could argue whether or not you are the sole designer of your AI.

Try this for what it really means:
http://en.wikipedia.org/wiki/Strong_AI
As far as I can tell from skimming it no one has any solid proof that it can exist.

Your using general purpose machine learning code it was designed to allow you to do this. It probably has a decent adaptive filter integrated into it, this would allow it adjust, try feeding ROT13 or something that does not have real integer readings. The person who wrote the code programed it, you are just changing the numbers in the memory inside it.

Please stop using SF handwaving, it does not help the credibility or your first paragraph. There is a large gap between "fuzzy logic adaptive filters" and actually rising up against humans. Only seen the movie.

Could any of these AI's do the others job or anything other than what it it wasn't designed to do. They are just making decisions based on code written by humans because humans aren’t good enough to make in the given time.

Work in a more general case than what the designer had accounted (accidentally or purposely) for.
If you change the AIs inputs and outputs in a way not accounted for it will break and the AIs goal is not going to magically change to something different what it set to.

How is that skynet? Its missing the most important parts and this could happen and has already been avoided with humans.

They are talking about AGI (something no one has manage to demonstrate to any degree) not some AI failing to achieve what it was designed to do. They are two very different ideas and latter could cause some damage.