John Gruber at Daring Fireball writes about why people get angry in situations like Apple disabling the SIM unlocking and third-party application hacks for the iPhone with a software update. It is a 'misguided mindset', he says, to 'to expect support after taking unsupported actions.' But he has come to understand that this mindset is due to the subtleties and motivations of software limitations as opposed to hardware. The article is quite well written, as usual, and I recommend that you read it rather than only my attempt at a summary. But I think he misses a key aspect of this phenomenon. He says:

"But when you do these things [take unsupported actions], you are assuming responsibility for any adverse effects caused by them, now or in the future."

"Now or in the future." It's not just about hardware vs. software. It's that with the advent of software and networked devices, the product you buy is not necessarily the product you will always have. Particularly with devices like the iPhone that are designed to be more software- than hardware-driven, the product's nature can change to a great extent after you buy it, through software updates. Of course the updates are usually for the better, and you can almost always opt out of them, but in my experience, you can only hold out for so long, and it's quite often a mixed bag of improvements and degradations.

You can't blame Apple for disabling the hacks with the software update, and you can't even necessarily blame them for completely breaking some people's phones in the process. There might be a legitimate reason that wasn't their fault. But you have to understand that people are going to be suspicious, and have emotional reactions, when they take an unsupported action like this and it works fine, and then a software update comes along and breaks their product. It seems to violate an unspoken tenet of traditional ownership. And it's that much worse when the company has a business incentive not only to disable such unsupported actions, but to discourage people from even trying to find new ways to enable them. (In the case of third-party applications, you could argue that it would be quite short-sighted of Apple to discourage it, because all the rumors point to them enabling it in the not-too-distant future. But this enabling will still be controlled and on their terms in some way that might otherwise encourage hacking.)

Gruber writes about the risk one takes when "replacing the current version of a kernel extension or other system component with an older version from a previous release of Mac OS X because you read on MacFixIt that it works." Of course there is risk involved there. But quite often it does work, and in some cases it's necessary for the user. And when you have to consider not only the current risk but the risk that at some point in the future, the company that made your product will decide it doesn't like what you're doing and break your product with a software update, I think it will have a chilling effect on a type of experimentation that can be very beneficial to users.