Using self-signed certs for these sorts of things is more secure.. theoretically. In practice, I'm not so sure, the grandparent has a good point. By using self-signed certs you rule out MITM, but that's only when the following assumptions are also true:

* The device has the self-signed cert installed.
* The device recognized the aforementioned certificate as the *only* valid certificate for a given website.

In practice, many (most?) phones do not support installing arbitrary certificates. I've never encountered even a desktop browser that allows me to configure a cert as being the only valid certificate for a given website. For these reasons, using self-signed certificates actually trains your user to always click 'Yes' when they see a certificate warning, and they were already clueless about security to begin with. Even if the device supports installing self-signed certificates, without enforcement one can hack into a CA, create a false certificate, poison the user's DNS and redirect traffic to his MITM website.

You can easily have that many images loaded if you have many tabs open. The modern web is full of images. Or just browse a few photo galleries in different tabs.

> There is no excuse for a web browser process to hit the GB mark, none.

So if all the images that are open in your web browsers all add up to 2 GB of uncompressed pixel data then the browser still shouldn't use hit the GB mark? I want my computer to be magical too.

No, it's about the cost of the bugs vs the cost of fixing the bugs. Suppose that a smartphone costs $400 in its current state. It has a few bugs here and there, not always noticeable, and when they show up they're annoying, but in general the device works fine. Now suppose that fixing those bugs and preventing new bugs from occurring costs the company $700 million in additional developer expenses (training, hiring ever better developers, improving Q&A) etc which causes the price of the device to jump to $1400. Would you buy the $400 device and take the bugs for granted, or would you buy the $1400?

It's faster than 3.6.

Uh, yeah, so? My point is that the filesystem installed on the device is what the OS sees and your post says nothing to refute that assertion. Your flash device is not going to emulate an FAT filesystem when there is in fact an ext2 filesystem on it.

He isn't. He's complaining about people's apparent false sense of entitlement. Avoiding GPL is perfectly acceptable. Avoiding GPL *and complaining that GPL has harmed you* displays a false sense of entitlement.

I believe his point was that BSD'ed code promotes standards by allowing anybody to easily incorporate the code. Suppose that the original TCP/IP stack was GPL'ed. Others could still have written their own TCP/IP stack, but would they have bothered to do that or would they just have invented their own proprietary standards instead of bothering with TCP/IP at all? BSD'ing the code makes TCP/IP the path of least resistance.

I don't necessarily agree with this point of view but I can see the reasoning.

Heck, "mission critical" is overrated. Some ATMs run Windows!

Your web browser speaks HTTP. However an operating system reading the contents of a mobile device involves treating it as a block device. There is no commonly-used high-level file access protocol over USB. Unless the device can emulate an FAT filesystem over a block device interface (is this even possible without huge amounts of resources?) the device's filesystem has to be formatted as FAT.

The alternative is to create your own file access protocol and force the user to install a custom application that speaks this protocol, e.g. how iTunes interfaces with the iPad/iPhone. This really, really sucks, you now have to install iTunes everywhere.