You can't /. Google, and it looks like you can't Google /. either. :)

The problem isn't the system requirements; infact, I'm able to run the latest Android 2.3.7 on an outdated HTC Magic (G1) just fine (the last official update was 2.1) The problem is: 1) Manufacturers want to give users one more reason to buy a new phone - that it runs the latest Android version; hence they stop supporting older devices. It also makes sense to them, financially. 2) Manufacturers like to customize the deivce and tack on their own UI and integrate crapware into the system. Sometimes this doesn't work well, as there just isn't enough space on the ROM for all that extra crap (case in point - the reason for the delay in releasing Gingerbread for the HTC Desire). Even if the device supported all that crap, the manufacturers still have to do extra testing to check if everything works fine with the new Android codebase, which introduces delays. 3) Carriers. If manufacturer bloat isn't enough, the carries have to add in their own bloat, or worse, more lock-ins/restrictions. And then you run into the same issues as above. It's easier to just say, "sorry, your device is too old". Android isn't the problem here, the carriers and manufacturers are.

I too have an HTC Desire, but I still continue to use my HTC TyTN because of all the handy buttons, and the QWERTY keypad. I just don't understand why phone makers these days seem to think buttons are un-necessary. :(

Just checked the troubleshooting section, and looks like this extension requires inbound and outbound UDP and XMPP. Oh well, good luck trying to use it behind any corporate network. Without the ability to work behind firewalls, I doubt the Chrome Remote Desktop extension will be very successful. I, for one, won't be recommending this extension to anyone.

I tried connecting to/from behind a squid proxy server at uni, and it doesn't work. However, TeamViewer works just fine. Guess this extension is using a non-standard port/protocol which is being blocked at my uni..

Except that you can't just run a random attachment without giving it execute rights first. Also, shell scripts by default open with a text editor (when launched from within Firefox). Say you decide to package your binary malware in an installer - which format would you choose? .deb? .rpm? .txz? What about all the dependencies needed (to do anything useful), if you add them how large would the file be then? You could go for shared libraries, but there's no guarantee that the target's system has the required versions installed. Heck, it's a challenge to install even legit third-party programs (those which aren't available on your distro's repo) due to dependency hell. Besides, the installer on most distros come with sufficient warnings/prompts before the package can be installed. Thanks to the fragmentation within the Linux community, it won't be feasible for a malware writer to bother packaging for different distros / versions / configurations. I'm not saying that it can't be done, I'm saying that if it is done, it has to be a very specific, targeted attack.

Could you explain how it would cost a significant amount of time/money just to deploy Firefox across the machines? If you have proper app-deployment infrastructure in place, it shouldn't take more than a few clicks. Then send out a note saying "you can use Firefox for all the non-intranet stuff, just remember that we don't officially support it". And you're all set. Firefox auto-updates itself so that's taken care of as well. Eventually one could start using IE Tab+Firefox and remove IE (the icon). IE Tab can automatically open pre-defined websites in an IE instance within Firefox itself, thus the end user only has to deal with one browser (icon). When they visit the IE6-only site, it opens up in IE tab, but all other sites would work in Firefox. Problem solved.

Firefox+NoScript is no good if the OS itself has bugs. I'd recommend Opera instead. Case in point: The Animated Cursor vulnerability a couple of years ago. Firefox allowed the exploit while Opera didn't.

Not to mention MARQUEE