I only stopped using a 1200MHz P3 a couple of years ago. It was nice because it was the newest machine I owned that you could get deterministic CPU timing results out of. Building LLVM on it was a pain (over an hour), but 4 of them would be quite reasonable. My NAS / media centre box is currently using an AMD E-350, which is a dual-core 1.6GHz part - I think this would probably be faster. If it has a well-supported GPU and a decent collection of SATA ports, I might be tempted...

Yes they do. But hardly anyone makes ARM chips that draw 5W (nVidia is the only one that comes to mind). Most are 1-2W under load, with the slower ones closer to 250mW.

We're just about to open source our MIPS IV implementation (I'll post something to Slashdot when it's done - lots of legal paperwork for creating a community interest company to coordinate it and so on). It's written in Bluespec, which is a high-level HDL and very easy to modify (we've been setting an exercise to replace the branch predictor[1] in it to students for a couple of years now and they're able to in a couple of hours and get the required prediction rates).

MIPS IV is nice, because it's a 64-bit ISA that's over 20 years old (the magic number for patents). FreeBSD 10 runs on it out of the box with the BERI kernel config on the Altera DE4 boards and in simulation and 10.1 should include a kernel config for the NetFPGA 10G board. These boards are pretty expensive, but we have a couple of configurations that will let it run on smaller FPGAs. Removing the FPU makes it a lot smaller and you can also build a microcontroller variant (simple static branch predictor, no MMU) that's even smaller. The simulator is slow, but just about useable (it takes about an hour to boot to single user mode, but it's enough for testing).

It's only in the last couple of years that FPGAs have become interesting for this kind of thing. There are a few high-level HDLs appearing, because hardware is sufficiently complex that the traditional approach of throwing it all away and starting again every CPU revision is increasingly impractical. The devices themselves are now fast enough that they're useable for prototyping and getting a reasonable feel for behaviour. We can get 100-200MHz with 4 cores in a single FPGA with the latest generation - not competitive with an ASIC, but fast enough that you can actually use them. I gave a demo that ended up being more compelling than I expected because I was showing people some things running on the UART console and I'd left the network cable connected so the screen kept being spammed with messages about invalid ssh connection attempts. Nothing I was doing said 'this is a real computer' quite as much as people on the Internet trying to attack it...

The CPU on any of these boards is probably adequate for most tasks. The big difference between Intel and most ARM SoCs for open source development? The GPU drivers. Intel releases documentation and code for their drivers. Most ARM SoC makers release blob drivers that work with a specific windowing system and kernel version. Trying to get X.org running on one that only provides Android drivers, or trying to get any non-Linux OS running on them with acceleration is painful.

Is the point to give the users something useful? I assumed it was to give the devs something to debug. Asking the user to transcribe a kernel stack trace is not likely to happen, but a QR code that encodes a bit of text saying 'please email to crash@linux.org' followed by the stack trace would mean that you'd actually get something vaguely useful in some cases, rather than the user just shrugging and power cycling the box.

And this is actually a fairly small amount of money. The UK government spends some number of hundreds of millions a year in license fees for Microsoft. I'd love to see that money spent in the UK, rather than being shipped to the US. It would do a lot more than all of the government's other initiatives to improve the state of the tech industry in the UK.

The stock price has dropped quite a bit in the last few days, but it's still above the average for the last year and a lot higher than it was this time last year. It's hard to draw intelligent conclusions from the Facebook stock price, it's better to use it as a source of entropy for your random number generator...

I actually think that it would be an easier sell if you went for the wider context, where the state-recognised mechanism did not in any way imply a sexual relationship. If two guys who are just housemates want to have the same legal protection, it's fine. It shouldn't offend the Christian right, because it's a purely legal arrangement that is not in any way like a marriage. And then, if the same legal framework were used for marriages, then, well, that's just a simplification of the legal code.

Okay, so how do you do it? I'm running 4.4, with the stock calendar app. I don't have a Google Calendar account sync'd with it, I want to just subscribe to a calendar. I've found lots of forum posts saying you can't - you have to add it to your Google Calendar and then let it sync indirectly. So, since you can do it, how about sharing the mechanism? To make it more concrete, here's a calendar URL that I want to subscribe to from my phone: webcal://www.talks.cam.ac.uk/show/ics/6330

Step by step instructions please, since apparently it's easy...

I don't see why the same legal instrument couldn't apply to both traditional married couples and hippy communes. The same set of issues apply: shared assets, being treated as a single entity for tax purposes, and having automatic power of attorney in case of incapacitation.

I'd have thought Amazon would be a natural choice, especially for funding the Android port of Firefox. Amazon has its own Android distribution (and app store) that has none of the Google additions. In recent versions, Google has stopped developing the AOSP web browser in favour of Chrome, leaving a gap for an independent web browser that Amazon can bundle.

These economies work differently for a government. When you're spending am amount measured in hundreds of millions a year on software, then you don't complain that an open source program misses some features you need, you just ask for them and either your supplier provides them or you get a new supplier. You don't have to worry that you can't open MS document formats correctly, because you are the one defining what the interchange format is. If other companies buy MS products and they can't open the documents that you send them, or you can't open the ones that they send you, then that's their problem, not yours.

I'd have said FreeBSD not Linux, but the question still remains. For an answer, how about £5.5M? To put that in perspective, the annual budget of the FreeBSD Foundation is about a tenth that, which funds new development work, subsidises some conferences and so on. The UK government is paying £5.5 just for security updates. I can point you at several companies who'd be happy to provide extended support for a particular branch of FreeBSD for a fraction of that cost and an even bigger number who'd do it for Linux. £5.5M, even including overheads, will pay for 50 developers working full time. Let's assume that there's a lot of overheads, shareholder profits, and so on and call it 20. Do you really think it takes 20 developers to backport security fixes for Windows? Oh, and if they were running FreeBSD then I can point them at a couple of UK companies who would happily take half that money, provide the same support, and keep the money in the local economy. Want to take a guess about where Microsoft will be spending that £5.5M?

Sounds like the USA sucks for phones. In the UK, I'm on a pay-as-you-go plan. I top up £10 every few months and pay 3p/min for calls (about 5), 2p/text (about 3.3) and 1p/MB for data (about 1.6). Most of the time I'm near WiFi if I want to do something online, so I rarely turn on the mobile data for my phone, and when I do I rarely use more than 5-10MB. Aren't there any carriers in the US that do pay-as-you-go data, so if you don't use any you don't pay for any?

Almost all of the apps I have installed on my Android phone are from F-Droid. I tried setting it up without a Google account at all, but there was one app (irritatingly, my Internet banking one) that required a Play Store account. I also have the Amazon AppStore installed for its free app of the day thing (it was NeoCal a few days ago, which is a really nice calculator app, but I use a calculator so rarely that I'd probably never have bought it).

The biggest limitation with iOS for me though is it's lack of some decent equivalent of OSMAnd - a map app that lets me download entire countries worth of vector maps and can do offline navigation, so I won't run up huge bills using it when abroad.