When you refer to this hypothetical steambox as open, what makes you assume that it will be? If it is custom hardware there is no reason to assume that it will be any more open than competing consoles. Linux on the inside doesn't guarantee root access, and if they drop bios for a signed bootloader it would be difficult to get into. The DRM in the system will be derived from steam, so I would assume that online activation will be mandatory. The piracy rate could actually be lower than the currently existing consoles as none of them enforce mandatory online activiation. It could also completely kill the second-hand game market for this device, so the economics of the market would quite new. As you wrote ealier, it would be up to Value to prove that gamers would buy into this market, although I assume the studios would love it.

Was about to post this. Also steal a part of ringbring's business model and pick up other supplies on route: popcorn, chocolate, soda, pizza etc.

Probably more showers... But they are less fun.

(that I'm too lazy to read the article for) is does this mean an end to passwords, and the beginning of proper identity management in the browser. I would love to authenticate myself once to the browser, and then have it negotiate identities with each website.

We have motorised desks at work. There is a pair of push switches on the front and they raise and lower maybe 1.5m (3-4ft). As long as you have fairly long cables on everything these make the transition very easy. It does make quite a difference as you can stretch your legs for an hour or two and then switch back. Oddly enough I've ended up in a weird slouched position with the chair reclined and the desk close to the floor, meh, the broken ergonomics are probably more to do with me than the office.

Fair enough, I think that I skipped over your penultimate paragraph. I'm not the poster that you were replying to though...

Yes, "among". Mankind taking a step off of this rock for the first time is a huge deal. It is as big as the other achievements that you list. Although Armstrong did not achieve it alone, most of the rest of your list are not solo endeavours either.

No. You are taking an extremely superficial view of the problems and claiming that we know the solutions. If you take your points together then they claim that we can build a viable closed eco-system. The experiments that have been conducted so far show that we cannot. We still do not know how to engineer a closed-loop that can survive for more than a few months and every experiment so far has had to shut down or risk killing the participants.

Sure we know how to build bunkers - do we know how to do large scale construction in a hard vacuum? Can we build a fission reactor large enough for a small city, launch it by rocket and land it on another orbital body? Or is it so easy to assemble it over there using our well-developed space construction skills?

It must be some utopian blend of crack in that pipe that you are smoking. Nobody is claiming that we can't learn how to do these things given the will and the experience, but to claim that they are already solved problems is just ridiculous.

Indeed, just because binary was good enough for the first few millenia of civilisation did not stop the invention of octal in the 19th century. It did not stop the widespread adoption of decimal across the world in the '60s (apart from america), and hexadecimal is currently being established as an international standard. Who knows? By the mid-century we may have progressed as far as a base-25 system.

What do you mean focus on playlists rather than albums? The only thing that I use playlists for in iTunes is overlapping sets of genres, I use albums the rest of the time.

Oh dear, you don't seem to grasp the point so well. If I replace a single processor that performs a task with a set of parallel processors then I cannot claim the power is reduced just because the time is shorter. The energy consumed depends on both the time and the power use of the set of processors. Given the discussion that followed in that thread it is quite clear that understanding the distinction would have added something.

Your logic is flawed:

the less time the machine spends active the more time it can spend in idle which quite obviously results in less heat and power consumption.

The caveat is that power consumption must not increase during the shorter period of operation for this to be true. In the case that we are offloading work from the CPU to the GPU then it is not only the length of time that is important, but the relative power consumption of the two devices while drawing the text. It may be that offloading the text rendering to the GPU reduces power consumption (Microsoft claim that in the article), but it does not logically follow from the claim that the operation can be performed in less time.

If you can do that in five minutes then are you not being a little harsh in not providing the info to the nouveau project?

Sounds Soggy, Dude.

What would calculating the theoretical peak tell them about the (real) sustained performance?

Partitioning the problem in chunks that can be distributed to the nodes in the cluster adds overhead. Assembling the finished results does the same. It is kind of hard to predict what this over will be as it depends on the interconnect. In this case they used 100Mb/s ethernet, but there was contention from running NFS over the same network. Building it and measuring it is the only way to find out what kind of performance you really get.