I'm fairly well-versed in interactive use of MATLAB and ipython/numpy/scipy. I also have an HP-50g next to my monitor, with the USB cable plugged in so I don't need to keep replacing the batteries every few months.

Generally speaking,

I use ipython if :
- the problem has a significant non-mathematical component (string processing or some other general-purpose programming language task), or
- I want good looking 3d graphs

I use MATLAB if:
- I need to plot a bunch of things (2d), or
- the problem is almost entirely linear algebra

I use the 50g if:
- the problem is short enough to quickly write on paper, or
- the programming only requires a handful or short routines, or
- neither of the above programs is running, and the problem won't take more than a minute or two to type in and run on the 50g

There's a strong inverse relationship between the power of a calculating tool and it's usability. For really short stuff with only one or two variables, the ergonomically-ideal 15c is much much quicker than the clumsy notation of numpy. The 50g, with it's math-optimized keyboard and GUI, is often much faster than either desktop option because the bottleneck is the typing process, not the computation process.

So, it doesn't have a physical keyboard, it's slower for all but the most complex operations that a 12c or 15c can do, it gets 1/1000th the battery life, and it has a monthly fee. Sounds great.

It's really a perfect illustration of the downside to device convergence. Your smartphone is jack of all trades, master of none. The 15c, on the other hand, has very little room for improvement: the screen contrast, battery life, and execution speed could be improved, and the price could be lower. The rest of it is basically perfect, and couldn't be improved without seriously compromising another feature. (eg. it could support more functions, but only by compromising the usability significantly.)

Nowadays, you can actually force ICC to emit code that will use up to SSSE3 on AMD CPUs, but only if you don't use runtime code-path selection. (More specifically, you have to tell ICC that the least-common-denominator code path should use SSSE3, which defeats the purpose of runtime code-path selection. ICC will always choose the slowest available code path for an AMD CPU, but you can prevent it from including a non-SSE code path.)

For the hardcore gamers who don't have unlimited budgets, it might be logical to buy the cheapest CPU that won't bottleneck your games, and pair it with the fastest graphics cards you can afford. Particularly if your games can use the GPU for the physics engine, you might not need even AMD's high-end CPUs to keep a pair of NVidia cards busy.

If a company like Intel is still making big mistakes like the Sandy Bridge chipset problem, what reason is there to believe that there will still be a correlation between design flaw frequency and experience if everyone under discussion has at least a few years and more than one product cycle under their belt? Do you really think that anybody who was at Intel during the 1980s is helping them prevent design flaws?

If all they get is a molten pool of radioactive lava, they'll be much better off than Chernobyl: there, the graphite moderator of the reactor caught fire and helped spread fallout far and wide. A pool of radioactive lava is only problematic while it's still too hot to build a new containment building around.

Which nVidia chipsets are you talking about? The only NV chipset problem I remember was the defective packaging of the 8600M, which was eventually recalled.

It's quite possible that sending frivolous takedowns could result in some kind of fraud charges for the copyright holder and FRCP rule 11 sanctions for the law firm, but the DMCA itself does almost nothing to dissuade frivolous notices.

According to the law at issue, the only portion of a DMCA takedown notice that is under penalty of perjury is that the person filing it is authorized to act on behalf of the copyright owner alleging infringement.

For a DMCA takedown counter-notice, the poster needs to assert under penalty of perjury that they have a good faith belief that the takedown was a mistake or misidentification. The lack of a requirement that the party issuing the takedown make a similar statement of belief under penalty of perjury is the real bullshit here, as it violated the principle of equal protection under law.

They don't have any immunity in the first place, as this notice is not about infringement (it's about circumvention devices), so it isn't covered by the DMCA's safe harbor provision.

It's not a real takedown. It's a normal cease-and-desist letter, and even if it were a real takedown notice, it's only perjury if the person submitting it isn't authorized to act on behalf of the person claiming ownership of a work being infringed. Neither the copyright holder or his lawyer are guilty of perjury if the copyright holder lies to his lawyer (or is mistaken) and thus causes a frivolous takedown notice to be sent.

Hit the brakes, and tell the computer in the car behind you to do the same. Also, the car behind you was using radar-assisted cruise control, so it actually wasn't tailgating you.

Obstacles like deer, weather, falling trees, etc. are far less dangerous than the other people on the road. If you can get the rest of them to drive safely and react reliably and quickly, then things like deer are much easier to handle.

The 6900s were originally intended to be build on a 32nm process and be released as the 7x00s, but then TSMC cancelled their 32nm node, so AMD axed half the features of the 7x00s and moved them to the 40nm process that they first used on the 4700s, and used on all the 5000s and 6000s. (The earlier 6000s had also been planned to be 32nm, but ADM switched back to 40nm for cost reasons before the 32nm node got canceled.)

Long story short, the 6900s are being manufactured on a process that is quite mature and that they have a lot of experience with, since it is roughly the 4th generation product on that process. It should be no surprise that yields would be very high out of the gate.

My setup: A regular mouse on the right for gaming and occasional casual use, and a tablet for anything requiring precision or lots of GUI interaction (and for some RTS games). That way, I'm not developing muscle memory that is in any way backwards, but I still get to take advantage of the better dexterity in my left hand.

I do quite like the layout of laptops like the MacBook Pro, with a big touchpad centered under the keyboard, so that you can use it with either hand. Even the multitouch gestures added by jitouch are ambidextrous, so I can switch between hands easily, depending on what side of the keyboard I need to use more.

You may not have to pay money to get the fuel, but it is limited, so efficiency matters.