That's what PPAP and APQP are dealing with. In practice, those processes are used to ensure that the vendors understand what is expected of their parts in their intended end-use.

They should be rightly shunned for acting on fantasies, as it were. What they believe here is a figment of a collective imagination of a whole bunch of people, unfortunate enough to be misled to believe such nonsense. Just sayin'.

In the particular case here, I don't quite get how an engine stall can cause loss of control? Going 50+mph, you don't care about power steering. Loss of brake assist might be an issue, and loss of ABS too if that was the case, but even then, I think in this one the driver shares 50% of the blame.

I've had hydraulic brakes fail, I've had ABS fail (with no MIL coming on), and in both cases some forethought and pre-training has immensely helped me. Yes, in both cases those were close calls, but through no fault of mine. No, nobody had me do it. I simply figured: I must know what to do in such cases, since there's never enough time to think. For example, if my accelerator-by-wire car ever had an UA on a road with some clearance ahead of me, I know that it wouldn't cause anyone any harm - I know how to handle it, and I routinely test my service brakes to ensure that they will stop the car in such a scenario even if I were to forget to switch to neutral (or the switching didn't have effect, as it may). The ECU doesn't have a brake throttle override, BTW.

Yes. I've found that there's no way to use it without knowing this. It's as simple as that, and I don't consider it a deficiency of git at all. You're supposed to understand the tools that you use. The things you speak of are not implementation details, they are part of the semantics exposed directly to the user. They make the whole thing work and useful.

Yes, the command line syntax is abhorrent, but then I almost never use it. SmartGit/Hg has won me over.

The initial checkout from svn doesn't clone the entire history. With git, a shallow clone (--depth 1) doesn't either. Git's smart http transport is bound to be faster than svn, since compression can be applied across multiple files, and if there are similar files in the repository, their common chunks will always be factored out by design. With svn, it's not guaranteed and depends on how file copies/moves were managed.

The shallow clone have fixed this, mostly. So far, I find no downsides to shallow clones, other than the loss of history.

First of all, I don't think that it's sane to use svn or git without a graphical front-end. That said, I find that git-svn works way better than svn itself. Subversion merging starts looking downright silly once you see how git does it. And I was one of those who actually argued that svn's merging is "good enough". Eventually I convinced myself, through experimentation, that it's nowhere near good enough. To me, subversion is a way of centrally storing git repos without having to configure server-side git, and also a way of doing partial checkouts when I don't need the whole thing.

Fuck yes!

Man, that MS Word source should be perfect for an allnighter port to modern winapi :)

Yeah, that's what I always tell people: the scale of things in the U.S. is quite hard to understand without actually being here. U.S. municipal budgets are like country budgets in eastern Europe. NYC tracks Poland, etc.

That's why I'm always smiling when people smirk at me when I tell them that a lot of firmware that I develop runs on devices with 64kb or less total memory (code + data, since it's Harvard). I'm not using frameworks with tens of thousands of lines of code, I can actually tell exactly how the whole thing works, top-to-bottom, and it can be explained in detail in about 150 pages of prose. A lot of that code is written in functional style (side-effect-free functions) and can be formally proven to agree with the specs. When you've got a lot more than that, it gets very, very costly to have formal proofs of anything, much less even formal specifications (formal not as in bureaucratic, formal as in using the language of mathematics).

Does that mean anything? An operating system by definition is totally aware of the files and processes that you run on it. So, the best I can make out of what you're saying, is just gobbledygook and wakalixes. Nobody forces you to use cloud backup of anything with Mavericks. I have no evidence that the app update mechanism is leaking contents of private files, but if you have some evidence to the contrary, I'm sure you could get some limelight time for it.

I think that you'll find that Windows 8.1 will perform like shit on those machines... Not only because all of the rendering will be software-based.

Care to elaborate? (seriously)

Again, what's wrong with it? You don't want to be seen, don't go out in public, it's that easy.