Good thing that the definition of "evil" is sooooo malleable.

Historical note: x86 is a bastadised rip-off of the PDP11 instruction set.

And as with most technological descendents, the folks who did the job botched it. Incredibly obtuse instruction decoding, special instructions that do five things at a time (most of which are not useful), and horribly slow to interrupt and restore.

The PDP11 was built as a "hardware Fortran machine" ie one instruction represents one Fort[r]an instruction as far as was achievable in 1970.

Uh, not really. The PDP-11 was designed as a general-purpose ISA, used as much for assembly code as Fortran. In addition, it hosted four OSes (RT-11, RSX-11/M, and RSTS/E from DEC and UNIX from an odd place called Bell Labs). The different OS'es used different tools. A lot of RT-11 code was used for industrial control and was done in Assembler (did some of that), RSX-11/M was their mainline OS for applications and was programmed in COBOL (the implementation here sort of sucked) or FORTRAN (a pretty brilliant implementation) or Assembler, and RSTS/E was an odd duck that had a BASIC interpreter. UNIX had C. The best thing about programming on the -11 (besides the nice, relatively orthogonal ISA) was the FORTRAN automatic overlay feature. It let you bundle code into overlay segments that were automatically swapped in when routines in the module were called. A performance killer when used improperly, it was the only way I could fit a FORTRAN program that took 320K on an IBM\360 into the PDP-11's 64K.

C is (just one) PDP11 assembly language!

I don't think I'd go that far. There were many things (conditional branches on overflow, control of interrupts/traps, computed gotos) that, although accessible via assembler, could not be easily done in C. That's why today you still have assembly modules and/or use of inline assembly in UNIX code.

The VAX instruction set was an attempt to achieve a higher level machine code, which worked quite well - most VAX assembly instructions are actually function calls to application specific microcode.

As were most instructions in those days. As for "worked quite well"? Well, there was that whole RISC/CISC thing going on and, you know what? RISC sort of won the technical war - it may be papered over with an ugly CISC instruction set on the inside, but internally, it's all condensed onto execution on a mostly RISC core.

X86 was a poor ISA when the first 8086 chips were made (but good, given hardware capabilities at the time). That was about 40 years ago. MIPS and Sparc (and ARM) are all better than x86.

Well, yeah. They have the benefit of hindsight and much less self-inflicted baggage. On the other hand, that baggage has kept Intel in the game while they try to catch up to ARM in power consumption.

So, according to your model, we'll all be able to drink hydrofluoric acid next year? I'll let you go first.

How does a single company make bad decision after bad decision so persistently?

A conundrum for the ages to be sure, but my humble opinion?

Ahem... A company that makes bad decision after bad decision does not understand the difference between a good decision and a bad decision.

Do I win a prize?

'nuff said...

Does da Vinci get paid every time someone looks at the Mona Lisa?

No, but the Louvre does.

Ownership hath its rights in our capitalist society.

No, I'm blaming Pandora for a shitty algorithm that always eventually directs people towards the most popular music (as defined by their overall population) regardless of what the individual indicates as his/her preferences.. Start out the day listening to punk and you'll soon enough end up with Katy Perry anyway. Bayesian logic and priors work well, except when they don't.

There's no such thing as "fairness" - it's a fairy tale concept that causes humans far too much suffering.

As does attempting to say that because things aren't fair, it's fine if everything stays as unfair as it currently is, or becomes more unfair. In general, the better we all behave, the better off we all are. If you deny this, you are an enemy of civilization.

Well, as they say, "Past Performance is Not Necessarily Indicative of Future Results". When you decide that sharks are what you're going to swim with, you don't get much choice in the bites that go along with them.

--- THIS IS AN AUTOMATIC MESSAGE, DO NOT RESPOND! ---

A level 2 Potential Lack of Mememetic Acknowledgement has been detected. Lack of attribution is very dangerous and could start a copyright infringement war and ultimately involve the use of lawyers. Please consider reading comments with your IP (Intellectual Property) detector engaged to avoid further chance of litigation.

Bill Nye is a poor man's Mr Wizard.

No he's not! He's a poor man's Professor Proton!

Christ almighty! Get your facts straight, Karmashack!

Yes! Everyone knows Clojure is the most awesomeness...

I suspect the latter was more scientifically based than the former.

But I bet the former would have made more money, as there are more people interested in astrology than horse racing.

You reap what you sew.

I wish. I can use a Singer (or a Pfaff, or...), but I've never reaped anything I've sewn. On the other hand, I have planted several gardens and, in that case, I've always reaped what I've sown.

Is people's illiteracy increasing, or is it just me? Sew and sow are such different words, I can't actually see a well-educated person making these sorts of mistakes (unless you transposed the e and o in a right-left hand confusion thing, in which case your proofreading needs work).

Well, I had quantum mechanics and solid state theory, logic design, processor design, and various courses on memory architecture. I was a bit skimpy on some of the database/compiler work, so I took courses in those topic as well. My education was a bit odd, though, starting with a BS CompE, followed by a failed MSEE (mainly because my marriage at that time was falling apart and a job offer seemed really good by that point) and thirty year gap between my first MS attempt and my second (successful) attempt at an MSCS (and yes, I did a thesis the second time). But, yes, I do believe I understand just about everything in a modern processor/software stack. Or, at least, with enough study of the particular system, I do believe I can understand the system and, if I'm lucky, why they chose the design decisions they did.