The government, of course.

#insert observations/law/drferris.h

(preprocessed for your convenience)

"Did you really think we want those laws observed?" said Dr. Ferris. "We want them to be broken. You'd better get it straight that it's not a bunch of boy scouts you're up against... We're after power and we mean it... There's no way to rule innocent men. The only power any government has is the power to crack down on criminals. Well, when there aren't enough criminals one makes them. One declares so many things to be a crime that it becomes impossible for men to live without breaking laws. Who wants a nation of law-abiding citizens? What's there in that for anyone? But just pass the kind of laws that can neither be observed nor enforced or objectively interpreted â" and you create a nation of law-breakers â" and then you cash in on guilt. Now that's the system, Mr. Reardon, that's the game, and once you understand it, you'll be much easier to deal with.â

A funny idea I had once was that after the revolution, the new government gets a 100 page blank notebook to write the laws in.

Once they fill up the last page, all of them are executed.

The next group of guys gets a 100 page blank book.

Have you tried using Windows XP on a touch device?

You know, you're right. Nobody should ever try anything new with voice interaction. We should leave that shit off because its buggy or only knows how to do web searches based on bad guesses.

We shouldn't spend any time putting this stuff in front of users and learning what works well, what doesn't work, what people like, what they don't like. God forbid we try and see if there are ways to integrate it with how people currently use computers.

Instead, what we should do is wait until the 23rd century, when we have starships. Even though we've done no incremental work between now and then, in the distant future, voice recognition and natural language processing is just going to be really, really good, because The Future. It's just going to build itself, and when we are bald and say "COMPUTER" to our starship, its going to listen and then do exactly the right thing, and nobody is going to ask why the bridge has so many buttons and levers and consoles and shit if there is a ship-wide computer with unlimited power and perfect human voice recognition. And we're going to gloss right over how a near-Ai level of natural language understanding still needs us to say COMPUTER first before it figure out who we're talking to, as if anyone else on the bridge could execute the command we're asking when we're staring off into nowhere instead of at another human in the same room...

Anyway, I'm running 9926 on two machines - neither of which are touch-enabled. I've never talked to the thing yet. It appears to run faster than 8.0/8.1. The start menu behavior is better, and you can flip back and forth between little-menu-on-desktop or "big screen of metro" with a simple gesture.

The UI feels positively snappy. The paradigm has been reversed entirely from 8 - now, metro apps run on your desktop - instead of your desktop is some weird bad neighborhood nobody wants you to go to.

I think a lot of people will like Windows 10.

It depends on the specific service member in question.

http://oathkeepers.org/

During the time of the US Civil war, Americans shot their literal brothers - not just their squad mates.

It starts with one soldier. How many follow, and when they follow, depends on the rhetoric of the separatists, how they conduct themselves, how they spread their message, and the counteracting rhetoric and actions of the government.

All of us are alive because people on both sides of the Atlantic with their finger on the "launch" button skipped opportunities to press it. Soldiers are people in difficult situations, trying to balance many opposing directives.

RT has desktop mode.

It's patently untrue that the web is the future for "the kinds of apps that made windows dominant"

Actually, windows was dominant for every kind of app. The growth in apps of all sectors - LOB, entertainment, etc -- is on devices, and people regularly pan device apps that are just thin shells around a browser control.

People want native apps on their devices. MDD (multi-device-development) is something enterprise is very interested in -- they need to deal with a BYOD workforce, and they always want to economize on IT spend.

If it had been feasible to make Win32 apps run well on ARM, don't you think we would have done that?

The most insightful thing you wrote is this:

"But yes, Intel hasn't been asleep, and ARM is no longer as much of a requirement for mobile devices"

Consider the following -- and note that while I work at MS, I am neither privy to, nor attempting to disclose -- any high level strategy

1) Microsoft delivers a lot of value to enterprise customers because of app compat
2) think back a few years at what the CPU landscape looked like -- think about the power consumption of Intel's offerings. Remember, there was no ATOM yet.
3) app compat, battery life, performance -- if you don't have a low-power native x86 processor, you can only get two of these at a time.
4) Enterprise customers want all three
5) Intel, years ago, didn't appear to have any intention to deliver a low-cost, low power x86 part
6) this meant that MS would be unable to deliver low cost, new form factor mobile devices that could still run legacy software
7) this would force a wedge between new form factors and the Microsoft platform advantages (great compatability)

Clearly, what needed to happen is that something had to convince intel to develop a low cost, low power, good performing x86 chip

Based on 20+ years history, considering ARM, AMD, dec Alpha, etc, what makes intel innovate well and do its best work?

A credible marketplace threat to Wintel.

Claim: The purpose of Windows+ARM was to force intel to develop a low-power, low-cost x86 chip. If Windows+ARM took off in its own right, great. But the main purpose has been to secure a $99 x86 windows tablet -- which means that enterprises have the price points and form factors they want, and the app compat they need.

Exhibit A:
http://www.amazon.com/HP-Strea...

I happen to like my RT tablet -- but the Surface Pro is a credible do-it-all device, and now software that runs on the Pro is the same software that runs on your $99 HP tablet and your $4999 gaming rig.

Back when windows+ARM started, the intel hardware to allow that continuum didn't exist.

As I said -- nobody at MS tells me how things really go down. But this is a high stakes game. The people at MS aren't stupid.

There is an unspoken assumption by "sensitive" people that they deserve your time, or that their ideas need to be heard irrespective of merit.

Disagreeing with them, correcting them, or challenging them is _offensive_ to them.

Your first response doesn't convey the additional problem the way the 2nd one does

What additional problem?

"Hi, I am the busiest man in the free software world, and you just consumed my valuable time with amateur level mistakes. We have a FAQ about this, which you either didn't read or didn't understand, and as a result, my productivity has suffered. Please re-read the contribution FAQ. If you are still confused, send a message to LKML, not to me"

"I've cc'd this response to LMKL so that others don't repeat your mistakes, and so that everyone else is expecting your follow up questions and amended patch"

It's not "abusive" to point out that when people don't do their due diligence, everyone else suffers. Set an expectation of excellence, and expect people to follow it.

Results first. Feelings later.

Do you have any evidence of this? At all?

Because here's what I see:

Linux has remade the software world in its own image. I'd hardly call that "failing". Real actual super computer companies (e.g. Silicon Graphics) stopped developing their own OS and started shipping Linux.

Microsoft, the arch nemesis of Linux and Open Source, is shipping kernel patches and releasing code under open source licenses.

What does "success" look like to you?

And lest you say "that's just a singular case", we can look at Theo and OpenBSD. OpenBSD has been wildly successful, both as a BSD fork, but also in its broader mission to cultivate a software culture of excellence and correctness, with results that speak for themselves.

Linux and OpenBSD are two of the oldest open source projects around, with two pretty intense personalities at the helm.

I see no evidence to support your claim whatsoever.

There's this idea in the US that you are never allowed to hurt anyone's feelings.

The problem with that is that people are irrational, oversensitive, and cannot control themselves. Interacting with such fragile daisies is like tiptoeing through a minefield.

It is not, to me, a foregone conclusion that assertive people should tolerate having to deal with sensitive people. That is the prevailing dogma in US business, but it's not clear why that should be the case.

The meta response to this entire conversation space should be "stop bothering me with your bullshit and submit some fucking code"

For many years, I ran an alix2d3 box with OpenBSD installed on it as my edge device. Excellent hardware, excellent OS.

pf.conf is simple for a basic configuration.

If you want to run off of a read-only flash file system, or have a router-style config experience, there are adaptations for that purpose also. But just plain old boring openBSD is a great place to start.

My favorite thing about openBSD is how lightweight the install is. There is very little garbage you'll want to shut off or remove.

For the canonical SOHO edge device, choose any x86 hardware you have, put 2 network interfaces on it, and you're done.

A basic pf.conf that gives you NAT and blocks everything evil from the outside is only a few lines, and well documented on the interwebs.

Put your samba server somewhere else.

Oddly enough, I finally retired my openbsd device and got a few Ubiquity EdgeRouters. My home network situation changed and I wanted a smallish device with POE support, but still wanted a real OS on it..

if you understand scalar assembly, understanding the basic "how" of vector/SIMD programming is conceptually similar

Actually, if you think back to pre-32bit x86 assembler, where the X registers (AX, BX) were actually addressable as half-registers (AH and AL were the high and low sections of AX), you already understand, to some extent, SIMD

SIMD just generalizes the idea that a register is very big (e.g. 512 bits), and the same operation is done in parallel to subregions of the register.

So, for instance, if you have a 512 bit vector register and you want to treat it like 64 separate 8 bit values, you could write code like follows:

C = A + B

If C, A, and B are all 512 bit registers, partitioned into 64 8 bit values, logically, the above vector/SIMD code does the below scalar code:

for (i == 1..64) {
    c[i] = a[i] + b[i]
}

If the particular processor you are executing on has 64 parallel 8-bit adders, then the vector code

C = A + B

Can run as one internal operation, utilizing all 64 adder units in parallel.

That's much better than the scalar version above - a loop that executes 64 passes..

A vector machine could actually be implemented with only 32 adders, and could take 2 internal ops to implement a 64 element vector add... that's still a 32x speedup compared to the scalar, looping version.

The Cray 1 was an amazing machine. It ran at 80mhz in 1976

http://en.wikipedia.org/wiki/C...

According to WP, the only patent on the Cray 1 was for its cooling system...

Actually, I'm going to disagree here.

I spent the last 10 years working on the unglamorous parts of the MS stack that deal with writing business software. I worked on ERP systems (big stuff, big money) and small stuff (VS Lightswitch).

It turns out that there are zillions of little apps that make businesses run. Some of them are no more robust than an excel workbook on a well known file share. Some are Access applications. Lots of VB6 apps are still keeping businesses running.

Integrating with existing systems is notoriously hard. The larger a system is, the more likely you are to need a VAR/integrator for an industry specific vertical to build customizations for you.

There is absolutely a need for tooling to build business critical apps - forms over data, basic workflow, reliable data storage, etc.

It is possible to roll-your-own on each of these pieces, but it turns out that tailoring any of them for business app usage, much less stitching them all together into some sort of recipe or middleware... is hard.

However, the smaller the problem domain is, the larger the advantage in NOT trying to integrate or use off-the-shelf software.

With LightSwitch, you can write a 1-5 form productivity app in a few minutes that will let a small team of people do data entry, with basic validation and reliable storage underneath. This makes it great for volunteer organizations and other low time / low budget endeavours...

I'm very interested to see if other people have had positive experiences apart from Lightswitch... our customers tell us its the only tool quite like it... unless you go back to PowerBuilder or something along those lines...