Please create an account to participate in the Slashdot moderation system


Forgot your password?

Comment Re:wtf? (Score 1) 605

You kill your cash cow IF your cash cow is lame and on life-support (basically the x86) AND you have a substitute cash cow that is so good, so novel and so ready that you can switch to it and gain more customers than you lose AND gain a time advantage over all your competitors of such magnitude that you'll have a very stable market before anyone is in a position to match you.

This does happen in the CPU world from time to time. The problem is that Intel has failed to make a new product of adequate maturity from the start, which is what it needs. If Intel had ditched its internal politics (which are horrible, BTW) and continued developing the Itanium until it got to where it is today BEFORE releasing it, it would have scored over big. (The extra time would have also been considerably shorter, as there would have been more designing, more working, less blaming and less avoidance.) Intel LOST a remarkable opportunity to eliminate the x86 market, because they had (and have) much the same culture as the old IBM, NASA and Lockheed Martin - pushing for good headlines rather than good products. I guess it's also the same as the Medieval Catholic Church. In the end, there is only one possible result - a catastrophic collapse in confidence and capability, where you go from market leader to borderline extinction in pretty much an afternoon. Sinkholes, rot and corporate failure all share one thing in common - you see nothing until it's all over.

ARM and MIPS have pretty much total control over the mobile world and the embedded world. However, the PC world isn't going to vanish. It will change, though. Why? Because there'll be more action in the ARM and MIPS toolchains, because people will want programs to work seamlessly across devices (not just talk, but actually run on multiple systems), because manufacturing is cheaper when you've a common base architecture, because software companies don't like supporting multiple architectures. And that means PCs will have to switch to the same architecture as the mobile/embedded world, albeit with performance considerations rather than power.

Comment Re:They could.... (Score 1) 605

There are still a huge number of programs (especially for Windows) for which 32-bit versions are the only versions. This is by no means a good thing - and at this point, there is no excuse for it. (Size may have been a factor at one point, but if you're capable of running 64-bits, you're probably working with a memory that can cope with the larger pointer sizes.)

Firefox, for example, is threatening to abandon the 64-bit line. You lose 32-bit support, you lose Firefox support. Now, that might not be such a bad thing (it's bleedin' bloatware, it's slow, the distinction between plugins and extensions is annoying, I hate the menu tree layout, they release major releases too often, there's no UTF-32, and I'm sure there are other problems I could find if I could be bothered, and they should Get Off My Lawn!). However, Chrome and clones don't have enough plugins yet, Opera isn't the performance giant it was, and there's really no other browsers out there of much significance.

Comment Re:Why would Intel want to kill the x86? (Score 1) 605

I would object to calling Intel's processors "the best". The best at what? I'd pit a MIPS64 CPU against an x64 CPU (equal cores, equal FPUs, equal bus speeds) for servers. I'd consider the T2 to be easier to radiation-proof than comparable Intel offerings. The AMULET processor series is the most original. For bus speeds, HyperTransport is still faster than PCI Express.

That doesn't mean Intel isn't "the best" at XYZ, what it means is that there is no generic "the best". There will ALWAYS be another processor that is better at something else.

Comment Re:evidence-based policy (Score 1) 1106

I have nothing against having the biggest government you could possibly have - 100% of the population actively working on some aspect or aspects of governance - provided that you have as small as is rational - ie: not so small it is essentially run by a single person as a personal dictatorship - government working on any given specific aspect.

In fact, I have rather more objection to the current state of affairs, where voters and pundits pronounce on issues they lack the information or education to comprehend. Education I've addressed lots of times (we need more of it, it needs to be considerably more intensive, it needs to be considerably more extensive and it absolutely needs to be devoid of creationists and conspiracy theorists), but that only provides the necessary foundation. The only place to acquire experience and current factual information on government is in government itself.

There is one other benefit. It's easier for lobbyists to bribe/coerce/blackmail 300 people to vote specific ways than to bribe 300 million.

Comment Re:Not the CPI please (Score 1) 1106

The "ideal minimum wage" would be the total cost of food, healthcare, transport and housing as recommended by the experts in each field for providing a quality and quantity of life such that for the simple majority of people on those wages, it would require a disproportionately large increase in wage to significantly impact that quality or quantity of life.

In other words, they are not suffering actual, measurable harm by being on the minimum wage versus the alternatives (leeching off others, going onto welfare, becoming a survivalist, becoming a member of the criminal class, etc).

(No, not all criminals are "just trying to survive", but anyone who is just trying to survive is going to consider crime a valid alternative to eating roots and berries.)

The "practical minimum wage" is the best available compromise between what wage-slave masters will tolerate and the ideal minimum wage. Wage-slave masters don't give an f about your health, even though it costs more to replace someone than to keep them healthy and functional. They don't give an f about anything beyond their day-to-day profits, the actual economics be damned. Odds are, many are guilty of criminal activity already, so to them one more crime means bugger all.

If you have no minimum wage, these people would pay as close to absolutely nothing at all that they possibly could. Some already pay nothing and use fear and intimidation to insure there's no trouble. If you raise minimum wage by anything at all, a few more bosses get added to the number. You can't escape it. But if you increase it by just the right amount, the number of people who gain truly functional lives outnumber the ones who become trampled on further.

Of course, if the US did something rational, like adopt universal "free" healthcare (ie: distribute the cost across everyone), legislate that nobody could be involuntarily homeless, and adopt other rules considered basic human standards by the rest of humanity, then you could actually LOWER minimum wage since there would be fewer expenses that minimum wage would need to cover.

Comment Re:Not an unexpected event.... (Score 1) 626

How do you get 3? At 10k per pop, 1 mil allows 100 sequences to be performed. Divide by 3 sources, they should be able to do 33 sequences each. The cost of sequencing is further reduced if you just want SNPs, as you can use a microarray for that. . But sequencing the entire genome is cheap these days. The cost has fallen very rapidly and continues to do so. It really is at the point where garage enthusiasts can perform many of the basic steps and will be able to do at-home SNP discovery within a few years. Of course, at-home work isn't criminology-grade, but just as enthusiasts can do basic work for pocket money, crime labs can easily swallow the cost of a clean room plus an Illumina or Oxford Nanopore sequencer.

(Oxford says it can get costs down to $1k per sequence within a year or two. They also are working on sub-thousand disposable sequencers, which may be useful.)

Comment Re:God, not this again. (Score 1) 292

The question has no meaning. By using a static spacetime diagram, there is no before or after. Time is merely a spacial dimension in this type of analysis.

Furthermore, "you" have no defined meaning in either of my two scenarios. In the first, simpler, framework, "you" can be either the individual particles in the body, the body as a collective whole, the instantaneous logical state of the brain, the collective logical state over a defined unit of time, or any combination thereof.

The Greek Ship paradox only occurs because you reuse the same label for utterly different aspects of a construct that is simultaneously logical and physical. By using a generic label, you can persuade yourself of almost anything. You must use specifics. And, yes, that means distinguishing object state from object dynamics from object encapsulation.

This is what I mean about uneducated. You lack the understanding necessary to comprehend my first post, you will doubtless fail to understand this one, and you cannot even be bothered to do the basic legwork to comprehend spacetime static waves (far simpler than m-theory, which I guarantee is as complex a model of existence as anyone has managed to achieve).

It is with nonsensical replies such as yours that I end up wondering if eugenics was such a bad concept. I still firmly believe better schooling would fix most examples of stupidity (I think 9 hours/day, compulsory between the ages of 3 and 23, narrow-band streaming per subject should suffice).

Comment Re:God, not this again. (Score 1) 292

Science is not done by straw poll, so the views of most (uneducated, I might add) people is unimportant. What matters is that physicists and mathematicians take the possibility seriously and have published papers on how simulation affects QM.

Something "exists" IFF there is a defined energy matrix superimposed on a defined probability matrix where said matrices cover non-zero, finite space, and non-zero, finite time, and interact with other such matrices of equal or higher number of dimensions.

The universe can be described also as a single object, static in 5D, with all possibilities as branches, that lies at the intersection of two membranes.

Comment Re:Pro Bono Opportunity (Score 1) 626

The cost of sequencing is negligible these days (around $10k for a full genome). Buying a machine isn't cheap - I looked at the cost of one a year or so ago, and they were still multi-million dollar devices. (Why? Bcause I'm a geek! Having a sequencer of my own would be bloody amazing! Useless, but amazing!) Buying a slot at a lab that can run a full sequence - dirt cheap.

This would, however, mean using REAL data and not the 7-12 markers they currently use for criminology. (NB: Genetic genealogists looking to see if two people are closely related would need to perform in excess of 100 STRs and a dozen or so SNPs - if both are male, PLUS a whole load of markers off the autosomnal region, PLUS a full mitochondrial sequencing. And even then, accuracy isn't great and falls off sharply. Nobody in genetics, even those who are experts in the criminology aspect, takes current DNA testing by police seriously. The probability of coincidental matches is too high.)

Comment Re:Not an unexpected event.... (Score 2) 626

Easy. Move to a system that focuses more on rehabilitation, retraining and (when an external element is a factor) removal of external factors contributing to the criminality. You still isolate from society (the sole benefit of prison) but with reduced or eliminated punitive element, there is no risk of punishing an innocent person who happens to be cojoined to someone who is guilty.

Comment Re:Not an unexpected event.... (Score 1) 626

Even in the case of identical twins, the genome differs by many hundreds of markers.

The sequencing of the entire genome would cost around $10,000 (USD) each. That, by my calculation, makes $30,000 (one per guy and one for the DNA sample). The cost of finding the closest match is the cost of writing about 15 lines of C code.

I am trying to figure out where the rest of the money goes.

Comment God, not this again. (Score 1) 292

Look, this is very simple. We don't even know if THIS universe is a computer simulation. (See arXiv for constraints.) If this universe is a simulation, it is by definition a cyberspace. If cyberspace does not exist, then no law governing anything within this universe is possible.

Since laws governing this universe are possible, one of the statements in that chain must be false. The one most likely to be false is that cyberspace does not exist.

If cyberspace is true, then it is just as possible to establish laws in cyberspace.

However, and this is the incredibly annoying part, the assumption by the original article was that you couldn't have cyberlaws AND laws within nations. The cables have a physical location AND a logical location, and therefore must be subject to laws in both.

Comment Re:Customized resumes?????! (Score 1) 113

True, but it doesn't matter what the company is like. When people look at your resume in the future, be it on paper or on LinkedIn, they notice the names. Having Big Names on your resume is significant, and anyone applying for a job involving Linux is applying for a job at a company that knows Red Hat and probably uses their enterprise products.

In the end, a resume is all smoke and mirrors anyway. Accurate, yes. Truthful, yes. But market-speak just the same. It has no relationship to your ability to do things in the future, all it can ever do is summarize in hyper-compressed form what you have done in the past, in a context your future employers probably don't have, where the hyper-compression necessarily takes out much of the essential information as to what the starting point was and how the end point was achieved from there.

The employer knows this just as well as you do. So employers pay a lot of attention to brand names, in the hope that vendors with a good reputation to uphold will have hired and/or kept on employees who help maintain that reputation.

Again, this goes back to Marketing 101 - it doesn't matter what you know, it matters only who you know. (Which is why industry is such a mess.)

Comment Customized resumes?????! (Score 2) 113

I do not, will not, customize a resume for Red Hat. The Starship Enterprise could be flying over and hiring, but they would get standard and that is that.

Those who have spent any serious time applying for jobs know that numbers matter. It is ALL a numbers game. There may be ten thousand, or maybe a hundred thousand, people who will apply for the position who technically qualify. The job market is overflowing with programmers who have "mad skillz" (and maybe even spelling skills). The odds of getting the job are very very slim and you will have taken 8 hours to customize the resume, format it perfectly, etc. It takes about 2 minutes to fire off the appropriate standard resume (I assume you've three or four standard resumes) and a marginally modified cover letter.

Assuming the probability of getting a specific job is about the same, you do the maths on which is the more productive approach.

Sure, Red Hat is a major prize, but so is the lottery. And you know how that is a really crappy investment.

Slashdot Top Deals

The "cutting edge" is getting rather dull. -- Andy Purshottam