Forgot your password?

Comment: Re:So many practice doing it wrong (Score 1) 154

by raymorris (#48026729) Attached to: New Research Casts Doubt On the "10,000 Hour Rule" of Expertise

> Properly designed databases should in theory never contain NULL-columns or values (unknown state)!

Codd and Date disagree with you. Date may or may not be using NULL is his new database architecture, but relations use nulls to represent "unknown". Of course, you want to reduce the instance of nulls as much as practicable.

> So he probably did it the wrong way (checking for special values),

Unfortunately, not even that. In his mind, and therefore his systems, zero, empty string, unknown, and several other values are all the same thing, and those values aren't checked for. So if we don't know how long a task took, it's recorded as having taken zero hours. We record that we paid $5,000 to have a task done, and it took zero hours, so our cost per hour is infinitely high. Before sending the report to policy makers, that information is aggregated, so you get:

hours cost
0 $5000
1 $50
2 $100
6 $300
1 $50

Totals: 10 hours, $5500
Our average cost is is $550 / hour, we tell the policymakers.

Comment: ctrl-F null returns empty (Score 1) 154

by raymorris (#48026671) Attached to: New Research Casts Doubt On the "10,000 Hour Rule" of Expertise

Doing ctrl-f search for "null" finds no mentions of them in that 580 page pdf, so I don't know what Date's current thinking is. In the last book I read by Date, he was basically suggesting that maybe we should have two types of null. Is that still what he's saying? If he's changed his tune, what is his current suggestion for unknown values?

Comment: Re:net metering != solar and 10% needs new physics (Score 1) 318

by raymorris (#48026623) Attached to: Energy Utilities Trying To Stifle Growth of Solar Power

> It is sad how people are so negative about change that all they can see is the problems at the next step, and not the potential three or four steps into the future.

The plan proposed by the pump-and-dump^H^H^^H^H^H^H^H kindly adviser you talked to:

1) Give us all your money
2) We'll breed magic unicorns who fart free power
3) We'll sell the power for lots of money
4) Profit!

Sure, "three or four steps into the future" looks great, but the fact that step 2 is impossible means you'll never get to step 3! It's a ripoff, dude!

Here's a company actually making, and selling, solar power units that actually WORK, and actually save the purchaser money:

Comment: Cover the entire US with 12 feet of water (Score 1) 318

by raymorris (#48026583) Attached to: Energy Utilities Trying To Stifle Growth of Solar Power

So you'd like to store energy in the form of water in a reservoir. Okay, there's precedent for that. Hoover Dam, located in Black Canyon, produces 4.2 billion kWh annually.

I told you I'd throw some arithmetic at your proposal. Let's start getting a rough idea of how that would work, then we can refine the numbers. The US uses 1,4000 billion kWh, so Hoover Dam provides 0.03% of our energy needs. 3,333 reservoirs the size of Lake Meade would do the trick, if we had huge 3,333 canyons to put them in, so they could all have dams 700 feet tall. Lake Meade covers 247 square miles and is 590 feet deep. That means 3,333 of Lake Meade would cover 823,251 square miles. The land area of the continental US is 2,959,064 square miles. So we'd need to cover 28% of the country, to a depth of 590 feet, in order to have all our energy come from hydro.

PUMPED hydro is different from traditional hydro in that we only need to store enough water to last a week, until the cloudy period is over. So we don't need the water to be 590 feet deep - 50 feet deep would do the trick. Of course, 28% of the country isn't canyons. Most of the country (everything between the Appalachians and the Rockies) is flat plains. When you try to flood the middle of the country to 50 deep, it spills out all over the rest of country, and you end up with the entire country under 12 feet of water in order to have enough water to provide a week of energy storage.

Sorry, I _told_ you that a little bit of arithmetic would make you look silly. And that's without even getting into the millions of giant pumps you'd need in order to pump billions of gallons of water every day.

Comment: PR and your money (Score 1) 318

by raymorris (#48024879) Attached to: Energy Utilities Trying To Stifle Growth of Solar Power

Walmart is putting panels on some stores that they say will provide 5%-10% of their power. So as you say, they aren't saving that,much money- skylights might well save more money. However, the money paying for the solar panels is mostly YOUR money, not theirs. They don't care so much about saving YOU money. The taxpayers pay most of the cost od building the solar panel factory, pay the company tax credits for running the factory, and pay Walmart credits for buying the panels. When you put $1 trillion of taxpayer money in to be spent based on politics rather than arithmetic, you end up with silly decisions.

Solar electric is also good PR, of course. The hot water in the Walmart bathroom sink could be 100% solar heated with $30 worth of black ABS pipe, but that wouldn't get them PR headlines.

Comment: half (Score 1) 318

by raymorris (#48024817) Attached to: Energy Utilities Trying To Stifle Growth of Solar Power

If sales and buyback rates were based on time of day , noontime power would be priced low going both directions. Agreed, that alleviates a lot of the problem I described. By more accurately representing actual costs, it also makes solar much less attractive- you'd no longer be able to sell the company 1Kwh of cheap noon power and use that money to buy 1Kwh of more expensive evening power. You'd need 15Kw of noon solar to pay for 1Kw of evening power, because that's the actual production cost. That could work, we'd just need solar systems 80% less expensive in order for it to do a lot of good.

I don't actually assume that the peak solar is greater than the demand at that time, though I am talking about a scenario where solar produces significant power. "Enough to meet 100% of noon demand " is just the simplest, most obvious point on the graph. The problem exists for any amount of net metering. It's at the point if zero net cost to consumers that it becomes OBVIOUS that the power company would definitely go out of business immediately.

The power company spends quite a bit of it's revenue on customer service and on infrastructure. Someone has to handle the billing, collect late payments, hook up new connections, etc. They have to maintain lines, transformers, generators, etc. IT staff buils and maintain all of the big systems for billing, dispatch, etc as well as desktops. People are paid to handle payroll, HR, legal and regulatory, etc. So let's guess that 50% of the cost is actually generating the electricity itself and 50% is running the organization. That number might be wrong, but we're just talking about the concept. At the generating station, you have a set of $XX million generators that need to be paid for and maintained, whether they are in use that hour or not. Even if the generator is turned off for an hour, the people who loaned the $XX million still want the power company to make their payments on time. The staff at the power station still needs to be paid, inspections still need to be done and certifications renewed. So figure at least 30% of the costs at the power station itself continue even if no power is sold from one of the generators for a couple of hours. In total, we can say that maybe 40% of the cost is fuel, and 60% is everything else. The exact numbers don't matter.

So let's say solar covers 50% of the power needs at noon. That reduces the company's FUEL cost by 50%, but most of their costs are unaffected- they still have to pay their IT guy the same amount. Actually, they have to hire another IT gal to build and manage the net metering system. Their total costs for the hour drop by 20% due to reduced fuel usage, while their revenue drops by half.

That bears repeating - if solar provides 50% of the power, their revenue drops by 50% while their costs drop by only 20%. Where do you think that 30% is going to be made up? They could fire all of the customer service staff (and go,outbof business when the phones go unanswered) or increase rates by 30%. Something major would have to change.

Comment: they claim the will (Score 1) 318

by raymorris (#48024511) Attached to: Energy Utilities Trying To Stifle Growth of Solar Power

"the company scaling up to build ... they claim"

For fifty years people have been claiming their company is just about to start making some magical new energy stuff. My uncle claims he's Napoleon. Call me when it happens. Fyi, if it costs $20,000 to make something, and the government (taxpayers) pays $15,000 of that through subsidies, that's still a cost of $20,000. We all can't subsidize ourselves for thousands of dollars per month.

Comment: choose 4 hours by direction (Score 2) 318

by raymorris (#48024483) Attached to: Energy Utilities Trying To Stifle Growth of Solar Power

I said most of the power is in a four-hour period. Your numbers match that, you just pointed your panels into the afternoon sun. You'd get more power, earlier in the day, by pointing them more upward. You might prefer less power later. Of course what time that is also shifts by an hour based on daylight savings time.

You can (and probably do) also buy a system that is incapable of converting all of the peak power. In that case, your power generation will flatline not because the amount of sunlight remained steady, but because your system was incapable of converting. all of the brightest sun - you get only got 3PM power out of 1:00PM light, even though the 1:00PM light was much brighter.

Comment: good thought. Most getting smart meters anyway (Score 2) 318

by raymorris (#48024447) Attached to: Energy Utilities Trying To Stifle Growth of Solar Power

That's a heck of an idea. Many places already have smart meters, or will soon.
It could certainly work like you said- noontime power would be very inexpensive if a lot of people had solar. Of course, that means the economics of buying solar panels would change significantly since the buyback would reflect actual costs. You'd choose between buying your noon electricity cheaply from the power company (from your neighbors, indirectly) or selling noon power at a low rate. Solar electric systems would probably have to get a lot less expensive before it would make sense for many people to buy them.

Using a piece of black pipe to heat most of your water for free - that already makes sense for many people.
It annoys me that most of the talk about solar energy ignores the inexpensive, effective type (solar heating) and focuses on the expensive, impractical type (solar electric) .

Comment: net metering != solar and 10% needs new physics (Score 5, Interesting) 318

by raymorris (#48024089) Attached to: Energy Utilities Trying To Stifle Growth of Solar Power

Electric companies don't like being forced to pay far above their normal cost for something they have to throw away by shunting it to ground. That's net metering, when done on a large scale. The light outside might LOOK ten times brighter than the lighting inside Walmart, but it's actually 10,000 times brighter. Your eyes are very good at seeing in a wide range of light - from candlelight to full sun, a million times brighter. They do so by using a logarithmic, rather than linear, scale for brightness. For the same reason, although the noon sun may APPEAR to be only twice as bright as the sun at 9:00 AM, it's actually much, much brighter. Virtually all of the solar electric is generated when the sun is bright, from about 10:00-2:00.

What that means is that if most people had solar panels, from 10:00-2:00 they could generate as much power as they use the rest of the day. Their electric bill under net metering would be zero. However, the power company still has to provide power to them the other 20 hours per day - for free. See how that could be a problem for the utility, having to provide power for everyone, but nobody has to pay for it?

The utility can't give them back the power generated ten hours earlier, because there is no effective way to store power at utility scale. I know someone who heard a stock tip about some cool new company with magic storage will want to argue with me on that, but I've looked into all of the options and nome of them work at scale. You can try to argue with me, but I'll make you look very, very foolish when I apply some arithmetic to your idea.

Net metering is survivable if only 1% of people do it, because their neighbors can use their noon power. If everyone is doing net metering, you need a magic free energy source the other 20 hours per day. If you decide that solar electric implies net metering, you only end up proving solar electric to be impractical, because net metering absolutely, positively cannot ever possibly work for more than a small fraction of the population.

On a related note, if your argument for solar power assumes that solar means solar electric, you're probably shooting yourself in the foot too. There are several varieties of solar power that work well. Solar water heaters are a no-brainer. Solar electric is probably the silliest approach that anyone seriously suggests, as shown by the trillions of dollars we've wasted on utter fail so far.

Comment: Re:When they a) unlawfully b) abuse c) monopoly d) (Score 1) 351

-- because the user ultimately has the choice, and oems have never been prevented by microsoft from installing a 3rd party browser or media player, nor have users

That is a difference, only Microsoft set IE (the only browser shipped oem) to refuse to download Netscape. Android will download Opera and Firefox just fine, you're right.

Oh, were you under the impression that it was Google who got caught inserting the "don't download Netscape " code in their browser?

> and some jurisdictions consider rooting a mobile device to be illegally breaking its drm

Citation? When I got this phone I'm using right now, I went to the website of the manufacturer (HTC) to look at their instructions for how to unlock the bootloader. You're saying it is illegal to use the phone in accordance with the manufacturer's instructions? Previously, I've visited Google's code site where I downloaded the SOURCE CODE for Android, and Google provides instructions for how to compile your own custom build if you wish, along with a license allowing you to do so. I don't think it's illegal to use Google's code according to Google's license- in any jurisdiction, except perhaps North Korea. I think you're confusing Android with Windows Phone. Microsoft does have DRM in their operating system.

Comment: So many practice doing it wrong (Score 4, Insightful) 154

by raymorris (#48023041) Attached to: New Research Casts Doubt On the "10,000 Hour Rule" of Expertise

The article had logic approximately like this:

Doing it for a long time doesn't always make you an expert.
Therefore, it's genetics that make you an expert.

All around me, I see my co-workers doing it _wrong_ for a long time. I just discovered that one guy who has been in the same job for over ten years was completely unaware of some of the most basic concepts anyone starting in the field should know. This is a database administrator and developer who didn't understand that there is a difference between the number zero, the empty string, and null. He just had never heard of null, it seems. After I explained the idea of null to him, he said our database system (DB2) doesn't support nulls. DB2 has supported nulls since it's first release in 1983. This is a guy who has spent 10-20 years as a professional DB2 developer.

He's had lots of practice, but apparently never opened a book, including the manual. So he's been practicing it wrong for 10-20 years. Surprise, he's not an expert!

Comment: When they a) unlawfully b) abuse c) monopoly d) an (Score 5, Informative) 351

Microsoft was busted for abusing their monopoly power to engage in unlawful, anti-competitive practices.

For example, in Microsoft internal emails, executives discussed the fact that they understood they were hurting their own company, in order to hurt the competitor more. It's okay to try to make your product better than the other guy - that's competition. Intentionally making your product worse, in order to cause compatibility problems for the other guy, is not okay.

99.99% of the time that's self-regulating - most companies can't go around intentionally harming their own company and products or they'll go out of business. A monopoly is a special case. In 1996 Microsoft had 99% share of the desktop market. Therefore they could intentionally damage the computer industry, costing themselves $4 billion, if by doing so they'd cost Netscape $3 billion and put Netscape out of business. Any ordinary company purposely costing themselves $3 billion would be committing suicide, but for a monopolist losing $4 billion in order to make your much smaller competitor go out of business is a "smart" move. That kind of thing is why there are laws about what a monopoly power can do and not do.

Android has 51% of the market. They aren't a monopoly. If Google purposely creates a problem that makes Android worse, in order to also cause a problem for iOS, Microsoft would be jumping for joy. Microsoft only has 3.5% of the market, but they also have $380 billion to spend taking advantage of anything stupid Google might do.

So Google isn't a monopoly, and their actions are competitive, not anti-competitive (in the legal sense).

Do you also wonder about the difference between what Hans Rieser did and what Miley Cyrus did at the MTV awards?

Comment: "As many as", or as few as zero (Score 1) 351

TFA says the contracts include "as many as" 20 Google apps. I'm familiar with that weasel - "as many as", "up to". We all know "up to 30 Mbps" means "3 Mbps, most of the time". So "as many as 20" might well mean "2", most of the time.

One carrier has a contract that mentions they might install up to 20. And?

Comment: /system is read-only for reliability, unless root (Score 1) 351

Unless you choose to root, the /system partition is read-only, so it can't be changed, accidentally or on purpose. That means you can always go back to a stock, working, bootable system. Bugs, hacks, and accidents can't make the phone unusable because the main functionality is unchangeable, unless you root and explicitly mount it read-write. This is one reason that it's safe to pull the battery out of your running phone, while it's not safe to pull power cord from a running computer.

This is now. Later is later.