Indeed. My FF 3.x kept bugging me to upgrade to FF 4.x. Once a week I'd try it, and it would say, "The following ad-ons won't work..." Since that list included Tab Kit, I declined the upgrade
One of the useful features FF should implement would be 'time for developers to respond to new version for extensions.'
E.g. if the developer for Tab Kit normally has an updated version 3 months after a new release of FF, then I'll wait for it. If the developer never responded to the last update, then I start looking fore TK replacements.
Many of the books I use have pages much larger than a kindle screen -- and because of the technical illustrations on them, actually need large pages.
Many of hte books I use have color illustrations. Until the book readers support color, I'll still use paper.
Many of the books I use bristle with post-it note bookmarks for frequently accessed pages. Until an ereader has an easy way to flip back and forth between 20-50 bookmakrs, I'm not interested.
Much of the time I have a dozen reference books spread out over a table. Until I have a screen that allows me to do that, I'm not interested.
***
On the otherhand, having field guides in a waterproof Kindle or Nook would be wonderful.
***
For light reading, ebooks are fine. But I'm not willing to pay more than 25% of the new price for an e-version. With the typical paperback, about 60% of the cost is associated with printing and shipping. Out of the remaining 40% the publishing company pays for its editing, marketing, the layout and the author's nickle, and the bookstore's margin. Selling direct to the public bypassing the bookstore should result in the same margins at a much lower price.
Capitalism works well when there is a competitive market. So the issue is to keep it competitive.
Things that might help.
1. A progressive income tax for corporations. The bigger you get, the more, as a percentage, you pay. Mergers are much less of a win.
This also has the effect of reducing the taxes on small companies, which favours startups.
2. Corporate directors are personally financial liable for everything their company does. This liability extends to holdings they have in other companies. This would tend to reduce directors having multiple seats, which all too often leads to conflict of interest.
3. All senior corporate staff and directors and paid on some form of acculated delay. E.g. This year you get 100,000 bucks and the dividends off of 100,000 shares for 20 years. Next year you get 100,000 and the dividends of of 100,000 shares for 20 years. When you retire, you get dividends for 20 more years. So make sure the company is run right for the long haul.
4. A company cannot own shares in another company. Shares have to be owned by an individual. Obviously the transiition would have to be gradual. The idea here is to bring personal responsibility back into the equation. There may be merit in still allowing a not-for-profit hold shares.
5. Taxes are no longer calculated on net profit, but on gross income. There are no deductions at all. Needless to say, the rate is much lower. This makes accounting much easier. It penalizes very low margin companies more than high margin companies. Again, there would need to be a transition form to not cause complete chaos.
This is not a visual system, but it is very very fast at creating content.
Markdown is a simplified environment for writing that avoids most of the tags and their associated visual clutter, leaving a visual ascii representation similar to email and usenet.
# This is an H1 headline.
## This is an H2 headline
* An unordered list element. * Another list element
* Nested list element
[Text to show as a link](web reference)
You get the drift.
You can't do complicated stuff directly, but most of the time you don't need to.
Template Toolkit allows you to build frames to dump the markdown into.
While a bit clunky, it has enough logic in it to build a menu system from a directory structure.
Layout is done entirely with CSS.
The effect of this is almost complete separation of design from content.
The latest version of MD is MultiMarkdown 3, which in addition to creating HTML has options for creating PDF, LaTeX. Which means taht if you design it properly, you can have one set of easily editable source documents that can provide both web pages and print ready, page numbered manuals.
It's not a complete system. Layout is still a matter of much tweaking of the CSS file, and reloading. Scripts are as ugly as ever.
If this were used for ranking, then I would expect web masters to attribute articles to Big Names.
I would hope that Google would have a policy of fingerprinting the articles. Most people's writing style is sufficiently unique that claiming that someone else wrote Foo is fairly obvious on analysis.
I hope also that there is a search tool so that I can find all articles attributed to me.
And suppose that Slashdot and phpBB support this tag so that I can find all the posts by a given author.
I would think that the police could be even more effective by published 3-4 times as many checkstops as they have.
"Yes, we planned to have one there, but the manpower was needed eslewhere so that one got cancelled..."
Similar to the 'cry wolfs'. If you were a police department you could get a device that would broadcast signals on police radar frequencies, tripping every radar detector for a half a mile or so. It was an easy way to get speed demons to slow down.
The only difference in this one is that the reacting material is a suspended, particular solid. There are commercial systems that use flow through batteries already.
See:
Flow batteries have historically had low energy densities, as the entire reacting mass was in solution. E.g. the article above cites 25-75 Wh/kg
Which means a 50 kWh system would require roughly a cubic meter of tank, or about 5 standard drums. This is acceptable for a house, unworkable for a car. (50 kWh is about 2 days average power for a non-electrically heated house. This size unit would give impressive load level capability to the grid, and would be awesome for off grid use.
Going to a slurry increases the energy density, but possibly with problems with charging. How do you keep the particles the same size during the charge cycle. In many conventional cell batteries, differential plating on charge is often the limiting factor on the number of cycles.
In principle, the working solution could be filtered and the large chunks broken up. From the description of the press release, however the slurry has the consistency of light grease. Filtering would be problematic.
This may require periodic extraction and reprocessing of the working material. Still, this should be easier than rebuilding a battery.
So, you use a device that is hard to use. Suppose you built into a ring. The ring has the ability to interface with a socket, and transmit a password. It's voice activated. "Present slashdot password" touch to socket.
You could also use it to store your public key and exchange it with other ring holders.
Of course it's new technology so initially the rings would need to be very cheap. And it will be the kids who will be early adopters. I suggest that they be given away in boxes of cereal.
Who knew that Captain Midnight Secret Decoder Rings could become reality?
Locally Future Shop pays their people on commission. Best Buy pays straight scale.
I shop at Best Buy. I'm far more likely to get good advice (for a sufficiently small value of good) than at FS where I will be steered to the most expensive alternative.
Indeed, I've had Best Buy people say, "Don't get that here. Get Foobar instead, at Acme Electronics"
In the long run the BB approach generates store loyalty, while FS generates short term higher profit.
The latest thing was the replacement of relays with vacuum tubes. While an individual tube is reasonably reliable -- better than a light bulb -- collect a bunch together and you always have one blowing out.
I heard that with one such school sized computer after WWII the basic flip-flop was a 7 vacuum tube circuit, wired so that ANY two of the tubes could fail and the device would still function.
Several flip-flops were in a drawer. Indicator lights on the drawer front showed status. Red light on drawer, open drawer. Each flip-flop board had indicator lights showing which tubes were good.
People would race though the corridors with shopping carts of tubes doing hot replacements while the machine was running. My recollection was that even so uptime was usually measured in minutes before some other thing would break.
In general, if you aren't the sole admin user on the box, you shouldn't play. You can co-locate, sure. But you must be the person who is responsible for everything on that box. The ONLY access the co-location provider should be able to have is the ability to execute a shutdown script to protect themselves from your rogue processes.
If you are not the sole owner, then you are dependent on their security being up to snuff.
If your application is small, you may be able to work with a virtual cluster of machines running on a single box.
I would give serious consideration to encrypting all sensitive information in the database. Even if the keys are wired into the application, it means that data extraction has to work through the application to get data.
I would also give serious consideration to putting large quantities of meaningless information into the database allong with some means of filtering for your own use. A Black hat who gets 2 million email addresses from your machine, but finds that 99% of them are bogus will not earn the gratitude of his client.
In Alberta, peak power use is the supper hour, followed by the shit, shower and shave time block in the morning. Late at night is the low point.
I recently dropped my power bill by $30 bucks a month by having the 50 Amp pole breaker replaced with a 35 A. one.
I've yet to trip it. I figure it will trip if I have the oven on, and start the table saw the same moment that the well pump kicks
A unit that could average power consumption would be a big deal to the utilities. Would mean far smaller wires needed for the distribution network. Consider: A standard service right now is 100 to 200 A. Which means 4 guage wire to the house. It has to be sized to handle the possibility of 20 KW. Most people, however use under 1000 kwhr/month, which is is on the order of 1 kW continuous. So having a box that:
A: Could hold roughly 1 day's power. B: Could be charged at a rate roughly 1.5 to 2 times the average consumption rate of the customer. C: Could be programmed by the utility company as to when to charge
would be a big win.
As a side effect, consider that in a power outage, everyone has a day's power at normal consumption, and with some frugality perhaps as much as a weeks.
Unless the carts are very sophisticated, it's an invitation to shop lifting. Scan one item into your cart, place 6 in your cart. Avoiding this would required that you have a scale in the cart accurate enough to register a single item's difference. If they come up with this, then just add a squirming kid to the basket. (fresh pork...)
Or you scan a can of el-cheapo discount icecream, but place Haagan Das in the cart. Or you fill a bag with cashews at the bulk bins, and tell the machine that it's wheat bran. Scan bananas, load avocadoes.
Now if I were an ISP and wanted to protect my users, I'd try to make these records as un-useful as possible:
1. Shorten the DHCP lease time.
2. Hack the dhcp server code so that it ignored the renew my address, but passed out a new address each time.
3. At the same time, hack the logging code so that time stamps were in microfortnights since last dhcp server restart.
4. In addition, tweak the format so that it was not trivial to parse. The easiest way to do this would be to have variable numbers of fields. Having multiple dhcp servers logging to a common log file with differing formats would also work.
5. Run your servers on Newfoundland time.
6. Do NOT keep servers time synchronized.
7. Use layers of NAT so that a client address at the ISP office level corresponds to hundreds of users.
Whenever a request came in, you would print a relevant block of data, but to save paper the log files would be printed in 6 point Squint font, and the block of data provided would run hundreds of pages. The output data format is neither by time stamp nor IP address so that someone has to wade thorugh the entire block.
Guru as team lead probably is NOT a good idea. Again, team lead is at least partially a social skill. I've known several guru class coders, and I wouldn't like to have any of them as bosses.
I can see merit in having a master/apprentice system too.
An experienced coder has a young apprentice. The apprentice does the scut work: Making the code pretty to whatever the flavour of the week is, reading the inline documentation, and adding to it as needed. (Or writing it from scratch...) Writing test suites. Assisting in the debug process.
Apprentices should be moved from journeyman to journeyman, say every 6 months. In this way they become familiar with various parts of the code base. They see a bunch of different work styles. They get practice at coming up to speed with other people's code. In addition they act as pollenators. "Mike showed me a neat trick with...."
After several years of this, they become journeymen. But now journeymen are paired to journeymen. The PAIR has to sign off on a milestone. For each segment, one codes, one documents/tests, both debug.
After a few years of these, if there is merit, they become masters, and are paired with an apprentice.
Some journeymen may not be master quality. At this point they are diverted into sales, and HR, and support....
The best way I can see to verify an account is with a system that phones you. Most people do not have more than a single land line and a cell phone. Often the land line is shared.
On the first N postings, the system phones your # of record, and simultaneously shows a 5 digit number on the screen. You respond to the phone call by punching in the number.
After N calls, you are verified only every M postings, where the value of M is inversely proportional to your karma.
I don't think that the Slashdot system works well at all. Way too many posts have low signal to noise ratios.
Slashdot has several issues:
1. A story can be modded up for either being funny or for being informative. In general the ones that are 'funny' are usually no more than distractions.
2. Too many of the comments are not well thought out. Many are too short, and boiled down consist of 'I agreee' or 'I disagree'
Ways I would like to see slashdot improved:
1. No mod points for funny.
2. It costs you X points to post. You are given N points per day, and they accumulate up to a max of M points. Short posts -- under 100 words, say -- cost more.
3. Threads are ranked by karma/post. People who have been modded up in the past are more worth listening too now.
4. Thread length is limited.
5. Threads are owned by the thread starter, who automatically has mod authority on the thread.
6. Thread starters have a responsibility to provide a summary of the discussion after the thread has closed. You can't post until you summary is complete. Summaries are also modded up or down. If you posted to a thread, you have an obligation to read and mark the summary.
The idea is to make it 'expensive' to post, so that posting isn't done for trivial reasons.
Part 2.
One of the things I like about slashdot is the usual cogent summary, a link to a meaningful story. However:
1. I wish I could subscribe to slashdot by sections. I have zero interest in games, lots of interest in technology. This could be done automatically by having an up/down set of buttons much like firehose that would increment/decrement all the tags associated with a story. Additinally, having a link on your profile where you could manually set the point values of every tag. One of the responsibilities of contributors would be to tag stories appropriately.
2. I wish the moderators would be a little more discriminating about stories. Sometimes we have the same story coming up several times in the space of a few days. I would much prefer a somewhat longer story, with several links to external articles. In particular the links to 'gee whiz' articles that have no content, and don't really have a break through behind them are anathema.
To me this concept is exciting: It's a way of collecting AND STORING solar energy.
Using the Mississippi at Itasca is a bit disingenuous. That's fairly close to the headwaters, where it is still small.
At the bottom end, the Mississippi has an annual discharge of about 500 km3/year, which corresponds to about 4.2 million gal/s
So that means you can't put one everywhere. Hydro power is a scale problem too. In hot climates this may be a good way to store power: Sunlight evaporates ponds. Generate power as needed from sea water + brine pond.
There are a fair number of big rivers that could be tapped.
In addition the Dead Sea, and parts of the Sahara are below sea level. Death Valley is probably too picturesque, but the Salton Sea may be suitable. If you want to do this on a large scale dike off the Sea of Cortez (between the Baha peninsula and the mainland Mexican coast.
The much larger gradient between ocean water and the Dead Sea means more watts per gallon of flow.
If done cleverly, the energy from mixing a few gallons of ocean water with a few gallons of brine may be sufficient to desalinate a few gallons of ocean water -- in part of the world where fresh water is scarce.
Looking at it another way if 13000 gal/s = 100 MW you're getting about 7500 J/gallon.
A gallon is about 8 lbs. Call it 40 newtons. So the energy is roughly equivalent to a lift of 200 m.
Now if the energy to be gained is proportional to the amount of dissolved salt, then for concentrated brine (about 30%) there should be about 75000 J/gallon.
Couple this with the idea of tower evaporators:
Desert climate: Build a tall tower in the middle of a large shallow pond. In the day, the pond gets warm, evaporates. Pump warm water to the top of the tower. Drop as a mist. Evaporative cooling makes the air heavier. Choose water droplet size so that about 3/4 of the water evaporates. Now you have cold wet air coming out of the base of the tower. At night this can create a heavy dew -- lots of desert plants can survive on dew.
The wind from the outrushing air may be enough to run the pumps. The brine + seawater runs the salt water battery system.
Salt can either be used as a feed stock for various chemical processes or stacked out of the way, like surplus sufur is now.
Remuneration for Company Officers I can hear the squeals from here...
Right now corporate executives of large companies are paid very large salaries with even larger bonuses, bonuses that they seem to get even as they waddle up to the trough for a government bailout.
There is no incentive to think beyond the next quarter, or at most, next bonus period. I watched an interview on one of the news documentary shows where an official admitted he knew the bubble would burst, but was going to make money as long as he could, then walk away.
Never mind bears and bulls, Wall Street seems to be owned by pigs. I'm sure I'm casting aspersions on many solid officers. Bad apples and all that. But where is the incentive to think beyond the next annual report?
Here's a solution. Tell me why it won't work:
Suppose that Company officers were paid a reasonable stipend -- no better than, say, a good doctor or senior engineer, but also received the dividends from a block of common shares of the company for the next 20 years. They would not receive the stock. Only the dividents from the stock. Or if they did receive the stock, it could not be sold until the 20 years was up.
So I get an offer to be CEO of Almagamated Consolidated. I get an up front salary of $200,000 per year. I also get the dividend from 100,000 shares of AlCo. Now AlCo has been a good company, and pays $4 a year in dividends. This, in effect, triples my salary. And, even if I quit AlCo this year, I will get another 19 years of dividends from AlCo. If I do a good job of running AlCo and the dividend rises to $6, then I've raised my salary for that year by a healthy amount.
Next year I get another 200,000 salary, and another block of shares' dividends.
At the end of 20 years I'm getting dividends from 2,000,000 shares of Almalgamated Consolidated. If I've done a good job of running the company, this should be a tidy bit of change. If ownership of the shares was part of the package at this point I can now sell the first block of shares. But even if I retire today, the payments will continue for another 20 years, and I'll be holding shares for another 20 years. I'd better leave the company in good shape with a good man at the helm.
Look at what this does:
The people who run the company now have a huge incentive not to look at the next quarter's earnings, but to try to plan for the long haul.
I won't guarantee that I've picked the right numbers. Depending on the goals of the company it might be 10 years. Or 30 years. It's not a stock option. It's a stock compulsion.
Indeed I think that this form of remuneration should carry down through the ranks. This would build a financial backing to company loyalty as well as make life run more smoothly when layoffs happen.
Until you are audited by the tax demons. In Canada they can audit for 7 years back, further if they suspect deliberate tax evasion. If you are running a business, you keep everything.
For us this amounts to a 3" thick folder of folders per year.
I also keep bills of sale for anything I've bought that is more than 'disposable' E.g -- tractor, truck, auger. So if there is ever a dispute about ownership, I can show good faith.
Unspoken however is the reasoning behind using such a crummy cell in the first place.
My suspicion is that they had a technique that had the potential for reducing internal resistance in cells, so they used a material that had a high internal resistance so that their efforts would show up in a way that could be measured.
If this technique can be applied to other cells, the effect is likely to be additive, not multiplicative -- ie it will take a 20% efficient cell up to 21. And be just as you pointed it it will be a small step forward.
I'm not wealthy. My small business doesn't even make me minimum wage. Yet of all the ways that the Government can spend money, well chosen basic research is high on my list.
Really basic research cannot be managed like a production line. It's closer to being prospecting, except the long term payoff is better. Most of the time you really don't know what you are going to discover. (If you did, it wouldn't be basic research)
It's not clear to me what the best way to fund it is, but I can the following as possibilities.
1. Multiple agencies similar to the National Science Foundation and the National Institute of Health. Their funding should be cyclic, varying about 60% over the time span of 3 grant cycles. This prevents a lot of the "I got it last time, so I'm a shoo-in this time" Each cycle, the sponsoring agency has to prune 2/3 of the deadwood. Having multiple agencies gives you flexibility to change your priorities, and encourages PIs to not be so specialized.
2. Industry consortiums. Any industry that had more than 5 participants could form a research consortium. They pool research money, and get a tax break on that. They have a board that can award research grants. All results are shared amoung the research pool first. For any given project the consortium can decide that it is valuable research, and can put a lid on it, keeping it internal, but to do so they have for forego tax writeoffs for the project until they open and publish the research. This requires some serious auditing
The limitation on numbers is mostly so that one firm cannot dominate the consortium.
3. More prizes. These focus attention on science and research.
4. As a result of automation, we are getting more and more people who don't really have meaningful work. The number of really brilliant scientists are few. We need a research system that encourages one PI to mentor/direct the not quite so bright ones. And we need to figure out ways that the PI doesn't spend his life as an administrator.
The transition to metric in Canada has been going on for decades. It's a pain in the butt.
One of the advantages of the imperial system was that more things are chunked in convenient units. Tinned veggies came in 12 oz and 20 oz cans. Now we get 373 ml cans. 350 ml cans. 342 ml cans.350 gram packages of frozen vedge. 500 gram, and kilogram packages.
Milk was mandated by law to move into integral liters. The milk industry used this as an opportunity to resize milk crates. Where they used to have an internal dimension perfect for moving LP records, now they are about 3/16 (sorry 5 mm) too small.
Meat is advertised at dollars per pound, but labeled in the store at dollars per kilogram. Cheese is an international product. In the same store, I can find it priced per pound, priced per hundred grams, and priced per kilogram all on the same shelf.
Potatoes can be bought in a 5 or 10 kg bag, or if they are from the US in a 10 or 20 lb bag with the metric unit in very small print somewhere. Dog food used to be universally in 40 lb bags. Now you can get it in 8 Kg, 18 kg, 44 lb (20 kg) and 50 lb bags.
If a company didn't want to change their production line, they just bought new labels. So you can buy a quart of something and get: 1000 ml if they did convert. 946.35 ml if they were a U.S. company and relabeled their quart. 1136.52 ml if they were packaging in British (Imperial) quarts.
Doing comparison shopping between 946 and 1136 is a challenge. If you do a lot of it, you end up with approximate conversion factors in your head, A british quart to is 6/5 of an american quart.
Our road signs are in kilometers. But the rural road network is in units of miles. Many people still think of milage in miles per gallon, even though we buy gasoline in liters and measure distance in kilometers. Andd officially mileage figures are in liters/hundred kilometers -- an inverse unit, so to convert you have to do do mental long division.
The housing industry revolted when they proposed changing the sizes of dimension lumber. Sheet goods were going to go from 4x8 feet to 125 x 250 cm. A 2x4 would have gone from 44.45 x 88.9 mm to 45x90 mm Which would mean a reno project would require every board to be cut to match. Or it would mean that the mills and the slupply stores would have twice the inventory, and things that were so close would be constantly swapped.
They reached a compromise. Most of the building trade was left alone, but sheet goods are in thickness measured in mm. So you can't buy a 4x8 x 1/2" sheet of oriented strand board. It's 4 feet x 8 feet by 12 mm.
BUT All the building codes are in metric. AND the plans you submit for a permit have to be be in metric. And for some obscure reason they settled on mm. I have seen property plans done with mm's as units. E.g 20,000 x 40,000 mm instead of 20 x 40 m.
Hydraulics, already a maze became twice the hassle. I know one hydraulics shop, you bring something in with a metric fitting, he says, "That's metric. Take it in to the city." It's not worth his while to double his inventory.
Plants only use about 1% of the energy in sunlight.
Most plants, the problem is CO2 availability, water, and heat.
Plants in a greenhouse can absorb ALL of the CO2 in a few minutes. One of the reasons that greenhouses are expensive to run is replacing and heating all that air. And you can get signficantly higher production by adding CO2 to the air.
Photosynthesis slows down as the temperature goes up. Reason: Evaporation picks up, and the plants close the stomata to reduce water loss.
The energy cycle to convert light into sugar is complex. Not all frequencies can be used effectively. Plants are green because that's the peak of the frequencies it doesn't use.
I'll point out that an aerial view of the boreal forest is close to black. While an individual spruce needle is green, the light that misses a needle has to bounce too many times to return to the sky. So in terms of keeping warm, spruce trees are doing a pretty good job. You'll note that spruce are not common at low latitudes.
Oh, and the purple and dark red leaved ornamental trees: In most of these, the leaf is overproducing purple and red pigments in excess, masking the chlorophyll. It is a disadvantage to them, making them less efficient at photosynthesis. These plants tend to grow more slowly, and be more disease prone, due to the shorter energy supply.
The problem is that there won't be useful work for a large number of people. Education can help, but there is still 50% of the population that is below average in smarts.
Now you can reduce this fraction somewhat by using different kinds of smarts. At present there is no good way to outsource or automate certain classes of jobs. It will be a while before we automate truck driving to the point where there is no driver in the cab. Outsourcing plumbing repair or furnace installation is a ways away.
Stuff is getting cheaper. Thanks to automated factories, outsourced factories, machine intelligence. For the same reasons stuff is getting better quality, although there is still junk out there.
So where does it end? Historically when we have made big gains in productivity, the work week got shorter. But the intellectual jobs, the ones with high salaries, are ones with long hours. Ask any sysadmin, and find out how many of them are working 32 hour weeks.
At the low end, here in Canada, many brick and mortar stores have 90% part time workers -- typically working 15 to 20 hours a week. If they are part time, they don't have to pay benefits. So people end up juggling 3 part time jobs to make ends meet.
If you want to see where it is going, take a tour on most Indian reservations. See the joy and the bliss of communities that have 70% unemployment, where the entire community economy is dependent directly or indirectly on government handouts.
Education is a partial answer. But here in Canada we're finding that a university degree doesn't necessarily get you a job. Indeed, do we need 75% of the population with a university degree? Some fields do. One of my former students recently finished his civil engineering degree, and was snapped up at 70,000/year. But for each guy like him there are a dozen who move back home after University. But not everyone has the savvy to be an engineer.
Our local tech school advertises, "Now that you have your degree, come and learn a trade so you can make some money and pay off those loans" And people are doing that in droves. Go back, and take training to become an electrician, welder, surveyor. A computer geek working for Best Buy can get $13/hour, and probably works 15 hours a week with no benefits. A high school kid with a grade 10 education can get a job rebuilding drill bits for $18/hour, plus $2/hour shift differential, plus 10-20 hours of overtime, plus 30-40% bonuses every 3 months. A welder can expect 27-32/hour.
Energy has to get a lot more expensive before it becomes more economical to do things locally again. Off hand, I'd guess 2 orders of magnitude more expensive. Right now it is cheaper to move apples from New Zeeland to Canada than it is to store apples in Canada. At the same time, my uncle, who grows apples in Washington state, ships containers of apples to India. A few thousand dollars takes a container anywhere in the world. The cost of the contents has to be small compared to that cost before the economics favour local production again. Apples wholesale around $10/box. A box is what, 2 cubic feet. A container is 8x8x40 feet -- 10,000 cubic feet. So the apples are worth 50,000. Shipping is 3-6000. Remember that rising energy costs will increase both prices.
H. G. Wells commented that 'progress is a race between education and catastrophy' It's not clear to me that education is winning, or if, indeed it can win.
Slashdot Mirror
Indeed. My FF 3.x kept bugging me to upgrade to FF 4.x. Once a week I'd try it, and it would say, "The following ad-ons won't work..." Since that list included Tab Kit, I declined the upgrade
One of the useful features FF should implement would be 'time for developers to respond to new version for extensions.'
E.g. if the developer for Tab Kit normally has an updated version 3 months after a new release of FF, then I'll wait for it. If the developer never responded to the last update, then I start looking fore TK replacements.
until this happens.
Many of the books I use have pages much larger than a kindle screen -- and because of the technical illustrations on them, actually need large pages.
Many of hte books I use have color illustrations. Until the book readers support color, I'll still use paper.
Many of the books I use bristle with post-it note bookmarks for frequently accessed pages. Until an ereader has an easy way to flip back and forth between 20-50 bookmakrs, I'm not interested.
Much of the time I have a dozen reference books spread out over a table. Until I have a screen that allows me to do that, I'm not interested.
***
On the otherhand, having field guides in a waterproof Kindle or Nook would be wonderful.
***
For light reading, ebooks are fine. But I'm not willing to pay more than 25% of the new price for an e-version. With the typical paperback, about 60% of the cost is associated with printing and shipping. Out of the remaining 40% the publishing company pays for its editing, marketing, the layout and the author's nickle, and the bookstore's margin. Selling direct to the public bypassing the bookstore should result in the same margins at a much lower price.
Capitalism works well when there is a competitive market. So the issue is to keep it competitive.
Things that might help.
1. A progressive income tax for corporations. The bigger you get, the more, as a percentage, you pay. Mergers are much less of a win.
This also has the effect of reducing the taxes on small companies, which favours startups.
2. Corporate directors are personally financial liable for everything their company does. This liability extends to holdings they have in other companies. This would tend to reduce directors having multiple seats, which all too often leads to conflict of interest.
3. All senior corporate staff and directors and paid on some form of acculated delay. E.g. This year you get 100,000 bucks and the dividends off of 100,000 shares for 20 years. Next year you get 100,000 and the dividends of of 100,000 shares for 20 years. When you retire, you get dividends for 20 more years. So make sure the company is run right for the long haul.
4. A company cannot own shares in another company. Shares have to be owned by an individual. Obviously the transiition would have to be gradual. The idea here is to bring personal responsibility back into the equation. There may be merit in still allowing a not-for-profit hold shares.
5. Taxes are no longer calculated on net profit, but on gross income. There are no deductions at all. Needless to say, the rate is much lower. This makes accounting much easier. It penalizes very low margin companies more than high margin companies. Again, there would need to be a transition form to not cause complete chaos.
This is not a visual system, but it is very very fast at creating content.
Markdown is a simplified environment for writing that avoids most of the tags and their associated visual clutter, leaving a visual ascii representation similar to email and usenet.
# This is an H1 headline.
## This is an H2 headline
* An unordered list element.
* Another list element
* Nested list element
[Text to show as a link](web reference)
You get the drift.
You can't do complicated stuff directly, but most of the time you don't need to.
Template Toolkit allows you to build frames to dump the markdown into.
While a bit clunky, it has enough logic in it to build a menu system from a directory structure.
Layout is done entirely with CSS.
The effect of this is almost complete separation of design from content.
The latest version of MD is MultiMarkdown 3, which in addition to creating HTML has options for creating PDF, LaTeX. Which means taht if you design it properly, you can have one set of easily editable source documents that can provide both web pages and print ready, page numbered manuals.
It's not a complete system. Layout is still a matter of much tweaking of the CSS file, and reloading. Scripts are as ugly as ever.
If this were used for ranking, then I would expect web masters to attribute articles to Big Names.
I would hope that Google would have a policy of fingerprinting the articles. Most people's writing style is sufficiently unique that claiming that someone else wrote Foo is fairly obvious on analysis.
I hope also that there is a search tool so that I can find all articles attributed to me.
And suppose that Slashdot and phpBB support this tag so that I can find all the posts by a given author.
I would think that the police could be even more effective by published 3-4 times as many checkstops as they have.
"Yes, we planned to have one there, but the manpower was needed eslewhere so that one got cancelled..."
Similar to the 'cry wolfs'. If you were a police department you could get a device that would broadcast signals on police radar frequencies, tripping every radar detector for a half a mile or so. It was an easy way to get speed demons to slow down.
The only difference in this one is that the reacting material is a suspended, particular solid. There are commercial systems that use flow through batteries already.
See:
Flow batteries have historically had low energy densities, as the entire reacting mass was in solution. E.g. the article above cites 25-75 Wh/kg
Which means a 50 kWh system would require roughly a cubic meter of tank, or about 5 standard drums. This is acceptable for a house, unworkable for a car. (50 kWh is about 2 days average power for a non-electrically heated house. This size unit would give impressive load level capability to the grid, and would be awesome for off grid use.
Going to a slurry increases the energy density, but possibly with problems with charging. How do you keep the particles the same size during the charge cycle. In many conventional cell batteries, differential plating on charge is often the limiting factor on the number of cycles.
In principle, the working solution could be filtered and the large chunks broken up. From the description of the press release, however the slurry has the consistency of light grease. Filtering would be problematic.
This may require periodic extraction and reprocessing of the working material. Still, this should be easier than rebuilding a battery.
So, you use a device that is hard to use. Suppose you built into a ring. The ring has the ability to interface with a socket, and transmit a password. It's voice activated. "Present slashdot password" touch to socket.
You could also use it to store your public key and exchange it with other ring holders.
Of course it's new technology so initially the rings would need to be very cheap. And it will be the kids who will be early adopters. I suggest that they be given away in boxes of cereal.
Who knew that Captain Midnight Secret Decoder Rings could become reality?
Locally Future Shop pays their people on commission. Best Buy pays straight scale.
I shop at Best Buy. I'm far more likely to get good advice (for a sufficiently small value of good) than at FS where I will be steered to the most expensive alternative.
Indeed, I've had Best Buy people say, "Don't get that here. Get Foobar instead, at Acme Electronics"
In the long run the BB approach generates store loyalty, while FS generates short term higher profit.
Really bad old days:
The latest thing was the replacement of relays with vacuum tubes. While an individual tube is reasonably reliable -- better than a light bulb -- collect a bunch together and you always have one blowing out.
I heard that with one such school sized computer after WWII the basic flip-flop was a 7 vacuum tube circuit, wired so that ANY two of the tubes could fail and the device would still function.
Several flip-flops were in a drawer. Indicator lights on the drawer front showed status. Red light on drawer, open drawer. Each flip-flop board had indicator lights showing which tubes were good.
People would race though the corridors with shopping carts of tubes doing hot replacements while the machine was running. My recollection was that even so uptime was usually measured in minutes before some other thing would break.
Good advice previously.
In general, if you aren't the sole admin user on the box, you shouldn't play. You can co-locate, sure. But you must be the person who is responsible for everything on that box. The ONLY access the co-location provider should be able to have is the ability to execute a shutdown script to protect themselves from your rogue processes.
If you are not the sole owner, then you are dependent on their security being up to snuff.
If your application is small, you may be able to work with a virtual cluster of machines running on a single box.
I would give serious consideration to encrypting all sensitive information in the database. Even if the keys are wired into the application, it means that data extraction has to work through the application to get data.
I would also give serious consideration to putting large quantities of meaningless information into the database allong with some means of filtering for your own use. A Black hat who gets 2 million email addresses from your machine, but finds that 99% of them are bogus will not earn the gratitude of his client.
In Alberta, peak power use is the supper hour, followed by the shit, shower and shave time block in the morning. Late at night is the low point.
I recently dropped my power bill by $30 bucks a month by having the 50 Amp pole breaker replaced with a 35 A. one.
I've yet to trip it. I figure it will trip if I have the oven on, and start the table saw the same moment that the well pump kicks
A unit that could average power consumption would be a big deal to the utilities. Would mean far smaller wires needed for the distribution network. Consider: A standard service right now is 100 to 200 A. Which means 4 guage wire to the house. It has to be sized to handle the possibility of 20 KW. Most people, however use under 1000 kwhr/month, which is is on the order of 1 kW continuous. So having a box that:
A: Could hold roughly 1 day's power.
B: Could be charged at a rate roughly 1.5 to 2 times the average consumption rate of the customer.
C: Could be programmed by the utility company as to when to charge
would be a big win.
As a side effect, consider that in a power outage, everyone has a day's power at normal consumption, and with some frugality perhaps as much as a weeks.
Unless the carts are very sophisticated, it's an invitation to shop lifting.
Scan one item into your cart, place 6 in your cart. Avoiding this would required that you have a scale in the cart accurate enough to register a single item's difference. If they come up with this, then just add a squirming kid to the basket. (fresh pork...)
Or you scan a can of el-cheapo discount icecream, but place Haagan Das in the cart. Or you fill a bag with cashews at the bulk bins, and tell the machine that it's wheat bran. Scan bananas, load avocadoes.
The ISP has to keep records.
Ok.
Now if I were an ISP and wanted to protect my users, I'd try to make these records as un-useful as possible:
1. Shorten the DHCP lease time.
2. Hack the dhcp server code so that it ignored the renew my address, but passed out a new address each time.
3. At the same time, hack the logging code so that time stamps were in microfortnights since last dhcp server restart.
4. In addition, tweak the format so that it was not trivial to parse. The easiest way to do this would be to have variable numbers of fields. Having multiple dhcp servers logging to a common log file with differing formats would also work.
5. Run your servers on Newfoundland time.
6. Do NOT keep servers time synchronized.
7. Use layers of NAT so that a client address at the ISP office level corresponds to hundreds of users.
Whenever a request came in, you would print a relevant block of data, but to save paper the log files would be printed in 6 point Squint font, and the block of data provided would run hundreds of pages. The output data format is neither by time stamp nor IP address so that someone has to wade thorugh the entire block.
What else could an less than cooperative IPS do?
Guru as team lead probably is NOT a good idea. Again, team lead is at least partially a social skill. I've known several guru class coders, and I wouldn't like to have any of them as bosses.
I can see merit in having a master/apprentice system too.
An experienced coder has a young apprentice. The apprentice does the scut work: Making the code pretty to whatever the flavour of the week is, reading the inline documentation, and adding to it as needed. (Or writing it from scratch...) Writing test suites. Assisting in the debug process.
Apprentices should be moved from journeyman to journeyman, say every 6 months. In this way they become familiar with various parts of the code base. They see a bunch of different work styles. They get practice at coming up to speed with other people's code. In addition they act as pollenators. "Mike showed me a neat trick with...."
After several years of this, they become journeymen. But now journeymen are paired to journeymen. The PAIR has to sign off on a milestone. For each segment, one codes, one documents/tests, both debug.
After a few years of these, if there is merit, they become masters, and are paired with an apprentice.
Some journeymen may not be master quality. At this point they are diverted into sales, and HR, and support....
The best way I can see to verify an account is with a system that phones you. Most people do not have more than a single land line and a cell phone. Often the land line is shared.
On the first N postings, the system phones your # of record, and simultaneously shows a 5 digit number on the screen. You respond to the phone call by punching in the number.
After N calls, you are verified only every M postings, where the value of M is inversely proportional to your karma.
I don't think that the Slashdot system works well at all. Way too many posts have low signal to noise ratios.
Slashdot has several issues:
1. A story can be modded up for either being funny or for being informative. In general the ones that are 'funny' are usually no more than distractions.
2. Too many of the comments are not well thought out. Many are too short, and boiled down consist of 'I agreee' or 'I disagree'
Ways I would like to see slashdot improved:
1. No mod points for funny.
2. It costs you X points to post. You are given N points per day, and they accumulate up to a max of M points. Short posts -- under 100 words, say -- cost more.
3. Threads are ranked by karma/post. People who have been modded up in the past are more worth listening too now.
4. Thread length is limited.
5. Threads are owned by the thread starter, who automatically has mod authority on the thread.
6. Thread starters have a responsibility to provide a summary of the discussion after the thread has closed. You can't post until you summary is complete. Summaries are also modded up or down. If you posted to a thread, you have an obligation to read and mark the summary.
The idea is to make it 'expensive' to post, so that posting isn't done for trivial reasons.
Part 2.
One of the things I like about slashdot is the usual cogent summary, a link to a meaningful story. However:
1. I wish I could subscribe to slashdot by sections. I have zero interest in games, lots of interest in technology. This could be done automatically by having an up/down set of buttons much like firehose that would increment/decrement all the tags associated with a story. Additinally, having a link on your profile where you could manually set the point values of every tag. One of the responsibilities of contributors would be to tag stories appropriately.
2. I wish the moderators would be a little more discriminating about stories. Sometimes we have the same story coming up several times in the space of a few days. I would much prefer a somewhat longer story, with several links to external articles. In particular the links to 'gee whiz' articles that have no content, and don't really have a break through behind them are anathema.
3. No more slashvertisements.
To me this concept is exciting: It's a way of collecting AND STORING solar energy.
Using the Mississippi at Itasca is a bit disingenuous. That's fairly close to the headwaters, where it is still small.
At the bottom end, the Mississippi has an annual discharge of about 500 km3/year, which corresponds to about 4.2 million gal/s
So that means you can't put one everywhere. Hydro power is a scale problem too. In hot climates this may be a good way to store power:
Sunlight evaporates ponds. Generate power as needed from sea water + brine pond.
There are a fair number of big rivers that could be tapped.
In addition the Dead Sea, and parts of the Sahara are below sea level.
Death Valley is probably too picturesque, but the Salton Sea may be suitable. If you want to do this on a large scale dike off the Sea of Cortez (between the Baha peninsula and the mainland Mexican coast.
The much larger gradient between ocean water and the Dead Sea means more watts per gallon of flow.
If done cleverly, the energy from mixing a few gallons of ocean water with a few gallons of brine may be sufficient to desalinate a few gallons of ocean water -- in part of the world where fresh water is scarce.
Looking at it another way if 13000 gal/s = 100 MW you're getting about 7500 J/gallon.
A gallon is about 8 lbs. Call it 40 newtons. So the energy is roughly equivalent to a lift of 200 m.
Now if the energy to be gained is proportional to the amount of dissolved salt, then for concentrated brine (about 30%) there should be about 75000 J/gallon.
Couple this with the idea of tower evaporators:
Desert climate: Build a tall tower in the middle of a large shallow pond. In the day, the pond gets warm, evaporates. Pump warm water to the top of the tower. Drop as a mist. Evaporative cooling makes the air heavier. Choose water droplet size so that about 3/4 of the water evaporates. Now you have cold wet air coming out of the base of the tower. At night this can create a heavy dew -- lots of desert plants can survive on dew.
The wind from the outrushing air may be enough to run the pumps. The brine + seawater runs the salt water battery system.
Salt can either be used as a feed stock for various chemical processes or stacked out of the way, like surplus sufur is now.
Remuneration for Company Officers
I can hear the squeals from here...
Right now corporate executives of large companies are paid very large salaries with even larger bonuses, bonuses that they seem to get even as they waddle up to the trough for a government bailout.
There is no incentive to think beyond the next quarter, or at most, next bonus period. I watched an interview on one of the news documentary shows where an official admitted he knew the bubble would burst, but was going to make money as long as he could, then walk away.
Never mind bears and bulls, Wall Street seems to be owned by pigs. I'm sure I'm casting aspersions on many solid officers. Bad apples and all that. But where is the incentive to think beyond the next annual report?
Here's a solution. Tell me why it won't work:
Suppose that Company officers were paid a reasonable stipend -- no better than, say, a good doctor or senior engineer, but also received the dividends from a block of common shares of the company for the next 20 years. They would not receive the stock. Only the dividents from the stock. Or if they did receive the stock, it could not be sold until the 20 years was up.
So I get an offer to be CEO of Almagamated Consolidated. I get an up front salary of $200,000 per year. I also get the dividend from 100,000 shares of AlCo. Now AlCo has been a good company, and pays $4 a year in dividends. This, in effect, triples my salary. And, even if I quit AlCo this year, I will get another 19 years of dividends from AlCo. If I do a good job of running AlCo and the dividend rises to $6, then I've raised my salary for that year by a healthy amount.
Next year I get another 200,000 salary, and another block of shares' dividends.
At the end of 20 years I'm getting dividends from 2,000,000 shares of Almalgamated Consolidated. If I've done a good job of running the company, this should be a tidy bit of change. If ownership of the shares was part of the package at this point I can now sell the first block of shares. But even if I retire today, the payments will continue for another 20 years, and I'll be holding shares for another 20 years. I'd better leave the company in good shape with a good man at the helm.
Look at what this does:
The people who run the company now have a huge incentive not to look at the next quarter's earnings, but to try to plan for the long haul.
I won't guarantee that I've picked the right numbers. Depending on the goals of the company it might be 10 years. Or 30 years. It's not a stock option. It's a stock compulsion.
Indeed I think that this form of remuneration should carry down through the ranks. This would build a financial backing to company loyalty as well as make life run more smoothly when layoffs happen.
And wouldn't that be an exciting change?
Until you are audited by the tax demons. In Canada they can audit for 7 years back, further if they suspect deliberate tax evasion. If you are running a business, you keep everything.
For us this amounts to a 3" thick folder of folders per year.
I also keep bills of sale for anything I've bought that is more than 'disposable' E.g -- tractor, truck, auger. So if there is ever a dispute about ownership, I can show good faith.
Mr. Landis's comment is apropos.
Announcing this as a 'breakthrough' is the issue.
Unspoken however is the reasoning behind using such a crummy cell in the first place.
My suspicion is that they had a technique that had the potential for reducing internal resistance in cells, so they used a material that had a high internal resistance so that their efforts would show up in a way that could be measured.
If this technique can be applied to other cells, the effect is likely to be additive, not multiplicative -- ie it will take a 20% efficient cell up to 21. And be just as you pointed it it will be a small step forward.
I'm not wealthy. My small business doesn't even make me minimum wage. Yet of all the ways that the Government can spend money, well chosen basic research is high on my list.
Really basic research cannot be managed like a production line. It's closer to being prospecting, except the long term payoff is better. Most of the time you really don't know what you are going to discover. (If you did, it wouldn't be basic research)
It's not clear to me what the best way to fund it is, but I can the following as possibilities.
1. Multiple agencies similar to the National Science Foundation and the National Institute of Health. Their funding should be cyclic, varying about 60% over the time span of 3 grant cycles. This prevents a lot of the "I got it last time, so I'm a shoo-in this time" Each cycle, the sponsoring agency has to prune 2/3 of the deadwood. Having multiple agencies gives you flexibility to change your priorities, and encourages PIs to not be so specialized.
2. Industry consortiums. Any industry that had more than 5 participants could form a research consortium. They pool research money, and get a tax break on that. They have a board that can award research grants. All results are shared amoung the research pool first. For any given project the consortium can decide that it is valuable research, and can put a lid on it, keeping it internal, but to do so they have for forego tax writeoffs for the project until they open and publish the research. This requires some serious auditing
The limitation on numbers is mostly so that one firm cannot dominate the consortium.
3. More prizes. These focus attention on science and research.
4. As a result of automation, we are getting more and more people who don't really have meaningful work. The number of really brilliant scientists are few. We need a research system that encourages one PI to mentor/direct the not quite so bright ones. And we need to figure out ways that the PI doesn't spend his life as an administrator.
The transition to metric in Canada has been going on for decades. It's a pain in the butt.
One of the advantages of the imperial system was that more things are chunked in convenient units. Tinned veggies came in 12 oz and 20 oz cans. Now we get 373 ml cans. 350 ml cans. 342 ml cans.350 gram packages of frozen vedge. 500 gram, and kilogram packages.
Milk was mandated by law to move into integral liters. The milk industry used this as an opportunity to resize milk crates. Where they used to have an internal dimension perfect for moving LP records, now they are about 3/16 (sorry 5 mm) too small.
Meat is advertised at dollars per pound, but labeled in the store at dollars per kilogram. Cheese is an international product. In the same store, I can find it priced per pound, priced per hundred grams, and priced per kilogram all on the same shelf.
Potatoes can be bought in a 5 or 10 kg bag, or if they are from the US in a 10 or 20 lb bag with the metric unit in very small print somewhere. Dog food used to be universally in 40 lb bags. Now you can get it in 8 Kg, 18 kg, 44 lb (20 kg) and 50 lb bags.
If a company didn't want to change their production line, they just bought new labels. So you can buy a quart of something and get:
1000 ml if they did convert.
946.35 ml if they were a U.S. company and relabeled their quart.
1136.52 ml if they were packaging in British (Imperial) quarts.
Doing comparison shopping between 946 and 1136 is a challenge. If you do a lot of it, you end up with approximate conversion factors in your head, A british quart to is 6/5 of an american quart.
Our road signs are in kilometers. But the rural road network is in units of miles. Many people still think of milage in miles per gallon, even though we buy gasoline in liters and measure distance in kilometers. Andd officially mileage figures are in liters/hundred kilometers -- an inverse unit, so to convert you have to do do mental long division.
The housing industry revolted when they proposed changing the sizes of dimension lumber. Sheet goods were going to go from 4x8 feet to 125 x 250 cm. A 2x4 would have gone from 44.45 x 88.9 mm to 45x90 mm Which would mean a reno project would require every board to be cut to match. Or it would mean that the mills and the slupply stores would have twice the inventory, and things that were so close would be constantly swapped.
They reached a compromise. Most of the building trade was left alone, but sheet goods are in thickness measured in mm. So you can't buy a 4x8 x 1/2" sheet of oriented strand board. It's 4 feet x 8 feet by 12 mm.
BUT All the building codes are in metric. AND the plans you submit for a permit have to be be in metric. And for some obscure reason they settled on mm. I have seen property plans done with mm's as units. E.g 20,000 x 40,000 mm instead of 20 x 40 m.
Hydraulics, already a maze became twice the hassle. I know one hydraulics shop, you bring something in with a metric fitting, he says, "That's metric. Take it in to the city." It's not worth his while to double his inventory.
Plants only use about 1% of the energy in sunlight.
Most plants, the problem is CO2 availability, water, and heat.
Plants in a greenhouse can absorb ALL of the CO2 in a few minutes. One of the reasons that greenhouses are expensive to run is replacing and heating all that air. And you can get signficantly higher production by adding CO2 to the air.
Photosynthesis slows down as the temperature goes up. Reason: Evaporation picks up, and the plants close the stomata to reduce water loss.
The energy cycle to convert light into sugar is complex. Not all frequencies can be used effectively. Plants are green because that's the peak of the frequencies it doesn't use.
I'll point out that an aerial view of the boreal forest is close to black. While an individual spruce needle is green, the light that misses a needle has to bounce too many times to return to the sky. So in terms of keeping warm, spruce trees are doing a pretty good job. You'll note that spruce are not common at low latitudes.
Oh, and the purple and dark red leaved ornamental trees: In most of these, the leaf is overproducing purple and red pigments in excess, masking the chlorophyll. It is a disadvantage to them, making them less efficient at photosynthesis. These plants tend to grow more slowly, and be more disease prone, due to the shorter energy supply.
The problem is that there won't be useful work for a large number of people. Education can help, but there is still 50% of the population that is below average in smarts.
Now you can reduce this fraction somewhat by using different kinds of smarts. At present there is no good way to outsource or automate certain classes of jobs. It will be a while before we automate truck driving to the point where there is no driver in the cab. Outsourcing plumbing repair or furnace installation is a ways away.
Stuff is getting cheaper. Thanks to automated factories, outsourced factories, machine intelligence. For the same reasons stuff is getting better quality, although there is still junk out there.
So where does it end? Historically when we have made big gains in productivity, the work week got shorter. But the intellectual jobs, the ones with high salaries, are ones with long hours. Ask any sysadmin, and find out how many of them are working 32 hour weeks.
At the low end, here in Canada, many brick and mortar stores have 90% part time workers -- typically working 15 to 20 hours a week. If they are part time, they don't have to pay benefits. So people end up juggling 3 part time jobs to make ends meet.
If you want to see where it is going, take a tour on most Indian reservations. See the joy and the bliss of communities that have 70% unemployment, where the entire community economy is dependent directly or indirectly on government handouts.
Education is a partial answer. But here in Canada we're finding that a university degree doesn't necessarily get you a job. Indeed, do we need 75% of the population with a university degree? Some fields do. One of my former students recently finished his civil engineering degree, and was snapped up at 70,000/year. But for each guy like him there are a dozen who move back home after University. But not everyone has the savvy to be an engineer.
Our local tech school advertises, "Now that you have your degree, come and learn a trade so you can make some money and pay off those loans" And people are doing that in droves. Go back, and take training to become an electrician, welder, surveyor. A computer geek working for Best Buy can get $13/hour, and probably works 15 hours a week with no benefits. A high school kid with a grade 10 education can get a job rebuilding drill bits for $18/hour, plus $2/hour shift differential, plus 10-20 hours of overtime, plus 30-40% bonuses every 3 months. A welder can expect 27-32/hour.
Energy has to get a lot more expensive before it becomes more economical to do things locally again. Off hand, I'd guess 2 orders of magnitude more expensive. Right now it is cheaper to move apples from New Zeeland to Canada than it is to store apples in Canada. At the same time, my uncle, who grows apples in Washington state, ships containers of apples to India. A few thousand dollars takes a container anywhere in the world. The cost of the contents has to be small compared to that cost before the economics favour local production again. Apples wholesale around $10/box. A box is what, 2 cubic feet. A container is 8x8x40 feet -- 10,000 cubic feet. So the apples are worth 50,000. Shipping is 3-6000. Remember that rising energy costs will increase both prices.
H. G. Wells commented that 'progress is a race between education and catastrophy' It's not clear to me that education is winning, or if, indeed it can win.