This is common in education: people want to teach X, so they bluntly approach X.
I recently had a conversation with K.A. Ericsson--a brilliant man who has driven research in the area of learning and developing expertise--and our exchange included a minor discussion of using his Deliberate Practice in education. His interest was largely targeted at incorporating better educational techniques into the classroom, and he lamented that he only saw an improvement in *learning* and not an improvement in the long-term career of the students; I debated the merits of teaching the students the techniques of learning, rather than simply using them.
To put it simply: he found a better way to teach programming, mathematics, music, and medical diagnosis; I told him he should teach *that* *method* to students so they can independently plan their approach to new information and skills. We didn't debate so much as discuss. His largest problem in the field is the scientific one: it's ridiculously hard to measure those things, and he's a scientist and needs to formally identify empirical results. I'm an engineer: at the cost of a constant drift toward dogma (which, hopefully, the scientists can counterbalance with their research), I take pieces and assumptions such as "education improves a person's ability to thrive" and work on the individual pieces along the way, assuming that optimizing each input will optimize the output, which acts as an input to the next piece. In other words: rather than asking, "If I make people more academically capable, do they actually become more effective and advantaged in their adult lives?", I simply say, "We must give these people the ability to learn more effectively so they have more tools to use in their adult lives, thus giving them an advantage."
Now go back to the original topic: What's a good software engineering project to teach high school kids programming?
If you want to be a software engineer, you should know about planning, risk, and research. You need a cursory grasp of project management, specifically of defining requirements and breaking down the work; you don't need EVM, schedule management, human resources management, procurement, or all that other stuff. You need to learn to make decisions about architecture. You need not only languages, but methods.
People want to teach students programming languages and call it programming. "Can you construct a conditional loop?" "Can you make a class, and call object members?" They think that's the starting point. It's ludicrous, like teaching someone construction by starting them on building a small shed, instead of on the fundamentals of wood and fasteners and the proper architecture of a stable structure. Apprentices are kept under strict watch and taught as they carry out tasks of increasing complexity, not loosed on their own to build rickety hovels that pass as homes.
When I learned what little I know about plumbing, the first lessons involved copper pipe and an acetylene torch; we joined six inch lengths of pipes together with flux and solder. The construction students went on to learn about routing water supply and waste systems, using PVC and CPVC, constructing stable floors, and architecting a complete building with proper building practices. It was not about building houses, but about knowing how houses were built and why they were stable.
You must learn planning, architecture, and design patterns before you learn to build something which uses planning, architecture, and design patterns. The common-sense answer that the student can somehow force something simple to work is broken: this teaches the student bad habits of thinking, and forces them to do excessive work laboring to produce something with little understanding of how to go about it. You should focu
Corporate taxes are near 40% in the US. Free trade tariff of 11% is unimpressive, but I get the difference in income tax versus a tariff: a tariff behaves more like an increased labor cost (it raises the monetary expenditure of production, same as having no tax and increasing wages or labor time per volume production). You can compare these as long as you don't take them as direct analogs.
The US government frequently works by bribes, power plays, and other unsound behavior. This brings favoritism leading to economic damage. In a nation such as DRC or DPRK, such behavior would create a relatively-rich noble class and a body of half-starved peasants; here it is less significant. That is what I meant when I said an industrial society would handle that just fine: taking a sizable chunk of enormous wealth is different than taking a moderate chunk of insubstantial wealth.
These are kind of cyclical methods of reasoning, anyway: an industrialized society facing enough pressure to hobble it would collapse back into a pre-industrial society; the people would not simply become poor slaves shoveling 97% of their wealth to the aristocracy.
So they have political troubles. An industrial society would deal with that--have you seen the tax structure and government corruption in the United States?--and still get along fine. An industrial society would also already have active markets, with high demand for products.
As humans designed new production methods using machines, the demand for steel increased. Note that a production method requiring 200 labor-hours won't replace itself with a machine-driven production method requiring 50 labor-hours and 250 labor-hours of machine construction and fueling (the 50 labor-hours are, of course, the time spent operating the machine--instead of 4 times as long hand-weaving cloth), all things (e.g. wages) equal. New steel-making methods and efficient engines reducing that to 50 plus 100--a total of 150--make those machines viable, leaving 50 labor-hours to do other things, which would divide up between e.g. 17 labor-hours of operating machines to make new products and 33 labor-hours of work building, maintaining, and fueling those machines. You've now got the labor of an entire person devoted to making one more product, without cutting back on making anything else, *and* ignoring the making of all this steelwork.
Simply dump materials onto a society and you get a market with more product than demand. Market saturation doesn't create jobs, and wastes labor producing a product nobody needs.
*Successfully* transition to a more industrial society too quickly and you get the Industrial Revolution. I advise you to read up on this. The argument about automation--that it will end all jobs and create an economic apocalypse--is blind to the realities of past events such as the Industrial Revolution.
If we transition onto the new machines over decades, we're just facing business as usual: a few jobs lost, concentration of buying power, new markets to sell goods (those consumers have unspent income), expanding production to capture those markets, and expanding employment to fuel that production (thus *replacing* the lost jobs). Everyone ends up with roughly the same proportional distribution of income; however, with the ability to make products with much less labor, your 0.0000000005% of the income translates to buying more (because it translates to 5.0e-9 of all stuff produced, and we produce more stuff).
If we transition onto this shit in a huge rush, we quickly raise unemployment to 50% or 70% or 95%. Jobs are lost at a rate of 1,000,000/month and gained at a rate of 3,000/month; do this for a few months and you have millions of jobs lost and growing unemployment. Those unemployed are no longer a reachable market, so demand for goods drops off; you cut production, and you're now poorer (lower per-capita production). This removes jobs even more quickly. Eventually, you're at a high rate of unemployment with no fast way to dig your way out of it.
All actions have consequences. There are always sets of actions whose consequences are completely offset by their benefits; we don't always *know* any of those sets, but they exist. I never said DRC wasn't bad; I said blindly hurling money, education, or international policies against child and slave labor at an underdeveloped nation has negative consequences. You will find a need to plan your interactions meticulously; you'll have to decide your political position, too: diplomatic negotiation, forced invasion, or a campaign to incite revolution and topple the government? All of these incur death, either by the delay (people are starving), the casualties of war, or the bloody consequences of a violent uprising. Your actual plans will also send working men to the sidelines, temporarily, where they may starve to death from time to time; over time, the general quality of life increases, and you pay the cost of greasing the machine with the blood of the lowest laborers.
This is why we have welfare in wealthy nations: your lowest, poorest class of laborer necessarily spend time without jobs. No plan of full employment ope
The qualifier is retarded. Information work is useful, but not "because everyone in America are innovators". We are still making things; they're less tangible.
You need a market before you can have progress. Granted, improving infrastructure can create productivity; that requires an investment of wealth, which requires productivity. Either they labor and produce and trade their labor, thus increasing their wealth by increasing their productive output per capita (and trading it to other nations to buy things they have no means to produce), and then apply this wealth to the development of new, more productive means; or we essentially invade, claim the mantle of benevolent dictators, bring resources to legitimize our control, and build their infrastructure for them.
The latter model is a way to destroy a culture, but it does bring change more quickly. It has the secondary effect of reducing deviation between various economic units; that is, it makes more countries and more groups of people *the* *same*. That lessens the broad spread of information available to the world via varied methods of thinking across cultural bounds, which ultimately reduces the likelihood of any given incremental economic advancement. This isn't a hard fact: it doesn't automatically make us poorer; it simply sterilizes the world somewhat, narrowing the range of things that are likely to happen, both good and bad.
In any case, simple education isn't enough. It's a start, so long as you're smart enough to educate them historically: teach them to use old, outdated, low-productivity methods which will turn them into a poor, medieval nation, but which they can actually achieve. If you teach them high-industrialized techniques, they won't have the capital to go about building the infrastructure. This is a decent compromise, and can be worked into a well-engineered plan to maximize the aji of a growing civilization without distorting its culture too much. Realize, however, that it won't erase hunger and harsh living conditions as quickly as coming into their countries and stomping the populous into the ground until they bow to your imperial rule and let you fix their outdated infrastructure.
Pointedly, none of these plans allow an *immediate* end to child labor. If you cut that off before *integrating* an equivalent method of productivity--the ability to produce the same goods with less labor, thus allowing the children to play and the adults to work--you will *reduce* the productive output per capita. That means less for each person, meaning worse living conditions: they have fewer goods per person, such as food, housing, clothing, and the like.
Once you've integrated a method of productivity providing gain, you have free labor. If the children are no longer necessary *and* you can spare some of the adults while meeting the same productive output, then you can convert those adults into teachers. This is the basis of a modernized society. From there, you can probably just leave, and they'll figure the rest out on their own.
Tell me this: If their economy is so great that jobs are abound, that children need not work, that adults enjoy the great luxuries of running water and steak dinner every night, then *why* do they send their children to the mines? What sadistic creature would send their own child to labor under heathens in unsanitary, deleterious conditions unfit for man or beast?
I'm sure you feel very proud of yourself for sending these children off to starve, saving them the pain of calloused hands, abused joints, and scarred lungs. It seems very ethical to take away a man's food when that food is not fit for men of dignity, and leave him hungry so as to save him from such savagery.
That's a distinctly odd case. Such cases have appeared before court many times--notably Apple v. Microsoft--and the courts have frequently claimed interface is not copyrightable. This is the basis of UNIX, as Bell Labs was unable to sue anyone for implementing a UNIX interface or defining a POSIX standard.
I don't understand how "structure, sequence, and organization" is in any way separable from an API. An API *is* structure, sequence, and organization.
You've only qualified the claims of air quality by pointing out that we're doing something about air quality. I live near 13 coal plants--one is a multi-gigawatt plant--and our air quality is decent; it could be better. I don't live in Beijing, I know that much.
The funny thing is it's still a non-issue. People aren't toxicologists, and they look at a 0.0012ppm increase in an atmosphere with 0.023ppm NOx and go, "OMG TEH CARZ WILL R KILL UZ ALL!" and talk about how poisonous these emissions are. This doesn't even account for either that *everyone* is doing this (you're not going to suddenly see tons of shit pumped into the atmosphere), that we have long-term atmospheric measurements (so it *hasn't* caused a problem), or that the concentration of NOx around the cars themselves is extremely high compared to anywhere else (highways and city air both carry *way* more of these emissions than anywhere with less-dense traffic, and even carry significantly less at night).
We've stumbled over a problem of non-compliance, not a problem of pollution. Pride, face, and the long-term goal of improved air quality tell us to squash this; however, logic and reasoning tell us to examine the performance benefits of these platforms--notably, extending our fuel supply--and weigh them against the cost. Have we reduced emissions enough to shift focus to fuel supply extension? Can we level off a little and aim for not guzzling so much gasoline? Those questions require us to admit we may have been wrong for a while--right initially, but then continuing along the same path when conditions changed.
It can't have *no* impact on the US economy. VW sells higher-quality cars at lower prices, meaning every American can buy a car and SOME OTHER STUFF, which puts more purchasing power into the hands of Americans. This spreads employment (creates jobs).
That's a global consideration, of course. It may create more jobs in other nations at the expense of other domestic jobs. The consideration at hand locally is the fluctuation of domestic job proportions: are we 90% local and 10% export, or 80% local 20% export, or what?
We find ways to export jobs, reducing the cost of products (by cost of labor reduction), thus reducing the spending of each consumer, thus leaving money in every consumer pocket, thus allowing us to sell them new goods, thus creating new jobs. Often, this doesn't change the balance: we wind up with 10% more jobs, and the same proportion are created locally as in China, and so we end up offshoring 30,000 American jobs to China and then creating 40,000 *new* American jobs and 5,000 new Chinese jobs, and thus end up with *more* local employment. It often *does* change the balance, as well, and we can offshore enough of the new jobs to create a local rise in unemployment (this is limited by the reduction of the consumer market, but it's even more limited by the simple fact that we're *constantly* in a state of having offshored as much as we could figure out how at a profit, and only do more when we find a new way to save costs with that mechanism).
The practical result of losing an import product, if the import product is cheaper than domestic, is a reduction in local employment and in standard-of-living at all income classes. Ricardo's theory of comparative advantage suggests this in fairly loose terms; I conclude the same using new methods.
Understand that the above is *not* well-formed theory. I have very well-formed economic theories at this point; however, labor export and exchange rates are two particular topics I've been putting a lot of theoretical research into. I have a loose sense of how these work, and can put it into words; I *can't* produce a concrete model (yet), so the overall idea is *probably* (but not guaranteed) correct, while many of the details of any lengthy explanation likely have flaws. I'll buff them out eventually.
Esperanto's structure is by context of word form. It's essentially word salad, and the words have affixes telling you if they're subject, verb, or direct object; they can come in fairly arbitrary order within a clause. I prefer the rigid structure of Japanese (as well as its extreme contextuality), but I must admit Esperanto does a fair job of telling you what you're talking *about*.
I doubt you could claim copyright on a language. It's information, in a generalized form: claiming copyright to a language is the same as claiming copyright on Microsoft software and then suing people who write books or design courseware to teach Microsoft software administration. Copyright doesn't cover ideas and information, but rather the form of expression of those ideas and information: poetry, textbooks, songs, paintings. The more inherently natural something is, the harder it is to copyright: you can copyright a dictionary, but you can't copyright it so much that another person writing the same dictionary with almost the same words falls to a plagiarism case--unless you show that he did, in fact, directly copy your work.
Put these together and you get a fairly direct legal conclusion: information *about* a conlang isn't copyrighted, because it isn't created in form (textbooks, dictionaries) by the creator of the conlang; and assembled information about a conlang ranges from highly-copyrightable coursework to barely-copyrightable dictionaries. Especially with interlingual dictionaries, the form "Foreign - Domestic" produces no new information, unlike a single-language dictionary in which a word has a descriptive definition. "Beer - Cerveza" is not copyrightable; "Beer - A fermented drink made from grain, usually barley, and hops" is *entirely* copyrightable. "Beer - An alcoholic beverage brewed by boiling malted barley and other grains with hops, then fermenting for several weeks" is another definition of beer, just as valid, with a completely different form; you could not do the same for "Beer - Cerveza" if writing a Spanish-English dictionary.
WB v. RDR is different. RDR not only copied a lot of the original work (rearranging it, which is a valid form of producing new information), but also charged for it. This makes RDR a commercial derivative work incorporating a lot of source material. The final resolution was to write more original text and incorporate less of the source material.
The term "effortless" is a meaningless word used as an undefined quantifier. It functions as a qualifier to suggest a steep learning curve, without suggesting what that might mean; typically it acts as hyperbole.
I've heard quotes from 4 to 16 times as fast--25% down to 6.25% as much effort--to learn; I don't know how much I believe that, and can't test on myself because I learn more quickly than others (I'm attentive to information; it's a habit that enables me to learn faster, when I'm not being incredibly lazy).
Any thoughts on Esperanto (International), Lojban (Semantic), Solresol (Representative of French), Sindarin (Tolkien), Drow (Dungeons and Dragons), and Klingon (Star Trek)?
I've noticed Esperanto seems to produce propaedeutic effects by either loading quickly (it's *fast* to learn) or directing more attention to the analysis of a language's structure (by nature, it encourages the student to do this). It's a very structured language, in terms of word construction.
Lojban is supposed to be unambiguous; I think Esperanto achieves that exactly as well, due to its grammatical structure, in so much that Lojban is *semantically* unambiguous (we know what in the sentence represents the subject, verb, direct object, adjective, adverb, etc.) but can be *conceptually* ambiguous. Your thoughts?
This leads to things like Solresol, Sindarin, and Klingon. They all seem to have a point: Solresol encodes French to music; Sindarin is supposed to "sound pretty"; and Klingon is supposed to sound harsh. How do people come up with this kind of thing? Is that even a valid concept? Is there any interesting aspect of these sorts of languages which I should consider, or are they just as essentially bland as any other?
Nod. Interesting, but I'm sure you see the point: less labor to get more gold. Imagine if that happened to dollars in a capitalistic fashion, where anyone (not just the Central Bank) could issue dollars, and the mediator for how many dollars they could issue was how hard it was to make dollars. What kind of currency would that be? How would that impact the stability of our economy?
I happen to like fiat currencies. I originally thought they were ridiculous, but I've had time to grow up. I don't like idiots in charge of fiat currencies; they tend to think they can solve all problems by printing more money, which causes serious problems.
The Oxford comma is used precisely because English language speakers insert a spoken pause. That is to say: when you write, "To my mother, Ayn Rand and God," you speak, "To my mother, Ayn Rand, and God." You still insert the pause in your vocal inflections.
Further, the serial comma does not always introduce ambiguity, and its absence does not always remove it. "Serena, a dancer, and a scholar" tells me there are two or three people; "Serena, a dancer and a scholar" tells me there is either one person or two people and a poor writer. You could use the forms: "Serena, who was a dancer, and a scholar"; "A scholar and dancer, Serena"; or others. For a single person, "Serena, who was a dancer and a scholar." For three people, "A dancer, a scholar, and Serena." Frequently, use of a semicolon improves form.
Ambiguity is a matter of form. Removing the serial comma due to ambiguity in some phrases is akin to replacing oil with gunpowder because oil burns when exposed to spark: gunpowder explodes in its fair share of situations, and also brings along horrifying logistics in storage and transportation.
Most current style guides advocate the Oxford comma; most detractors are journalistic publications.
Well, yes. We went from picks and shovels to powered machines, so instead of 10 hours of daily labor from 600 people digging a hole we spent 10 hours of daily labor from 250 people in total producing machines, refining oil, shipping oil, operating the machines, and so forth, and got the same amount of gold.
We've also had gold mines start out nice and full of rich veins, then eventually run down as we dig out all the big chunks and leave behind the little bits. Of course we came back and got some of those little bits when we invented machines to do some of the work, since the oil and steel needed to build and power the machines took less labor than just using a pick and shovel to get the gold.
Well we know that. It has been the policy of the NY Times, The Guardian, and even The Economist to not use the serial comma. All English style guides recommend use of the serial comma (with inadequate discussion); yet traditionally-print publications object to its use. This hands down from over a century ago, when typesetting all those extra commas was significantly expensive in aggregate over thousands or millions of papers. They threw out good writing in favor of saving pennies.
You find it surprising they'd throw out good reporting in favor of emotional arguments and sensationalized narratives? Throwing out the serial comma is not the beginning of a legitimate slippery slope and should not have been your first warning; in hindsight, however, it's an obvious part of a bigger problem which we're finally aware of: absolute tunnel vision directed squarely at cost-cutting and profiteering. The missing comma isn't the beginning of the decay, the first crack in the foundation which eventually crumbled from beneath a great industry; it is only a clear representation of the extent to which they will squeeze every drop of blood they can manage from the reams of broadsheet.
Bitcoin is more of a commodity, yeah. I don't always make consistent arguments; my arguments are all ad-hoc, and losing a firm grasp of the state-of-mind that generates an argument tends to lead to a different, incomplete argument.
In this case, I forgot about the common terminology of "commodity currencies" I've used when describing gold-backed dollars and precious metal money systems. I always use that terminology when describing *gold*, for example, because it has an important connotation: if you find a new gold mine or a way to use less human labor to produce more gold (thus making an existing, low-yield mine profitable), the amount of money changes. That means you can get sudden inflation (if you can suddenly mine more gold--and thus suddenly *do* mine more gold) or sudden deflation (if your ability to produce other goods grows faster than your ability to produce new gold). You also get the insanely-complex interactions between gold as a labor-produced commodity and gold as a currency: gold has a non-trivial cost of production and is itself desirable, so it's a product, and thus a gold system is barter. (I do not like terms like "intrinsic value" because the "intrinsic value" of gold will become 0 if you find a way to magic gold out of thin air, and so we are really only discussing the labor cost of production of gold as a product.)
You can't magically find new bitcoins; you might find a new way to compute a bitcoin hash chain in polynomial time. For the most part, bitcoins don't act like a commodity in the physical sense; they do act as a commodity in the economic or financial sense. The simplification is that bitcoin is a limited-supply currency with a deflationary behavior--both more and less correct than calling it a commodity.
I didn't think deep enough into it on the first pass. Oh well.
(Yes, I did just argue that gold coin, gold-standard currency, and bitcoins are the same thing, and all bad for the same reasons.)
We're not all going to the same place, so mass transit becomes incredibly inefficient. People like to think only of fuel and operating cost, but not of time and the aggregate cost across population. Imagine if all commutes were 90 minutes longer each way.
Statistics show fewer incidents per traffic volume in roundabouts than at intersections. There are fewer than 40% as many contention points in a roundabout than at a traffic signal.
Wandering into a locale with roundabouts everywhere *will* put you under a *lot* of stress if you're not used to driving them, though. The mental approach to a roundabout is completely different. While the brain can easily handle the task, it handles it about as well the first time as it does driving in general: it strains, and then becomes confused. Once you've learned how to react to roundabouts, they're a lot easier to navigate than a city riddled with stop signs and traffic signals.
We're actually looking to replace one traffic light intersection with a large roundabout here. I expect a *lot* of complaints and unsteady drivers trying to navigate it for the first 4-6 months, and then an enormous reduction in the number of collisions. This intersection experiences frequent traffic incidents, and even ambulance route around it; fire engines crash straight through because a fire engine colliding with a passenger vehicle is just a passenger vehicle turning into confetti. (Really. In a few towns in CT, if you park in front of a fire hydrant, the fire engine will simply crash full-speed into your car to move it out of the way.)
Government built and builds roads, rails, and space ships because those infrastructure projects were *enormous* and not very profitable. When no private corporation can front the money *but* the return is huge, you ask if the Government can do it in reasonable taxation. The answer is usually "not yet," until it becomes financially feasible. Such infrastructures transition to the private sector if they don't require a centralized, non-direct-revenue administration (e.g. power lines, railways, and Internet carry a product to customers; roads cannot be monetized in that way).
When new steelmaking techniques reduced the labor required in making steel, rail became feasible. Prior to certain blast processes, laying steel for rail would have required us to stop making all the products human society was consuming and devote years to making and transporting steel; maintaining the rails *and* the engines would have been more than the entire body of human labor time available. At a point, it became technically possible... if we all stopped eating. Then, we found lower-labor methods which made it technically possible, but all the wages involved made such transit so expensive that it would be cheaper to just drag them by cart--or at least to ship them along the coast (which is what we'd done all that time).
Eventually, we developed steelmaking processes which used little enough human labor to lay down rail and operate freight transit at acceptable prices, but not within the financial capability of the private sector; and then the Puddling process (and, later the Hot Blast furnace) and iron rolling processes finally lead to commercial creation of localized rail. We didn't get a trans-continental railway until the Government stepped in with tax money in the form of bonds issued at 6% interest. The project was simply too big for any given single corporation or trans-continental partnership to fund and expect a profit, so the Government built it with taxes (taxes it collected at 6% interest, since those bonds were gaining that right up until they were paid back).
Today, we have several companies actively developing self-driving cars. They all have access to GPS map data and, besides, any widely-distributed navigational data sells for cheap enough: an enterprising corporation could gather any such information cheaply enough, then resell it at a mark-up to any car company wanting full trans-continental or global annotated GPS map data. Sensors and central collectors in cars to gather and organize the data in real-time after deployment would be, essentially, free, as they'd use the same hardware and data networks *required* to operate the cars in the first place.
The government obviously doesn't need to issue bonds and fund this shit via tax money. The Democrats want their name on the face of progress, economic freedom, new technology, and environmental friendliness (autonomous cars are more efficient drivers than lead-foot teenagers, and will probably end up being mostly electric or hybrids); the Republicans want their name on the face of firm economic support, the growth of new businesses, and the freedom to pursue the American Dream by creating a new product and establishing an empire. Either party has good reason to throw tax money at the problem: PR.
There never was any reason to think Bitcoin would be better than the existing financial system. People believed in bitcoin for ideological reasons rather than pragmatic ones for the most part.
A lot of people were on this limited currency thing, where it becomes hard to make new bitcoins. That creates a deflationary currency, which is a good way to make sure debt becomes worth more over time, rather than less. In other words: mild inflation good; hyperinflation bad; deflation potentially worse, but hyperinflation is pretty fucking bad. It's kind of like saying you're small, but a Pygmy Mouse Lemur is smaller: that's not helping your case.
Bitcoin was never a financial system anyway; it was a money-making pyramid scheme.
Slashdot Mirror
This is common in education: people want to teach X, so they bluntly approach X.
I recently had a conversation with K.A. Ericsson--a brilliant man who has driven research in the area of learning and developing expertise--and our exchange included a minor discussion of using his Deliberate Practice in education. His interest was largely targeted at incorporating better educational techniques into the classroom, and he lamented that he only saw an improvement in *learning* and not an improvement in the long-term career of the students; I debated the merits of teaching the students the techniques of learning, rather than simply using them.
To put it simply: he found a better way to teach programming, mathematics, music, and medical diagnosis; I told him he should teach *that* *method* to students so they can independently plan their approach to new information and skills. We didn't debate so much as discuss. His largest problem in the field is the scientific one: it's ridiculously hard to measure those things, and he's a scientist and needs to formally identify empirical results. I'm an engineer: at the cost of a constant drift toward dogma (which, hopefully, the scientists can counterbalance with their research), I take pieces and assumptions such as "education improves a person's ability to thrive" and work on the individual pieces along the way, assuming that optimizing each input will optimize the output, which acts as an input to the next piece. In other words: rather than asking, "If I make people more academically capable, do they actually become more effective and advantaged in their adult lives?", I simply say, "We must give these people the ability to learn more effectively so they have more tools to use in their adult lives, thus giving them an advantage."
Now go back to the original topic: What's a good software engineering project to teach high school kids programming?
If you want to be a software engineer, you should know about planning, risk, and research. You need a cursory grasp of project management, specifically of defining requirements and breaking down the work; you don't need EVM, schedule management, human resources management, procurement, or all that other stuff. You need to learn to make decisions about architecture. You need not only languages, but methods.
People want to teach students programming languages and call it programming. "Can you construct a conditional loop?" "Can you make a class, and call object members?" They think that's the starting point. It's ludicrous, like teaching someone construction by starting them on building a small shed, instead of on the fundamentals of wood and fasteners and the proper architecture of a stable structure. Apprentices are kept under strict watch and taught as they carry out tasks of increasing complexity, not loosed on their own to build rickety hovels that pass as homes.
When I learned what little I know about plumbing, the first lessons involved copper pipe and an acetylene torch; we joined six inch lengths of pipes together with flux and solder. The construction students went on to learn about routing water supply and waste systems, using PVC and CPVC, constructing stable floors, and architecting a complete building with proper building practices. It was not about building houses, but about knowing how houses were built and why they were stable.
You must learn planning, architecture, and design patterns before you learn to build something which uses planning, architecture, and design patterns. The common-sense answer that the student can somehow force something simple to work is broken: this teaches the student bad habits of thinking, and forces them to do excessive work laboring to produce something with little understanding of how to go about it. You should focu
They remind me of the Drake Equation.
Corporate taxes are near 40% in the US. Free trade tariff of 11% is unimpressive, but I get the difference in income tax versus a tariff: a tariff behaves more like an increased labor cost (it raises the monetary expenditure of production, same as having no tax and increasing wages or labor time per volume production). You can compare these as long as you don't take them as direct analogs.
The US government frequently works by bribes, power plays, and other unsound behavior. This brings favoritism leading to economic damage. In a nation such as DRC or DPRK, such behavior would create a relatively-rich noble class and a body of half-starved peasants; here it is less significant. That is what I meant when I said an industrial society would handle that just fine: taking a sizable chunk of enormous wealth is different than taking a moderate chunk of insubstantial wealth.
These are kind of cyclical methods of reasoning, anyway: an industrialized society facing enough pressure to hobble it would collapse back into a pre-industrial society; the people would not simply become poor slaves shoveling 97% of their wealth to the aristocracy.
Everyone wants fruity pebbles.
So they have political troubles. An industrial society would deal with that--have you seen the tax structure and government corruption in the United States?--and still get along fine. An industrial society would also already have active markets, with high demand for products.
As humans designed new production methods using machines, the demand for steel increased. Note that a production method requiring 200 labor-hours won't replace itself with a machine-driven production method requiring 50 labor-hours and 250 labor-hours of machine construction and fueling (the 50 labor-hours are, of course, the time spent operating the machine--instead of 4 times as long hand-weaving cloth), all things (e.g. wages) equal. New steel-making methods and efficient engines reducing that to 50 plus 100--a total of 150--make those machines viable, leaving 50 labor-hours to do other things, which would divide up between e.g. 17 labor-hours of operating machines to make new products and 33 labor-hours of work building, maintaining, and fueling those machines. You've now got the labor of an entire person devoted to making one more product, without cutting back on making anything else, *and* ignoring the making of all this steelwork.
Simply dump materials onto a society and you get a market with more product than demand. Market saturation doesn't create jobs, and wastes labor producing a product nobody needs.
*Successfully* transition to a more industrial society too quickly and you get the Industrial Revolution. I advise you to read up on this. The argument about automation--that it will end all jobs and create an economic apocalypse--is blind to the realities of past events such as the Industrial Revolution.
If we transition onto the new machines over decades, we're just facing business as usual: a few jobs lost, concentration of buying power, new markets to sell goods (those consumers have unspent income), expanding production to capture those markets, and expanding employment to fuel that production (thus *replacing* the lost jobs). Everyone ends up with roughly the same proportional distribution of income; however, with the ability to make products with much less labor, your 0.0000000005% of the income translates to buying more (because it translates to 5.0e-9 of all stuff produced, and we produce more stuff).
If we transition onto this shit in a huge rush, we quickly raise unemployment to 50% or 70% or 95%. Jobs are lost at a rate of 1,000,000/month and gained at a rate of 3,000/month; do this for a few months and you have millions of jobs lost and growing unemployment. Those unemployed are no longer a reachable market, so demand for goods drops off; you cut production, and you're now poorer (lower per-capita production). This removes jobs even more quickly. Eventually, you're at a high rate of unemployment with no fast way to dig your way out of it.
All actions have consequences. There are always sets of actions whose consequences are completely offset by their benefits; we don't always *know* any of those sets, but they exist. I never said DRC wasn't bad; I said blindly hurling money, education, or international policies against child and slave labor at an underdeveloped nation has negative consequences. You will find a need to plan your interactions meticulously; you'll have to decide your political position, too: diplomatic negotiation, forced invasion, or a campaign to incite revolution and topple the government? All of these incur death, either by the delay (people are starving), the casualties of war, or the bloody consequences of a violent uprising. Your actual plans will also send working men to the sidelines, temporarily, where they may starve to death from time to time; over time, the general quality of life increases, and you pay the cost of greasing the machine with the blood of the lowest laborers.
This is why we have welfare in wealthy nations: your lowest, poorest class of laborer necessarily spend time without jobs. No plan of full employment ope
The qualifier is retarded. Information work is useful, but not "because everyone in America are innovators". We are still making things; they're less tangible.
You need a market before you can have progress. Granted, improving infrastructure can create productivity; that requires an investment of wealth, which requires productivity. Either they labor and produce and trade their labor, thus increasing their wealth by increasing their productive output per capita (and trading it to other nations to buy things they have no means to produce), and then apply this wealth to the development of new, more productive means; or we essentially invade, claim the mantle of benevolent dictators, bring resources to legitimize our control, and build their infrastructure for them.
The latter model is a way to destroy a culture, but it does bring change more quickly. It has the secondary effect of reducing deviation between various economic units; that is, it makes more countries and more groups of people *the* *same*. That lessens the broad spread of information available to the world via varied methods of thinking across cultural bounds, which ultimately reduces the likelihood of any given incremental economic advancement. This isn't a hard fact: it doesn't automatically make us poorer; it simply sterilizes the world somewhat, narrowing the range of things that are likely to happen, both good and bad.
In any case, simple education isn't enough. It's a start, so long as you're smart enough to educate them historically: teach them to use old, outdated, low-productivity methods which will turn them into a poor, medieval nation, but which they can actually achieve. If you teach them high-industrialized techniques, they won't have the capital to go about building the infrastructure. This is a decent compromise, and can be worked into a well-engineered plan to maximize the aji of a growing civilization without distorting its culture too much. Realize, however, that it won't erase hunger and harsh living conditions as quickly as coming into their countries and stomping the populous into the ground until they bow to your imperial rule and let you fix their outdated infrastructure.
Pointedly, none of these plans allow an *immediate* end to child labor. If you cut that off before *integrating* an equivalent method of productivity--the ability to produce the same goods with less labor, thus allowing the children to play and the adults to work--you will *reduce* the productive output per capita. That means less for each person, meaning worse living conditions: they have fewer goods per person, such as food, housing, clothing, and the like.
Once you've integrated a method of productivity providing gain, you have free labor. If the children are no longer necessary *and* you can spare some of the adults while meeting the same productive output, then you can convert those adults into teachers. This is the basis of a modernized society. From there, you can probably just leave, and they'll figure the rest out on their own.
Ethically and morally sound?
Ridiculous drivel.
Tell me this: If their economy is so great that jobs are abound, that children need not work, that adults enjoy the great luxuries of running water and steak dinner every night, then *why* do they send their children to the mines? What sadistic creature would send their own child to labor under heathens in unsanitary, deleterious conditions unfit for man or beast?
I'm sure you feel very proud of yourself for sending these children off to starve, saving them the pain of calloused hands, abused joints, and scarred lungs. It seems very ethical to take away a man's food when that food is not fit for men of dignity, and leave him hungry so as to save him from such savagery.
That's a distinctly odd case. Such cases have appeared before court many times--notably Apple v. Microsoft--and the courts have frequently claimed interface is not copyrightable. This is the basis of UNIX, as Bell Labs was unable to sue anyone for implementing a UNIX interface or defining a POSIX standard.
I don't understand how "structure, sequence, and organization" is in any way separable from an API. An API *is* structure, sequence, and organization.
It's not that; the statement is so obvious it's stupid. Where else would new threats come from? Aliens? Asteroids? The sun exploding?
You've only qualified the claims of air quality by pointing out that we're doing something about air quality. I live near 13 coal plants--one is a multi-gigawatt plant--and our air quality is decent; it could be better. I don't live in Beijing, I know that much.
The funny thing is it's still a non-issue. People aren't toxicologists, and they look at a 0.0012ppm increase in an atmosphere with 0.023ppm NOx and go, "OMG TEH CARZ WILL R KILL UZ ALL!" and talk about how poisonous these emissions are. This doesn't even account for either that *everyone* is doing this (you're not going to suddenly see tons of shit pumped into the atmosphere), that we have long-term atmospheric measurements (so it *hasn't* caused a problem), or that the concentration of NOx around the cars themselves is extremely high compared to anywhere else (highways and city air both carry *way* more of these emissions than anywhere with less-dense traffic, and even carry significantly less at night).
We've stumbled over a problem of non-compliance, not a problem of pollution. Pride, face, and the long-term goal of improved air quality tell us to squash this; however, logic and reasoning tell us to examine the performance benefits of these platforms--notably, extending our fuel supply--and weigh them against the cost. Have we reduced emissions enough to shift focus to fuel supply extension? Can we level off a little and aim for not guzzling so much gasoline? Those questions require us to admit we may have been wrong for a while--right initially, but then continuing along the same path when conditions changed.
It can't have *no* impact on the US economy. VW sells higher-quality cars at lower prices, meaning every American can buy a car and SOME OTHER STUFF, which puts more purchasing power into the hands of Americans. This spreads employment (creates jobs).
That's a global consideration, of course. It may create more jobs in other nations at the expense of other domestic jobs. The consideration at hand locally is the fluctuation of domestic job proportions: are we 90% local and 10% export, or 80% local 20% export, or what?
We find ways to export jobs, reducing the cost of products (by cost of labor reduction), thus reducing the spending of each consumer, thus leaving money in every consumer pocket, thus allowing us to sell them new goods, thus creating new jobs. Often, this doesn't change the balance: we wind up with 10% more jobs, and the same proportion are created locally as in China, and so we end up offshoring 30,000 American jobs to China and then creating 40,000 *new* American jobs and 5,000 new Chinese jobs, and thus end up with *more* local employment. It often *does* change the balance, as well, and we can offshore enough of the new jobs to create a local rise in unemployment (this is limited by the reduction of the consumer market, but it's even more limited by the simple fact that we're *constantly* in a state of having offshored as much as we could figure out how at a profit, and only do more when we find a new way to save costs with that mechanism).
The practical result of losing an import product, if the import product is cheaper than domestic, is a reduction in local employment and in standard-of-living at all income classes. Ricardo's theory of comparative advantage suggests this in fairly loose terms; I conclude the same using new methods.
Understand that the above is *not* well-formed theory. I have very well-formed economic theories at this point; however, labor export and exchange rates are two particular topics I've been putting a lot of theoretical research into. I have a loose sense of how these work, and can put it into words; I *can't* produce a concrete model (yet), so the overall idea is *probably* (but not guaranteed) correct, while many of the details of any lengthy explanation likely have flaws. I'll buff them out eventually.
Esperanto's structure is by context of word form. It's essentially word salad, and the words have affixes telling you if they're subject, verb, or direct object; they can come in fairly arbitrary order within a clause. I prefer the rigid structure of Japanese (as well as its extreme contextuality), but I must admit Esperanto does a fair job of telling you what you're talking *about*.
I doubt you could claim copyright on a language. It's information, in a generalized form: claiming copyright to a language is the same as claiming copyright on Microsoft software and then suing people who write books or design courseware to teach Microsoft software administration. Copyright doesn't cover ideas and information, but rather the form of expression of those ideas and information: poetry, textbooks, songs, paintings. The more inherently natural something is, the harder it is to copyright: you can copyright a dictionary, but you can't copyright it so much that another person writing the same dictionary with almost the same words falls to a plagiarism case--unless you show that he did, in fact, directly copy your work.
Put these together and you get a fairly direct legal conclusion: information *about* a conlang isn't copyrighted, because it isn't created in form (textbooks, dictionaries) by the creator of the conlang; and assembled information about a conlang ranges from highly-copyrightable coursework to barely-copyrightable dictionaries. Especially with interlingual dictionaries, the form "Foreign - Domestic" produces no new information, unlike a single-language dictionary in which a word has a descriptive definition. "Beer - Cerveza" is not copyrightable; "Beer - A fermented drink made from grain, usually barley, and hops" is *entirely* copyrightable. "Beer - An alcoholic beverage brewed by boiling malted barley and other grains with hops, then fermenting for several weeks" is another definition of beer, just as valid, with a completely different form; you could not do the same for "Beer - Cerveza" if writing a Spanish-English dictionary.
WB v. RDR is different. RDR not only copied a lot of the original work (rearranging it, which is a valid form of producing new information), but also charged for it. This makes RDR a commercial derivative work incorporating a lot of source material. The final resolution was to write more original text and incorporate less of the source material.
The term "effortless" is a meaningless word used as an undefined quantifier. It functions as a qualifier to suggest a steep learning curve, without suggesting what that might mean; typically it acts as hyperbole.
I've heard quotes from 4 to 16 times as fast--25% down to 6.25% as much effort--to learn; I don't know how much I believe that, and can't test on myself because I learn more quickly than others (I'm attentive to information; it's a habit that enables me to learn faster, when I'm not being incredibly lazy).
Any thoughts on Esperanto (International), Lojban (Semantic), Solresol (Representative of French), Sindarin (Tolkien), Drow (Dungeons and Dragons), and Klingon (Star Trek)?
I've noticed Esperanto seems to produce propaedeutic effects by either loading quickly (it's *fast* to learn) or directing more attention to the analysis of a language's structure (by nature, it encourages the student to do this). It's a very structured language, in terms of word construction.
Lojban is supposed to be unambiguous; I think Esperanto achieves that exactly as well, due to its grammatical structure, in so much that Lojban is *semantically* unambiguous (we know what in the sentence represents the subject, verb, direct object, adjective, adverb, etc.) but can be *conceptually* ambiguous. Your thoughts?
This leads to things like Solresol, Sindarin, and Klingon. They all seem to have a point: Solresol encodes French to music; Sindarin is supposed to "sound pretty"; and Klingon is supposed to sound harsh. How do people come up with this kind of thing? Is that even a valid concept? Is there any interesting aspect of these sorts of languages which I should consider, or are they just as essentially bland as any other?
Nod. Interesting, but I'm sure you see the point: less labor to get more gold. Imagine if that happened to dollars in a capitalistic fashion, where anyone (not just the Central Bank) could issue dollars, and the mediator for how many dollars they could issue was how hard it was to make dollars. What kind of currency would that be? How would that impact the stability of our economy?
I happen to like fiat currencies. I originally thought they were ridiculous, but I've had time to grow up. I don't like idiots in charge of fiat currencies; they tend to think they can solve all problems by printing more money, which causes serious problems.
The Oxford comma is used precisely because English language speakers insert a spoken pause. That is to say: when you write, "To my mother, Ayn Rand and God," you speak, "To my mother, Ayn Rand, and God." You still insert the pause in your vocal inflections.
Further, the serial comma does not always introduce ambiguity, and its absence does not always remove it. "Serena, a dancer, and a scholar" tells me there are two or three people; "Serena, a dancer and a scholar" tells me there is either one person or two people and a poor writer. You could use the forms: "Serena, who was a dancer, and a scholar"; "A scholar and dancer, Serena"; or others. For a single person, "Serena, who was a dancer and a scholar." For three people, "A dancer, a scholar, and Serena." Frequently, use of a semicolon improves form.
Ambiguity is a matter of form. Removing the serial comma due to ambiguity in some phrases is akin to replacing oil with gunpowder because oil burns when exposed to spark: gunpowder explodes in its fair share of situations, and also brings along horrifying logistics in storage and transportation.
Most current style guides advocate the Oxford comma; most detractors are journalistic publications.
Well, yes. We went from picks and shovels to powered machines, so instead of 10 hours of daily labor from 600 people digging a hole we spent 10 hours of daily labor from 250 people in total producing machines, refining oil, shipping oil, operating the machines, and so forth, and got the same amount of gold.
We've also had gold mines start out nice and full of rich veins, then eventually run down as we dig out all the big chunks and leave behind the little bits. Of course we came back and got some of those little bits when we invented machines to do some of the work, since the oil and steel needed to build and power the machines took less labor than just using a pick and shovel to get the gold.
Well we know that. It has been the policy of the NY Times, The Guardian, and even The Economist to not use the serial comma. All English style guides recommend use of the serial comma (with inadequate discussion); yet traditionally-print publications object to its use. This hands down from over a century ago, when typesetting all those extra commas was significantly expensive in aggregate over thousands or millions of papers. They threw out good writing in favor of saving pennies.
You find it surprising they'd throw out good reporting in favor of emotional arguments and sensationalized narratives? Throwing out the serial comma is not the beginning of a legitimate slippery slope and should not have been your first warning; in hindsight, however, it's an obvious part of a bigger problem which we're finally aware of: absolute tunnel vision directed squarely at cost-cutting and profiteering. The missing comma isn't the beginning of the decay, the first crack in the foundation which eventually crumbled from beneath a great industry; it is only a clear representation of the extent to which they will squeeze every drop of blood they can manage from the reams of broadsheet.
Bitcoin is more of a commodity, yeah. I don't always make consistent arguments; my arguments are all ad-hoc, and losing a firm grasp of the state-of-mind that generates an argument tends to lead to a different, incomplete argument.
In this case, I forgot about the common terminology of "commodity currencies" I've used when describing gold-backed dollars and precious metal money systems. I always use that terminology when describing *gold*, for example, because it has an important connotation: if you find a new gold mine or a way to use less human labor to produce more gold (thus making an existing, low-yield mine profitable), the amount of money changes. That means you can get sudden inflation (if you can suddenly mine more gold--and thus suddenly *do* mine more gold) or sudden deflation (if your ability to produce other goods grows faster than your ability to produce new gold). You also get the insanely-complex interactions between gold as a labor-produced commodity and gold as a currency: gold has a non-trivial cost of production and is itself desirable, so it's a product, and thus a gold system is barter. (I do not like terms like "intrinsic value" because the "intrinsic value" of gold will become 0 if you find a way to magic gold out of thin air, and so we are really only discussing the labor cost of production of gold as a product.)
You can't magically find new bitcoins; you might find a new way to compute a bitcoin hash chain in polynomial time. For the most part, bitcoins don't act like a commodity in the physical sense; they do act as a commodity in the economic or financial sense. The simplification is that bitcoin is a limited-supply currency with a deflationary behavior--both more and less correct than calling it a commodity.
I didn't think deep enough into it on the first pass. Oh well.
(Yes, I did just argue that gold coin, gold-standard currency, and bitcoins are the same thing, and all bad for the same reasons.)
We're not all going to the same place, so mass transit becomes incredibly inefficient. People like to think only of fuel and operating cost, but not of time and the aggregate cost across population. Imagine if all commutes were 90 minutes longer each way.
Statistics show fewer incidents per traffic volume in roundabouts than at intersections. There are fewer than 40% as many contention points in a roundabout than at a traffic signal.
Wandering into a locale with roundabouts everywhere *will* put you under a *lot* of stress if you're not used to driving them, though. The mental approach to a roundabout is completely different. While the brain can easily handle the task, it handles it about as well the first time as it does driving in general: it strains, and then becomes confused. Once you've learned how to react to roundabouts, they're a lot easier to navigate than a city riddled with stop signs and traffic signals.
We're actually looking to replace one traffic light intersection with a large roundabout here. I expect a *lot* of complaints and unsteady drivers trying to navigate it for the first 4-6 months, and then an enormous reduction in the number of collisions. This intersection experiences frequent traffic incidents, and even ambulance route around it; fire engines crash straight through because a fire engine colliding with a passenger vehicle is just a passenger vehicle turning into confetti. (Really. In a few towns in CT, if you park in front of a fire hydrant, the fire engine will simply crash full-speed into your car to move it out of the way.)
Political influence. They want their name on it.
Government built and builds roads, rails, and space ships because those infrastructure projects were *enormous* and not very profitable. When no private corporation can front the money *but* the return is huge, you ask if the Government can do it in reasonable taxation. The answer is usually "not yet," until it becomes financially feasible. Such infrastructures transition to the private sector if they don't require a centralized, non-direct-revenue administration (e.g. power lines, railways, and Internet carry a product to customers; roads cannot be monetized in that way).
When new steelmaking techniques reduced the labor required in making steel, rail became feasible. Prior to certain blast processes, laying steel for rail would have required us to stop making all the products human society was consuming and devote years to making and transporting steel; maintaining the rails *and* the engines would have been more than the entire body of human labor time available. At a point, it became technically possible... if we all stopped eating. Then, we found lower-labor methods which made it technically possible, but all the wages involved made such transit so expensive that it would be cheaper to just drag them by cart--or at least to ship them along the coast (which is what we'd done all that time).
Eventually, we developed steelmaking processes which used little enough human labor to lay down rail and operate freight transit at acceptable prices, but not within the financial capability of the private sector; and then the Puddling process (and, later the Hot Blast furnace) and iron rolling processes finally lead to commercial creation of localized rail. We didn't get a trans-continental railway until the Government stepped in with tax money in the form of bonds issued at 6% interest. The project was simply too big for any given single corporation or trans-continental partnership to fund and expect a profit, so the Government built it with taxes (taxes it collected at 6% interest, since those bonds were gaining that right up until they were paid back).
Today, we have several companies actively developing self-driving cars. They all have access to GPS map data and, besides, any widely-distributed navigational data sells for cheap enough: an enterprising corporation could gather any such information cheaply enough, then resell it at a mark-up to any car company wanting full trans-continental or global annotated GPS map data. Sensors and central collectors in cars to gather and organize the data in real-time after deployment would be, essentially, free, as they'd use the same hardware and data networks *required* to operate the cars in the first place.
The government obviously doesn't need to issue bonds and fund this shit via tax money. The Democrats want their name on the face of progress, economic freedom, new technology, and environmental friendliness (autonomous cars are more efficient drivers than lead-foot teenagers, and will probably end up being mostly electric or hybrids); the Republicans want their name on the face of firm economic support, the growth of new businesses, and the freedom to pursue the American Dream by creating a new product and establishing an empire. Either party has good reason to throw tax money at the problem: PR.
There never was any reason to think Bitcoin would be better than the existing financial system. People believed in bitcoin for ideological reasons rather than pragmatic ones for the most part.
A lot of people were on this limited currency thing, where it becomes hard to make new bitcoins. That creates a deflationary currency, which is a good way to make sure debt becomes worth more over time, rather than less. In other words: mild inflation good; hyperinflation bad; deflation potentially worse, but hyperinflation is pretty fucking bad. It's kind of like saying you're small, but a Pygmy Mouse Lemur is smaller: that's not helping your case.
Bitcoin was never a financial system anyway; it was a money-making pyramid scheme.