Yet it is certainly false that if you are able to get CS basics right, you will be any good at deep engineering.
354 karma · joined July 17, 2007
(https://play.google.com/store/apps/details?id=keepass2androi...)
The thing that I found incomprehensible, and somewhat tangential to the grandparent, is how ASUS manages to screw up distribution so badly. They release slightly different models at slightly different times in different geographic regions, and it is usually impossible to buy, say, the European or Asian versions, in the US. A year ago I was quite literally ready to throw money at them in exchange for a UX303LN to replace my aging U36SG, but my dream model was only available in Europe. The only two options I could get in US both had something wrong with them. Repeat: they actually manufactured a laptop to the exact specifications that I wanted, but didn't give me a way to buy it.
Considering they have global distribution channels, it don't imagine it actually being hard to make all models globally available, so the only other explanation I have is forced market segmentation: the two models available in US were pretty much the lowest and the highest end, with nothing in the middle.
I had been waiting to see if UX303UB will make it to US, or if it will suffer the same fate. Instead, I just pre-ordered a Surface Book.
There is another potential effect: as part of the labour certification process, the company is required to post up the applications of all foreigners it is hiring in a "prominent location" for around two weeks, to make sure all other employees are aware of it. It can be rather awkward to post up an LCA with a number on it that is significantly higher than what others in the company are getting.
My experience has been that many people are paid significantly more than the prevailing wage on the application. As mentioned, this is also just the base rate, without bonuses or equity.
I would say this analysis is useful for drawing comparisons between companies and observing trends, but less useful for the absolute numbers.
[edit: for a searchable interface to this data, as well as complete approved applications, check out http://dolstats.com/]
I disagree when it comes to reading books, though. My Kindle Paperwhite, and even the Kindle2 before that, are superior reading devices. The Fire is a decent video/games device, but the only reason I use it in preference to the Nexus7 is to get access to Amazon Prime streaming.
To expand on your points, based on my own experience:
* Rather than using the centroid to represent the cluster, pick a representative peak from the hue histogram. This tends to make things less muddy.
* Not only do I throw away extreme S/L values, I aggressively weigh everything by S * (0.5 - |0.5 - L|), so bright, saturated colours dominate. Black and white are usually very thin slivers, and I almost never have grey.
* As a last step, I convert to Lab space and merge any colours that are perceptually similar (dE < 5.0). This means that large gradients (like an unevenly lit surface) only occupy one sliver of the pie. Unfortunately, in my current implementation this conflicts with my first point above, so the colours I end up with aren't always in the original image -- something I need to fix.
Here's my result for one of your images: https://dl.dropbox.com/s/azl2bag84riugg8/imghist.png, notice how the orange has a disproportionally large weight because it is so saturated.
Wait, I should use more bandwidth than I need, and impose a client-side performance hit? What is wrong with a little CSS to serve the right assets to 1x, 1.5x and 2x pixel ratio devices?
Perhaps of some use to those annoying coupon code sites that go to great lengths to prevent copying of the coupon code in order to funnel users through their affiliate link. [Edit: To be an inconvenience to scrapers, you'd want to generate a random font per session.]
10: PRINT "hello"
GOTO 10
label: PRINT "hello"
GOTO label> type safety is loosened and a few keywords are borrowed from JavaScript (var)
C# is strictly and statically typed. The var keyword is only similar to Javascript's in using the same three letters. What the author calls loose typing is derived by the compiler, at compile time, in a completely static fashion. However, you get the power of being able to type variables even when you cannot name the type (for example, anonymous types).
> with LINQ, they decide to borrow the single most verbose language on the planet (SQL), and add that too
LINQ is about SQL about as much as Javascript is about Java.
edit: Oh yeah, and Java never really added generics. It's just some syntax sugar which in no way actually closes a hole in the type system.
Is it not even conceivable to the author that members of hard-core programming forums advocate C# because it is actually a decent language backed by a well-implemented runtime?
As for the environmental impact and total carbon footprint, that of wind farms is not small either. Vast areas of otherwise-usable land concreted, carbon expended on building, maintaining and dismantling wind mills, backup power required to buffer against weather conditions, and the like. I haven't seen any comparison of various energy forms with all the costs included, but would certainly like to take a look at one.
[edit:
A quick search led me to this: http://www.nrdc.org/nuclear/pnucpwr.asp
It seems that while existing nuclear power plants (constructed over 3 decades ago) provide marginally cheaper electricity than fossil fuel based powerplants, new nuclear powerplants are indeed not viable economically, at least as compared to fossil fuels. The report suggests that including a carbon emission tax of $100/tC will bridge that gap.]
[another edit:
A paper (http://gabe.web.psi.ch/pdfs/Annex_IV_Dones_et_al_2003.pdf) claiming that total carbon footprint of nuclear power, from mining to waste disposal, is in the range of 6-12g CO2-equivalent per kWh, mostly attributable to mining, compared to 500g/kWh for natural gas and >1000g/kWh for coal.
Hydro weighs in at anywhere from 3 to 3000g CO2-equiv per kWh depending on region (alpine hydro stations are clean, flooding tropical forests is not) with weighted average somewhere in the 300g range, wind is at 14-21g/kWh depending on weather conditions, and finally solar at 79g/kWh with potential to fall to 30-40g with future technologies.
The concluding graph shows only [European] hydro beating nuclear in terms of total carbon footprint.
Lots of other interesting reports on that site (http://gabe.web.psi.ch/research/lca/), most offering similar results. One 2007 paper also considers land use, ecotoxicity, and radiation emissions, allowing certain types of wind generators to approach nuclear in terms of total environmental impact. Solar is still way off due to high levels of inorganics and toxic non-greenhouse air pollutants. ]
True. Also consider the relative efficiency of the generation methods. Supplying the world's demand for energy with nuclear power would probably require far fewer powerplants than doing so by other means.
> I guess if reactors were located on land that was already unusable (eg. desert, esp Antarctica)
Except for the melting of the Antarctic ice and all that, in case of an accident :) Most technology seems to involve a trade-off of evils.
Neither. More like making generalisations dangerous (Although I am sure you will be able to pick apart my post to find plenty of generalisations, too).
> The average American diet is, to use your inflammatory language, "outright suicidal for an average middle-American". It's high fat, high sugar, low fiber, and low in nutrients (unless "fortified" with artificial concentrations of vitamins and minerals), and it is nothing like previous American generations diet.
I did not at any point advocate that Americans eat the average American diet. Nor did I mean any offense to America, merely to the diet. I don't eat such a diet, nor do I eat the average Russian diet since leaving that region. The point I am making is that accepting any extreme view on dietary requirements is a strange thing to do considering how little we understand about the workings of our metabolism, regardless of how many cultures have enjoyed such a diet for countless millennia. [edit: I realise you did not advocate such a position, either. But plenty of people misinterpret statements like the one I quoted to do just that.]
I am sure tofu is great, and has lots of science behind it. So does consuming red meat. Or milk. They also have lots of science pointing to the contrary. But most of those scientific arguments apply to extreme cases. In the right environment, eating an occasional steak will not give you a heart attack or colon cancer, just as eating a few soy beans will not make you impotent. But excluding foods on the basis of current scientific understanding strikes me as more damaging, if it turns out that those foods carried nutrients we previously did not understand. Of course, dietary requirements based on religious or moral principles are something else entirely.
Seems like we agree on the general principles, the devil is in the implementation details, and one's risk profile.
Several of those disasters, of scales arguably larger than Chernobyl, were caused by attempts to generate power. And although I do not have any evidence to back this, I suspect the total number of casualties resulting from burning fossil fuels over the past century far exceeds that from nuclear power related accidents.
The one aspect of nuclear accidents I'd be willing to concede is the long-term contamination of the area, but while rendering chunks of land unusable, it is not a direct threat to life.
And let's not forget that a large fraction of casualties after the Chernobyl accident is attributable to the fact that the general public was not informed for an entire day after the accident, in fact not until a powerplant in Sweden reported heightened radioactivity levels. This, with Kiev, Ukraine's multi-million capital, only 100km from Chernobyl.
That's a dangerous statement. Your lifestyle is different to the lifestyle of the Chinese over the past few millennia.
People in northern Europe eat very large quantities of cholesterol-rich foods: salted/cured pig fat, and a thick meat-based jell-o made by boiling pig legs for hours, are two popular dishes in Russia. This diet is not only tolerable, but is recommended during the colder months, as (1) fat is the best energy store we have, and (2) cholesterol reduces the freezing point of lipid cell membranes, helping maintain their fluidity. On the other hand, such a diet would be outright suicidal for an average middle-American.
The same cultures also eat large quantities of heavily-salted preserved vegetables during the winter; the salt is not particularly good for you, but the negatives are outweighed by the benefits of vitamins in the vegetables, which would otherwise be unavailable in the winter months. Again, not recommended for the average American who has access to fresh fruit and vegetables all year.
In Australia, people are hospitalised every year from over-eating after a rich Christmas turkey meal in 40C heat. The tradition evolved in the English winter, where energy demands are higher and digestive systems operate at a greater efficiency.
I do agree with you on the fads. Human metabolism is still sufficiently misunderstood that any claim that advocates particular extreme dietary behaviour will ultimately be proven wrong. The sequence of events usually goes something like this: x provides essential nutrients -> x is added to every product on the market -> scientists claim x causes cancer! -> scientists discover a way to emulate x with y, a healthy alternative! -> y is added to every product on the market -> scientists claim y causes cancer! -> ...
Those claim are generally true, technically. Everything you eat kills you; the trick is balance and moderation.
Great! So now, all you have to do is give some sort of a comparison that lets people visualise exactly how big an area 50,000 times the size of U.S. is. May I suggest a metric based on whales?