Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
msellout
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
32 ms
·
241.
▲
by
msellout
11y ago
If 'trapped in financial instruments' includes standard investments, like deposited in a bank, loaned to a business, or buying equity in a business, then your money is not trapped at all but spreading through the economy in multip
242.
▲
by
msellout
11y ago
Stadiums and kind of businesses that appear nearby take up valuable space that could be used for higher-value economic activities. There was a rush of US mayors sponsoring stadiums in the 1990s. Almost all of those turned out to be net nega
243.
▲
by
msellout
11y ago
On a global scale? I thought all those Hans Rosling "Mind the Gap" videos were telling us the world is increasingly equal.
244.
▲
by
msellout
11y ago
Convention. FYI for non-Pythonistas: the prevailing convention is lowercase_with_underscores for everything, except CamelCase for class names. Of course, the core types are mostly lowercase. I think it's somewhat like being a famous sc
245.
▲
by
msellout
11y ago
Yes, the lambda was equivalent. However, the second example offers that it might be better to be more verbose and avoid the use of anonymous functions. In general, I prefer to define a block and name parameters. As the slideshow illustrated
246.
▲
by
msellout
11y ago
Correct, though I find that a language is more about the community and culture than the syntax itself.
247.
▲
by
msellout
11y ago
Correspondingly, some objects don't need names at all. sorted(pairs, key=lambda p: p[1]) Though in that case you might prefer using the operator module. sorted(pairs, key=operator.itemgetter(1)) The aspect of Ruby tha
248.
▲
by
msellout
11y ago
If they cannot borrow Euros, they will have two options: default on promises to Greeks (pensions, etc.) or leave the Eurozone and return to their own currency which they can "borrow" from hypothetical wealthier future taxpayers.
249.
▲
by
msellout
11y ago
Although we can theoretically handle up to 2^64 bytes of RAM (16 exabytes), the practical limit is much lower. I think someone on Wikipedia said it's somewhere around 8TB, but I imagine the performance of random access into 8TB RAM is
250.
▲
by
msellout
11y ago
Yes, that's an unlikely scenario. As you say, they can mine bitcoin.
251.
▲
by
msellout
11y ago
I certainly agree there's a rich-get-richer phenomena, but I'm not sure there's evidence that mid-size VCs are disappearing. The generic phenomena is "preferential attachment", which creates a power-law distribution
252.
▲
by
msellout
11y ago
Doh. I should have said, "Use Jython."
253.
▲
by
msellout
11y ago
It's not technically a cast to a list. It's using the list constructor with the iterator as an argument.
254.
▲
by
msellout
11y ago
My professor once said, "If you want to make money as a consultant, invent an index." You'll make the low scores panic and pay for advice. You'll make the high scores preen and pay for your lectures.
255.
▲
by
msellout
11y ago
No, I wasn't thinking of free markets. I was thinking of the bundled concepts of money, markets, capitalism, and the whole mess of technologies we use to allocate resources among the population. It's more interesting to think abou
256.
▲
by
msellout
11y ago
Thinking of it another way: money is a technology for distributing wealth. Without money we would need to rely on some centralized authority, a chief or pharoah, who would directly or through a bureaucracy distribute wealth. Money allows so
257.
▲
by
msellout
11y ago
You can ignore the type hints syntax. I plan to. Python 2: map(func, iterable) # returns a list Python 3: map(func, iterable) # returns an iterator list(map(func, iterable)) # returns a list
258.
▲
by
msellout
11y ago
There are many ways to get around the GIL. Use PyPy. Use processes. And are you really doing that much computation? Threads on standard CPython are perfectly fine for I/O.
259.
▲
by
msellout
11y ago
First you need to define what you mean by "complex" as that could refer to many things. A useful definition for the term will be focused on informing a decision. Why did you ask the question? Further, definitions that rely on one
260.
▲
by
msellout
11y ago
Which answers in particular? I thought most of the answers were quite prescient of features that would be introduced to the standard library some years later.
261.
▲
by
msellout
11y ago
I believe it was, despite not calling the name out explicitly. Are you suggesting that simulated annealing does not belong in the categories of "repetition strategy," "alteration strategy", nor "ensemble strategy&qu
262.
▲
by
msellout
11y ago
As with many projects, there are people who use the code in production. What do you mean by "production ready"?
263.
▲
by
msellout
11y ago
One strategy is to consider establishing a success metric as the first step to every project. If you can't design a measurement for success/failure, maybe you shouldn't be doing that project.
264.
▲
by
msellout
12y ago
First, find out from the investor how much of the money he can take as salary. That can make a big difference. Second, realize there's no global "blacklist", just people, friends, and long memories. It can be a small world so
265.
▲
by
msellout
12y ago
Actually, with deep learning the fact of success is leading the theory. We don't know why deep learning works as much as we know it does work. Edit: in certain domains such as images and speech
266.
▲
by
msellout
12y ago
"Best of all possible worlds."
267.
▲
by
msellout
12y ago
Why not? It's effort to cancel them and has equivalent results.
268.
▲
by
msellout
12y ago
This fails when two networks have the same name and different characteristics, and you want connections to be automated. It's frustrating when I accidentally connect to a poor network just at the edge of range.
269.
▲
by
msellout
12y ago
You're using a different definition than the one I learned in school. The purpose of diversification, as opposed to simply picking the best asset, is to decrease variance. Given a choice between two assets, or portfolios, one first max
270.
▲
by
msellout
12y ago
Randomness is an easy way to ensure you're not doing something systematically wrong.
More ›