1,299 karma · joined April 11, 2010
As in, Oracle installing adware with their Windows installer is currently damaging to Java as a brand, and making Oracle Java be less of a "sane default" than Sun Java was.
As for institutional investment, this would create incentives for the founders (or other equity holders) to exaggerate the valuation of the company. On a different note, this kind of vesting would mean more equity for people who worked for the company while it was growing slowly (and who took on more risks - that's why you give them equity and not hard cash) before seeing exponential growth, and no equity growth for people who joined after the company had reached a stable state.
The next point is, what's whith people who want to disinvest from their position? This is both the case for investors after a company has reached a certain size (e.g. Zappos before they were acquired by Amazon), but it will also be the case when people who get partly paid in equity want out of their position.
Vesting cliffs unilaterally benefit the current equity holders, because it creates a skewed incentive structure where you either should leave at the start or stick it out until the vesting cliff. (And, that's what it was with Microsoft in the 90s/2000s: many people stayed on exactly for the time it took for their equity to vest).
For that, you'd need some "unbundling" strategy where not just the Star Alliance (or whatever the companies are called) can sell seats, but another entity can buy seat contingents and sell them at similar conditions. Oh well, phone service, search, and flights are all different, with different rules.
Of course, this system will still be constructed according to the biases and fears of real humans. Most people wouldn't call it an "AI government" because there's already an elected government calling the shots ... to a certain extent.
over(max, scan(0, max(+, 0))
slightly more readable, even if that still gives me a higher-order-headache.
Because, y'know, it's confusing to have symbol salad with a weird mixture of infix and postfix operators, even more so if you have type raising (or the moral equivalent thereof) thrown in for good measure.
(From what I understood, Rust gives you similar control over memory allocation as [a sane subset of] C++, so intuitively it should come out relatively close)
Which is about as jaw-dropping as what people spend on WoW and other pay-to-play games to get to the end game.
Hangeul has very simple rules, and is generally simple, but if you want Hangul Jamo (i.e., archaic letters that don't compose neatly into "standardized" syllables), you'll be in a similar place as if you want to produce mediaeval ligatures or thorn letters. Þis is hardly ever neſceſʒary in modern text.
Ideology starts when you automatically view an instance of a decision on a particular topic as a person's disposition towards that topic, and automatically view a person's disposition towards one topic as a reason to put a label on them and ascribe to them dispositions towards totally unrelated topics.
A non-ideological stance can consist in noting that, while many occasions have shown a co-occurrence of sexism-related idiocy and trolling, they are not the same thing (see Lennart Poettering being harassed for non-sexist reasons), and hence we have multiple angles of attack: combating the trolls through effective lawful action (which is a bit tedious, but still what OP thinks should be done when the guy in person has a well-known track record of harassing people in ways that frequently go beyond what's funny or harmless), combating people's willingness to act on behalf of trolls (e.g. reddit and 4chan's ban on personal-army type posts), as well as acting on the polarizing issues that make people more prone to get drawn into these astroturfing-initiated conflicts (which is a hard thing to do).
Do we have a moral common ground to do any of this? Absolutely. Even if we suspend contentious arguments and ideological blame-shifting.
competition introduces a superlinear factor in the time/pay equation
superlinear factors on average hurt women more than men since the latter are more likely to be on the upper end of workload (assuming a linear correlation of time/pay, you'd already get from a 77% "wage gap" to a pay gap of around 92%, which would tighten further or disappear if we account for "winner-takes-it-all" distribution of pay)
I find it relatively easy to believe that today's working environment leads to an incentive structure similar to all-pay auctions, where the employer reaps huge benefits due to everyone's fiction of attainable winning, and which hurt a majority of people, but women more visibly than men.
I response to your implied question why this is perceived as much more interesting than an older submission of a talk by Anita Sarkeesian (which I didn't watch, I'm just extrapolating here), it's because it separates out the ideology part from the description of the trolling phenomenon. You don't need to drink SeriousPony's koolaid to hear her argument about trolling in general and this troll in particular.
You can't expose trolls by preaching to the choir, you have to talk everybody's language to subvert trolls' efforts to create a polarized discussion.
Note how OP's post was met with weird complaints that she "thought prostitution and being a victim of domestic violence were somehow 'shameful.'" - this is again trying to use ideology to polarize a discussion that should be troll-targeted and not ideology-targeted.
Note: I'm not claiming that we shouldn't try to discuss ideologies, just that it's something that trolls will try to exploit. And that the difficulty of moderating such discussions is why Hackernews doesn't even try to be welcoming to discussions that are or look ideological. Let's put it this way: the moderators here have a good track record for creating a pleasant, constructive discussion around technical stuff (including Windows and Linux and upstart and systemd and a host of contentious issues) and around business/startup stuff. As we find out how to have a constructive, non-polarizing, non-troll-prone discussion about other issues (global economy, misogyny/misandry, whatever), these things will be received more positively on Hackernews. (Or maybe: as much as they are somewhat related to YCombinator's business with startups).
For some people, BSD init works just fine and they're happy with it and they abhor the additional complexity. Without any doubt, Systemd also has its flaws - it's not compatible with *BSDs, or generally non-Linux systems, it's a departure from the "everything is accessible using a simple text editor" principle that has brought Linux where we're now, and there's a non-zero switching cost for each and every package.
Many people know polipo-audio (later renamed to pulseaudio) and the crashing propensity of its earlier versions, and who are kind of intimidated by the whole DBus/ConsoleKit tangle that introduces many moving components that are hard to debug when they fail. A system that is "perfect in theory" but crashes often is not that great.
So people are kind of apprehensive when the same guy who brought them crashing sound demons a couple years ago now comes over happily with a solution to replace the most central component for their system. Thus far, I've been pleasantly surprised by Systemd silently doing its thing and working as advertised. And I wouldn't swap it out for upstart if someone gave me the choice.
The morale? Yes, people grow up, complex systems become more manageable with time when people write debugging aids for them. (Including things such as VirtualBox which make debugging central system components much easier - imagine living in a world where your systemd crashes and bugs simply get closed with a WORKSFORME tag and ignored).
"communities are full of bullies who won't hesitate to call you or your mother names" - I think this is best put as "doing controversial things will attract trolls and haters".
"you must equally be a bully enough" - no, you just need to tune out the haters and continue to work towards your goal.
Lennart and the others underwent a ginormous effort to grow systemd from a borderline dangerous "I'm smarter than everyone else" toy project into something that's actually useful.
They did that despite the fact that the problem in question had been solved a couple times over (upstart was the last in a series of "improved" versions of SysVInit), and despite the fact that systemd adoption incurs a large switching cost of re-writing init scripts for all packages.
The art of leading a successful open source project is partly in tuning in to useful feedback and tuning out the haters who have nothing to contribute. Yes, the haters function based on ways that are similar to bullying. But while you can keep them off the mailing list, you can't ban them everywhere.
The more visible a community or an undertaking, the more visible it is for haters and bullies, and the more effort is lost on policing them or tuning them out. There's no way around it, other than educating the general population or staying a small, very technical in-group.
This carries the danger of people shortening it into "I have the loudest voice" in their mind, which begs the question how other communities unite the role of "last instance for contentious questions" (aka benevolent dictator) with the ideal of being nice to everyone.
(i) you want to tag other texts, and
(ii) a license for the Penn Treebank (i.e. the standard training set for English newspaper text) sets you back by about $3k,
plus another couple thousand if you want a commercial license for the Stanford tools (although GPLv2 means you can use it in a SaaS without getting one)
(i) people in NLP actually bothering to tell others why their approach is interesting
(ii) other people being interested in the same / a similar kind of thing [avoiding the discipline-of-one problem that niche AI applications would have] and
(iii) NLP having a reasonably developed "canon" about what counts as must-cite papers. This canon is heavily biased towards US work, and towards people who write decent explanations of what they do, but at least it makes sure that people know about the big problems and failed (or not-quite-failed-as-badly-as-the-others) solutions.
What you see in other conferences is that the "Best paper" awards get to (i) more theoretical papers which still have issues to solve before people can use the approach (nothing wrong with those!), in (ii) subfields that are currently "hot". Whereas the most-cited papers are (i) more obviously about things that a dedicated person could apply in practice, and (ii) in a subfield that is obscure at the time but will become more popular in the following years.
For both country governments and the governments of corporations, having some amount of both democratic control (through elections, or through laws that say what corporations can and cannot do - we don't want corporations to poison our kids any more than governments) as well as some economic responsibility (which is difficult - both for CEOs and for government officials it is frighteningly easy to disguise bad performance as "properties of the global economy").
taxing you to death for subpar services? Hello, Comcast and AT&T!
brainwashing your kids? Disney or Activision would never do that!
violent action? Corporations only do that in places with weak governments, which is a good thing. Never mind that the US health care system means that firing people also means taking their health care away.
can put you into deadly situations? Well, corporations would never do that. No one died in Dow Chemical's (now Monsanto) huge chemical spill in Bhopal. Also, both corporations and governments get rather bad PR when they kill people.
You whip out your strong-authentication-of-the-day device, do the magic bellydance (i.e., enter some numbers into that device, or scan something on the screen) and whoop, you just transfered a sum of money. To someone in a different European (SEPA) country. Without paying anything (except if you live in Italy and have a greedy bank). With the money coming in within two business days at their side.
And of course you have an option to perform monthly payments that way. Some landlords put it in the rent contract that the tenant has to use this method.
France was one of the last country to abandon checks. Why? Because everything is more efficient when it's electronic, and there is less forgery. (You can still send a wire transform via mail with a forged signature, but with these, you can usually go to the recipient and give them a hard time, which is harder with a check someone cashed out two days ago).
The successful answers to these questions have always been those where people choosing B actually get something (be it a warm fuzzy feeling of supporting good movies, a nice box, or just the convenience of an interface that allows you to simply find and watch a movie).
If you want to rely purely on IP protection for that movie, good luck - people can watch other people's movies, the next YouTuber's shoestring budget entertainment, or a football game? Why wouldn't they read a book, or spend an evening knitting nice red socks.
There is no inherent law of nature that says that we need more film producers or film directors, or that they should make more money than the singer-songwriter in the bar down the street, or that guy who makes his living composing music for films and adverts once in a while. Or more than the guy who spent 5-10 years of his life trying to sell pet food over the internet and noticing that the business model is not competitive.
It's the investors betting on (a) people finding your film more attractive than knitting red socks, and (b) you being able to monetize that fact properly.
"Monetization" can come in all sorts of shapes: selling individual copies of the movie, selling licenses to distribution companies (including Netflix), putting adverts on the movie, or streaming.
Given that we have Google Play and Netflix, you should think that there is some value in allowing an app store model of people buying any film they like. That this doesn't happen is partly due to the rightsholders of these films (usually larger companies) not seeing the point of that when people could also watch new movies (= more revenue for them).
You could also make an argument about culture. It would open your eyes to the many ways that the entertainment industry - or the majority of it in terms of capitalization and revenue - is not concerned about culture at all, and to the many ways that what we have now - in terms of grassroots/pastime content creation, publically funded culture (e.g. orchestras and theaters in many places), and large projects (such as films or video games) - of creating, revising and sustaining cultural artifacts.
Will it have legal consequences? Most likely no. Will your coworkers or employees stab you with a rusty fork for getting their (most likely) favorite search engine to block them? Absolutely.
It is not blue-sky research, though: it's pretty clear what the value proposition is, who the (potential) customers are, etc.; IBM just put on the publicity upfront with their "winning Jeopardy!" grand challenge to put themselves on the map in a space that is currently dominated by small boutique companies and a very limited numbers of bigger players such as Thomson-Reuters and (partly) Nuance.
Which means that IBM will have an easier time selling to existing enterprise customers who want to get in on this "big data" thing using text analytics (and, frankly, it's also good advertising for their existing data analytics business).
Seeing the paper at http://www.jucs.org/jucs_10_7/total_functional_programming/j... it's more related to proofs, and indeed Idris seems somewhat similar to Coq (a "programming" language for writing proofs). Or maybe I'm just totally lost.
So, incf, which is CL's +=, could be defined as
(defmacro (incf lexpr val) `(setf ,lexpr (+ ,lexpr val))
which would evaluate "lexpr" both as a place to put the new value in and as an expression to get the old value out.
If you wanted to do that in terms of purely-functional monad application, it gets a bit hairier... Basically, you've got things like "+=" which are mutators (i.e. functions Int->Int). And you've got places, which either (i) take an X mutator and turn it into a mutator for Y, or (ii) take a Y and return an X (accessor).
In your example,
list_of_objects[index].attr = 'abc'
would mean that you take a "constant value" mutator that takes some string and returns "abc" and turn it first into a mutator that changes the .attr value of something ... to something that changes the .attr value of index [index] of the list_of_objects member of the current data.
Why would that kind of functional style be useful? Basically, because you could write something like a Python program that still allows for search/nondeterminism/backtracking etc.
Some of the weirdness I see in Mutagen seems to come from a partial (or just differing) understanding of how Python and Haskell, respectively, do things. Shoot me a PM if you want to discuss more.
Honestly, I've always missed defsetf or a similar construct in languages without macros. Python's __setitem__ and __setattr__ are already nicer than C++'s pretense of using references, but setf gives you nice improtvement.
For non-lispers:
In common lisp, if you assign (setf) an expression, the compiler looks for the corresponding setf-method.
As an illustration, consider
ball.roundness() = true;
where the outermost expression is a .roundness() call, which you could transform into
ball.set_roundness(true)
without much confusion.
Note that this is conceptually more complex if you want to do this in a purely-functional language.
- inverse causation - common cause - random correlations
The first two should be relatively stable and reproducible, and we could then proceed to find out about causation with an intervention study (e.g. if "good education" directly causes "good job prospects", will it help if we give people good education that wouldn't normally get one? Or does it improve people's educational achievement when we give them better access to jobs?)
The third isn't reproducible, but should normally be relatively rare. Why we're seeing more and more non-reproducible results is usually
- People fishing around in data sets for correlations, or tweaking experiments until they find a correlation
Because of this, we end up with a great deal of correlations that have nothing to do with causation.
For a related but different perspective, see this article ("Language is never, ever random", Adam Kilgariff) http://www.kilgarriff.co.uk/Publications/2005-K-lineer.pdf
underage = banned, also on reddit