mazelife

I'd somehow never heard of this guy. The piece of his they link to doesn't contain any microtonality (at least not the parts I listened to) and sounds _very_ much like mid-period Scriabin. I was a little disappointed at first, but then dug a bit more and found a couple of the microtonal piano pieces:

https://www.youtube.com/watch?v=WCvSkmIMRMY and https://www.youtube.com/watch?v=VOat_xsGUOw and https://www.youtube.com/watch?v=3EGriPbFmKE

These were more interesting, and had some good moments. You can kind of see what he's getting at with the color wheels (particularly with #5 from his Twenty-four Preludes in Quarter-tones - https://www.youtube.com/watch?v=WCvSkmIMRMY). But ultimately I'm not sure how illuminating those wheels are from an analytical perspective.

Also, the article's contention that "standard pianos cannot play this kind of music," is incorrect. There are lots of microtonal works for piano, it's just necessary to properly tune one to perform them.

On a somewhat tangential note, there seems to be a lot of synesthesia (specifically Chromesthesia) among 20th century composers. This guy, and Scriabin, and Olivier Messiaen, Amy Beach, Leonard Bernstein, György Ligeti, Michael Torke. (And going further back, Liszt and Sibelius). I don't know what the incidence is in the general population, but I wonder what the connection is, if there is any connection.

Here are some links that might interest you:

http://yaledailynews.com/blog/2015/11/02/sae-denies-charges-...

http://features.yaledailynews.com/blog/2015/04/16/harassment...

http://yaledailynews.com/blog/2011/05/17/miller-announces-dk...

http://qz.com/546403/yale-student-protests/

It's a shame no recordings of this instrument are known to exist [1]. Given the mechanism was quite similar, it seems likely it would have sounded like a Hammond Organ.

[1] http://www.historyofrecording.com/Telharmonium.html

Yes, but there are an infinite number of ways it could possibly not work. How long will they go on discovering them? Edison could say something like that because he actually was able to successfully build and market a lightbulb.

Good piece in the Baltimore Sun from a few months ago about what a boondoggle this project has turned out to be: http://www.baltimoresun.com/business/federal-workplace/bs-md...

"But after 17 years of research and $2.7 billion spent by the Pentagon, the system known as JLENS doesn't work as envisioned. The 240-foot-long, milk-white blimps, visible for miles around, have been hobbled by defective software, vulnerability to bad weather and poor reliability."

I dunno, taking aim at jargon seems to be central to the article's purpose based on my reading of it:

"The idea that writing should be clear, concise, and low-jargon isn’t a new one—and it isn’t limited to government agencies, of course. The problem of needlessly complex writing—sometimes referred to as an 'opaque writing style'—has been explored in fields ranging from law to science. Yet in academia, unwieldy writing has become something of a protected tradition."

That's pretty much the thesis of the article. Anyway, I've certainly encountered academic writing with "tortuous syntax and sentences mostly devoid of meaning," although it's a small minority of writing in my experience. This article seems to suggests it's a conspiracy to keep the public out while signaling some kind of in-group membership. But I've always let Occam's razor guide me: a much simpler explanation is that many people (and academics are by no means free from this) are not good writers.

I'm not going to say that there's isn't plenty of obscurantism, needless jargon, and just plain bad writing in academia. But, this article paints a pretty simplistic and one-sided picture. Let me take a quote that's emblematic of the problem from one of the professors (!?) in the article: “Academics, in general, don’t think about the public; they don't think about the average person, and they don't even think about their students when they write,”[Deborah S. Bosley] says. “Their intended audience is always their peers. That’s who they have to impress to get tenure.”

Specifically this notion that "[an academic writer's] intended audience is always their peers," is treated as though it's somehow incriminating. But it's not, it's what most academic writing is for! Many academics do write for the public—and that's great, and more should do so—but they don't publish that in journals or with an academic press and when they do write for the public, their writing is usually pitched at a more accessible level. But most of the writing academics do is aimed at their peers because that is how ideas are transmitted among specialists and how a body of knowledge is gradually expanded. The notion that that they only do this just to "impress" or get tenure seems to be a really bad-faith argument.

Look, any highly technical endeavor is going to have highly technical jargon that’s pretty impenetrable from the outside. There's nothing wrong with that: it allows specialists to communicate about complex topics efficiently. Whether the field is medicine, philosophy, or computer science is immaterial. Frankly, if you want to see a place where obscurantism, needless jargon, and just plain bad writing are used as tools to disguise the fundamental banality of the ideas being presented, look no further than a lot of business writing. I'd take academic writing over management-speak any day.

Or EXPLAIN (ANALYZE, BUFFERS), which is very useful when you suspect disk IO may be a choke point.

Since when does supporting the righting of an injustice require you to also support rioting? I'm sorry, that just doesn't logically follow. You're confusing the tactics of a tiny minority with the substance or a huge movement.

So when a small group of individuals riots that somehow validates, for you, the injustices experienced by the larger majority? That suggests a pretty shallow commitment in the first place.

Look, I'm a Baltimore resident and I hated to see the rioting that occurred here back in the summer. I understand why it happened and what motivated it, but it didn't really move the ball forward for anyone that cared about fighting for justice. However, in the days and weeks leading up to the riots, there were large scale, peaceful protests in the city involving many, many more people then rioted. (A fact the national media paid virtually no attention to.) Most people out there don't think activism = violence or necessarily leads to violence. It's quite possible to make a distinction between rioting (and I think most of our Baltimore rioters were just using the situation as a pretext to rob and loot anyway) and activism.

There was a pretty good long-form piece about Yahoo News in The Baffler a while back [1]. It was by Chris Lehmann, a former YN editor. Although one expects that there's going to be a certain amount of one-sidedness in an account by an employee that was let go, it makes a pretty compelling case that one of Yahoo's flagship properties was disastrously mismanaged. He's highly critical of Mayer: "When editorial knowledge is deployed to elevate advertising, it stops being either editorial, or knowledge; it’s hackery, sold out to the highest, or just the bare middling bidder. [...] Under Mayer’s leadership, the company has brought this ugly art form to sophisticated new heights — and in the process blown through all the basic rudiments of journalistic ethics." There's definitely some corroboration for that view point, BTW. [2]

And it wasn't just the way the firewall between editorial and advertising was dismantled by Mayer. He documents the atrocious working conditions, the extent to which the executives Mayer hired to manage YN had no idea what they were doing, YN's reliance on gimmicks to try to juice page views (which didn't even work), and the way the site's editorial mission seemed to shift from week to week. Lehmann tells a particularly hilarious story about a Yahoo exec trying to pitch a Crossfire-type blog where a rightwing blowhard and a leftwing blowhard would argue with each other. The exec really wants to hire Robert Novak, not realizing the guy had died two years ago.

Anyway, after reading this and then reading the OP...not surprising.

[1] http://thebaffler.com/salvos/purple-reign [2] http://deadspin.com/dan-gilbert-didnt-like-a-yahoo-blog-post...

I have very mixed feelings about this. On the one hand, it's great to see a younger and/or less-traditional audience attending symphony concerts. But I wonder what percentage of these first-timers would ever come back for a non-videogame music performance? Because—the title of this article notwithstanding—if these concerts don't do anything to cultivate an audience who actually engages with classical music (or "art music" or "western concert music", call it what you like) then videogames are not going to save the symphony. Not even close.

People in the comments rightly jumped all over the guy in the article who said these concerts were “akin to Mouton Rothschild using its wine to make and sell sangria.” But there is a point buried under all the snobbishness: video game music (and film music) are not supposed to be concert music. In fact a good film composer would probably tell you that if their score for a film is so compelling and interesting that it can be extracted out and plunked down in the concert hall as a fully-fledged work, they're doing it wrong. Film/videogame scores are (by definition) supposed to be subservient to the whole of which they are a part. Thus it's the narrative of the game or movie that drives the formal aspects of the music. This isn't to say that there can't be greatness and even genius in scores, but that it's not being used in service of producing a stand-alone concert work. Which is why even very good scores usually have long passages of music that successfully create an atmosphere for a particular scene, but taken out of that context would sounds rather boring and aimless in a concert hall. Similarly the kind of drama evoked in a concert piece where you have musical themes that are presented and then developed and transformed at length just doesn't work for film scores, where you are working in much shorter time increments (like the length of a scene). This is why even the greatest film scores (Maurice Jarre's score for Lawrence of Arabia and John Williams' score for Star Wars, for example) are known for having a few really memorable, resonant melodies and/or for the way in which they contribute to a memorable scene.

I've gone to film music concerts and enjoyed them, but I'm under no illusion that it's necessarily the most challenging or profound stuff. I'm also aware that it's partly the memories and associations I have with the movies themselves that I am enjoying when I listen to the music. In any case, it's not at all the same experience as a classical music concert. I hope that some of the people who have dipped their toes in the water listening to an orchestra perform "The Legend of Zelda" might be curious enough to go back and hear what other, far more interesting kinds of things an orchestra can do. Because if not, these concerts will be a nice windfall that can help pad an orchestra's bottom line (and god knows they could use the money). But they won't do what orchestras need most, which is to develop a real, life-long appreciation for classical music in younger audiences. Unfortunately this article presents zero statistics one way or the other on that point.

You're point about "Irish songs seem to be simple enough for the NN to understand it back to front," is a good one. There are a number of aspects of the compositional parameters here that I think make it much easier to generate this kind of music.

Firstly, it's modal harmony, not diatonic. In this case, dorian mode in A. If you've every improvised in say, a whole tone scale, you notice that you can play almost anything and it sounds good. Modal music works in a similar way. A lot of the things you find in diatonic harmony: the tension between tonic and dominant, chord progressions in general, key changes, chromatic inflections, etc....are all absent in this kind of music. Which isn't to say anything bad about modal music, just that it's much simpler. Because of the added complexity of diatonic harmony, there are many many more ways the music could "go wrong" so to speak. Most people are so grounded in diatonic harmony that they would easily perceive even small mistakes (or statistically speaking, deviations from the norm) without necessarily being able to explain what rule, exactly, is being broken.

It's also monophonic music; there's just the one voice and no accompaniment (other than a completely static rhythmic accompaniment that was added to the performance).

Finally, even within this much simpler framework, I'd argue this tune gets it wrong in a big way: it doesn't know how to come to an end. It just sort of stops, in medias res. In a lot of folk music, you'll find that the way a tune is ended still tends to hearken back to diatonic harmony: some sort of motion from dominant (e), maybe even with a raised 7th scale degree to tonic (a) that's outlined by the melody. That doesn't happen here, which is why the tune sounds like it just got cut off.

I find these NN experiments in music generation quite interesting conceptually, but so far the results--as music--have been pretty disappointing. I suspect that you could actually build a model that would allow for algorithmic generation of folk tunes that would produce music that would probably be more satisfying. The number of rules that govern a lot of kinds of folk music are small enough that you could encode many or at least most of them in your model. [1] However, at the end of the day you'd still just have a model that would only generate a fairly limited spectrum of folk music--say Irish gigues, reels, hornpipes, etc--whereas the dream with NNs, markov models and other statistical methods is that you could plug in any corpus of songs without understanding a thing about their harmony, form, structure, melodic patterns, etc. and get back music that sounds the same.

[1] and this is really massively simplifying on my part w/r/t the varied amount of folk music out there, some of it quite complex

I'm not sure if you read incorrectly, or it the article was updated after your comment, but I don't see that quote anywhere. What I do see is:

"Just as nuclear scientists developed norms of ethics and best practices that have so far helped ensure that no bombs have been used in attacks for 70 years, AI researchers, he urged, should embrace a similar ethic, and not just make cool things for the sake of making cool things."

That said, attributing nuclear disarmament solely to "norms and ethics" of scientists is pretty reductive on the part of the author. In fact there were large nuclear disarmament movements in Japan, UK, and the US after WWII. It was as much public pressure as anything that lead to the Partial Test Ban treaty in '63. Similar movements in the 80's led to further reduction in arms. One million protestors marched in Central Park in '82, protesting the nuclear arms race. That's a staggering number and indicative of how widespread and popular the nuclear arms reduction movement was at the time. Comparing that to the "Anti-AI" movement—which just about everyone would agree is pretty niche—is problematic at best.

No doubt there is interesting data to be mined here, but I wouldn't rely on it solely. The vast majority of patients are not in a good position to evaluate the quality of care they received. The kind of things they typically judge by are often fairly superficial attributes:

* did I have a long wait time?

* was the office staff polite and helpful?

* was my physician/NP/PA polite and helpful?

* were there billing problems after the appointment?

These things do matter, but there are more important factors in play that patients are not in a position to judge:

* were standards of care followed by the provider?

* were unnecessary tests ordered or unnecessary drugs proscribed?

* were necessary tests skipped or necessary drugs not proscribed?

* if they had access to my medical history, did the provider bother to review it before seeing me

* etc, etc.

Quality of care can absolutely be evaluated, and that's happening more and more, it's just that patient satisfaction is only one small part of the whole picture.

In addition, the realities of managed care and reimbursement rates mean that--for many types of practices--doctors have to see a pretty large volume of patients every day to be able to stay afloat. No doctor I've talked to likes this state of affairs, but it's a reality. This leads to dissatisfaction because patients feel that their doctor is rushing through the appointment without giving them adequate time. It doesn't necessarily mean they're getting substandard care, but it's also reasonable to feel angry or annoyed by that kind of interaction.

Conversely, many practices will do everything they can to maximize physician utilization, including scheduling lots of back-to-back appointments. Inevitably, if a doctor is actually doing a good job, some visits will uncover more complex health issues than planned and require more than the 15 or 30 minute allotted to address. This means that appointments start to back up, people start having long wait times to see their doctor, and patient dissatisfaction ensues. Again, this isn't a problem with individual doctors or practices-- although it may seem that way to patients--it's systemic.

- Violent offenders represent 53% of the population of state prisons and 7.9% in federal prisons [1]. So what's everybody else doing there? Also, it's necessary to present evidence that our unheard of incarceration rates are what actually casued violent crime to drop to unheard-of lows. Some people have proposed, for example, that the most compelling reason for the drop in violent crime in the US was attributable to lower levels of lead exposure in children of the 70s and 80s, due to the elimination of lead from gasoline and paint.

- [edit: I misread what you were saying here, the point about the death penalty is a non-sequitur on my part] Your second point isn't one I would disagree with but, but I don't think your example is a particularly good one: there is ample evidence indicating that the death penalty does not serve as a particularly strong deterrent. For example, murder rates in the US are actually lower in death penalty states, and have been since the 90s.

- "That means changes need to occur with the electorate, not elite legal minds." The author never asserts the legal profession is the only thing that can reform our broken criminal justice system. However he points out that the legal services we have now are vastly inadequate to the job of protecting the rights of most Americans caught up in the criminal justice system, and that part of the reason so many people are incarcerated is that they recieve little-to-no represenatation. Furthermore: "even apart from the millions of pending criminal cases for which people are not being provided a well-resourced and zealous attorney, every one of the thousands of unlawful stops, searches, home raids, beatings, taserings, shootings, and arrests that take place every day forms the basis for a freestanding constitutional civil rights suit. A quiet tragedy of the legal system is that these rampant daily violations are almost never litigated."

The broader point he is making is that the criminal justice system is only able to grind through tens of thousands of lives each day becuase those involved in it--and that includes lawyers and judges--are commited to keeping that system well-oiled and functioning. "Imagine a world in which lawyers stood ready, en masse, to use their skills and training and intellects to vindicate these constitutional rights every day. Such a social movement of lawyers would dramatically alter the nature of the legal system and our society. The system of modern policing, which depends on callous indifference to vindicating basic rights, would crumble at our simple willingness to hold it to its own formal rules. We can do it, but only through massive collective action to act on our professional and moral values."

[1] http://en.wikipedia.org/wiki/Incarceration_in_the_United_Sta... [2] http://www.nber.org/papers/w13097.pdf

That's a good point. A sum type/disjoint union (by my understanding) means the type is in the union of subsets and that the subsets are pairwise disjoint, but it doesn't say that the subsets have to be finite. Since this post is really aimed at Haskell beginners, I was trying to avoid going down the rabbit hole as regards type system or set theory and I might have oversimplified in the process.

Thanks for pointing that out; I think the syntax highlighter may have clobbered some code. It's now fixed.

Author here. Thanks for pointing me to pylogsparser, I'll definitely take a look at that. Your point is well taken: without building a parallel implementation in Python, we don't have a way of knowing for sure if Haskell is faster. The only data points I have are that we've built a couple of logparsers for custom formats in Python before this, and the number of lines parsed/second was far smaller than the attoparsec-based parser.[1] It's not apples-to-apples, since the formats differ a bit, but I don't think that it has no predictive value. So in the second part of this post, which I'm working on now, I'm hoping to be able to provide a fully-functional NCSA combined log format parser in Haskell alongside the blog post. I think that would be fairly easy to benchmark since it's a common-enough log format.

[1] that's just measuring the time to parse log files into some sort of structured data, not necessarily to do anything with it

This seems to be similar to Euterpea, a domain-specific language embedded in Haskell for computer music composition: http://haskell.cs.yale.edu/euterpea/.

I'm curious what the differences are, either in capabilities or in overall goals, if anyone has looked at both.

There's a secondary source of government funding beyond earmarked grants to universities: the tuition assistance it offers to students, which then goes straight into the pocket of the school. Consider pell grant spending, which has grown about 8x since it was established in 1975: https://www.americanprogress.org/issues/labor/news/2011/07/2...

Even if we factor in the number of students receiving the loans (which has grown), the average amount awarded to students roughly doubled since 1994.

I'm not getting mixed up about the direction the money flows; see my quote down-thread: "endowments come from donors by definition"

When asking why is tuition growing, I was saying exactly what you state: that for universities that are focusing on increasing the size of their endowments, something has to cover the costs that would formerly have been funded by endowments, that "something" being tuition.

No, endowments come from donors by definition. The question is—for schools that are sitting on large endowments—why is tuition so expensive? And why are there so many cutbacks to things like number of full-time faculty when the school is sitting on a 1.5 billion dollar endowment?

Mostly anecdotal, which is why I said "suspects". I'd love to see a comprehensive comparison of university endowment growth over the last few decades. I suspect the data's all there, but nobody's pulled it together. I've come across a number of articles discussing large increases in endowment size and a number of universities of late, seems to be a common theme as more and more people are starting to ask why college has become so expensive. For example there was recently a story in the Washington Post about Washington and Lee University which covered how it had gone in the 90s from a school with a relatively paltry endowment to land in the top 30 overall.

These two sources are interesting:

http://www.jbhe.com/features/57_pellgrants.html

http://en.wikipedia.org/wiki/List_of_colleges_and_universiti...

The first source leaves little question that it's happening at top tier universities, but I'm interested in the trend and second- and third- tier schools.

edit: As another commenter points out, most schools aren't sitting on large endowments. There probably isn't a single unified cause to explain rising costs. Endowments are probably part of the story for higher-profile schools, but not so much for others.

This reads like a press release written by York University administrators, or pretty much college administrators everywhere. Some facts that offer a useful corrective:

1. Rising college costs have massively outpaced inflation in the last 2 decades [0]

2. Salaries for tenured professors have not seen anything even close to that kind of growth in the same time period [1]

3. Most institutions actually employ fewer full-time faculty than they did 20 or 30 years ago, shifting more and more to part-timers and adjuncts who get paid a pittance [2]

So...college is massively more expensive than it's ever been, but full-time professor salaries have not risen significantly and there are actually fewer of them overall. So where has the money gone?

In the US, I don't know that there is a complete answer on that, but there are definitely some suspects:

* offsets to deal with slashed state funding (although federal funding has risen significantly in the same time period)

* university endowments

* university administrators [3]

* building sprees and tons of shiny new facilities

[0] http://www.bloomberg.com/news/articles/2014-11-13/college-tu...

[1] http://nces.ed.gov/programs/digest/d13/tables/dt13_316.10.as...

[2] http://nces.ed.gov/programs/digest/d13/tables/dt13_315.10.as...

[3] http://www.bloomberg.com/news/articles/2012-11-14/bureaucrat...

The problem is covered in Chapter 5. There's isn't really a silver bullet, depends a bit on your use case. He outlines a couple of approaches:

* Try Again Later: queue failed parts of the transaction for retry, moving towards eventual consistency

* Abort the Entire Operation: On failure issue a compensating transaction to roll back the previous successes

* Distributed Transaction: have some overall governing transaction manager process that orchestrates consistency (e.g. duplicate the transaction process of your monlithic data store up in the application layer). You could do a two-phase commit where each participant in the distributed transaction tells the transaction manager whether it thinks its local transaction can go ahead. "A single no vote is enough for the transaction manager to send out a rollback to all parties."

For each of these approaches there are issues, obviously, and to his credit he's pretty even-handed about them. He also says:

"If you do encounter state that really, really wants to be kept consistent, do everything you can to avoid splitting it up in the first place. Try really hard. If you really need to go ahead with the split, think about moving from a purely technical view of the process (e.g., a database transaction) and actually create a concrete concept to represent the transaction itself."

He gives an example of a domain model for a retail company of an order-in-process: instead of attempting to capture a user's order as a single database transaction, model the concept of an order flow of through fulfillment, shipping, marketing, accounting, etc.

Does it? I don't see where he's making the claim that the reason you should listen to him is because he is a better designer than the folks at Medium. In fact, he says "[m]edium’s ho­mo­ge­neous de­sign works and reads well." The claim he's actually making seems to be about issues of openness, ownership, and authorial compensation, specifically that users are trading those things away for the fit and polish of medium's platform.

Unrelated, maybe it's a platform-specific problem? I found his site to be quite readable. Things like line-height and line-length appear to have been chosen with care, and he takes great pains to get all the typographical details right. The use of a "◇" to indicate links is...eccentric, but whatever.

...which is why I usually turn to this: https://pypi.python.org/pypi/unicodecsv/0.9.4

It's a "drop-in replacement for Python 2's csv module which supports unicode strings without a hassle."

The cut-and-paste aspect of Messiaen's music is definitely one of it's many unique features. That said, it's always been my belief that Messiaen got the idea from Stravinsky, specifically the Rite of Spring, which takes the same approach.