Here's a really cool video that describes in lay terms how Chopin "noodled" one of his compositions:

https://www.youtube.com/watch?v=r9LCwI5iErE

So, sorry; I find that line so offensive.

Hey Dave,

I think this is a case of misunderstanding, see below...

But unless you are using lightpad or something similar, then you do, by necessity. Keep in mind he's not claiming that you are building something in your head over the course of an hour, then coding it up. The bar he's measuring against is instant feedback. He discusses this a lot in the inventing on principle talk.

This is as it should be. He's saying that your tools for haphazardly building and tweaking suck. Your tools don't give you instant feedback, and they don't allow you to see (physically see) the structure of your algorithms.

He would agree completely. He want you to experiment and fail. He just wants the experimentation to not be 50% experimentation and 50% fighting with your tools.

I think those are fine, and are probably an improvement over the for loop. But that's not the point. 1. Those examples don't allow you to "see" the whole process at once the way his visualization do. 2. What if your task is to teach, specifically, the "for" loop?

2 days!! This is exactly the point of the article. 2 days is insanely long. The "concept" here is something people already understand. We should be looking at a timeline of minutes. You could explain what the loop is doing in English to anyone, right? So why this massive divide between natural language and computer language? Why do you accept it so readily, and defend it? That's what he's questioning. You really need to step waaaay back from your own experience with computers and all the stuff you've learned and are proud of knowing to appreciate the point he's making... we are all the inmates running the asylum.

You're mixing up two different things: the achievements of mankind and the limits of what one person can hold in his brain at once. He's talking about the latter thing. And even if you're an incredible genius, that limit gets reached pretty fast. That doesn't mean you can't achieve amazing things. We do it by working around that limitation: mathematicians create a symbolic language (they only need pen and paper), so do musicians, engineers, etc.

He actually talks about this specifically. Yes, there are some people with the ability to hold lots more in the heads than the average person.

But you seem to be saying this is the only way to create great art. This is absolutely untrue. If you look at most musicians today, they don't work in their head. They play around, they find things, they build them up. There are many ways to skin a cat. The noodling around way is the way most people create, and he wants to empower it.

The advances in technology have allowed many people to express musical creativity who couldn't have previously. He wants to do the same for computers.

Also: How do you know that the great classical composer couldn't have composed more works, or better works, if they had better tools themselves?

I guess I wasn't clear enough on my examples. This is probably more self-explanatory:

"Hello, Student: This is how you can rotate a triangle:"

"And you see we rotated this triangle by 10 degrees.

But we can approach this by another perspective. Suppose you wanted to rotate the triangle 90 degrees, but yuck... you don't want to compute 90/5, do you?

Let's take a different perspective:"

My issue is based around the idea that he is showing *way* too much to a new person. There is nothing particularly wrong with taking things step-by-step. I argue that it is far more important to teach and empower the student with the ability to look at things from multiple angles. Conquering this little baby-step stuff isn't anything to be ashamed of as long as you can show someone how to hit the same problem from multiple perspectives. Yes, his may be a slower start, but the payoff in the long-run will be miles above any immediate gains.

Take a super-simplistic even/odd example and exploring different perspectives of this problem:

v1
v2
v3
I'm not sure if the author is a musician or creates music, but I can say with full certainty that learning to even get sounds out of an instrument takes more than a few days of noodling. The ability to bend your mind and fingers to behave is a very long process and a process that really never ends. You can't be Segovia overnight. Mastery, and even half-ass proficiency, comes with the ability to look at the same thing in multiple perspectives. A guitarist could always play at the top of the neck, but one day, he learns that he can play that same scale from the 5th fret, then the OH **** moment comes when he learns he can play that scale from from 1 down to 12 and play it forward and backwards. It's this love of exploring and finding those gems that keeps him going.

The "grammar" of music is extremely difficult to learn, much less master, especially to that level explained in that video I linked to. Take the idea of a Blues master. He spends years and years honing his craft to a decent level, and the Blues is, for the most part, 5 notes, and you may be surprised to know that the Blues seldom deviates from those 5 basic notes. To get to the level where you put out something decent and tolerable for masses takes many many years of composition and thought. That is why the debut album from some artist is quite often so good, but then the output afterwards is trash.

Music theory is the "grammar" of music, and it is highly unintuitive, and for this reason, it's a terrible argument on his part to make. However, learning theory helps you become a better musician and learning that stuff helps you create music much faster. For all the trash-talk about long-haired Rock pot head not knowing any music theory or grammar, 90% of that talk is pure myth. Most of them can play scales upside and backwards and have the Circle of Fifths virtually tattooed on their ass. There are a few exceptions to the rule, but they are far and few between. The real masters learn and study to use this "grammar" to consistently create new music that isn't utter trash. The theory, first and foremost, guides them, then they create from that, not the other way around.

So, to answer your question: I honestly don't think it would matter either way if they had our technology or not. Remember that the Beethovens et.al. created 300+ pieces, and many of those pieces were created when he was basically deaf. It was that constant practice that made them masters, not the instruments. I think the proof is in the results. We've been messing around with piano machines for 50+ years and no one has created anything worth a damn with them. I have my own thoughts on why this is, but it basically breaks down the impure sound of these machines.

There is something to be said for a profession that you can actually become employable at after a scant year or two of learning. What is the big rush and why do people feel that we have to be overnight Dijkstras with everything? It's not the way the world works in writing, math, music, or any other discipline where you have to break away from your world view and bend your mind against its will to achieve success. The initial mind-bending in programming is very shallow compared to music, painting, and concise writing. The latter three takes many many years of practice, alone time, and deep thought to get to the level that someone is willing to pay you money. Programming? It doesn't seem to follow the so-called 10,000 hours rule that these other disciplines demand.

The elephant for me is this other idea that many of the things he suggests has already been tried in some other iteration. According to one person who is an expert in Flash, "it is basically a movie." Isn't plug&play a'la WordPress or --install essentially the same concept in a nutshell?

The final thing I need touch on is the idea of historical perspective. There is no single person that invented the Next Big Thing. All of it was a result of people before us and some lucky bugger getting the final eureka moment. We are very lucky to be able to witness the speed of progress we see today. I believe that we'll all get there, and I totally respect this guy for putting his thoughts out there, as I can see how some of his ideas could be used and I've seen how his ideas have been used.

Regardless, he offers tons of food for thought and it isn't drivel. I think that his concept of some innate creativity borne from noodling and spreading paint around haphazardly is misguided. It appears haphazard and dreamy, but the reality is that it took many many years of study and trial-and-error to learn how to make those strokes and motifs. He let illusion grab a hold of him, but it is really no different than his zoning in programming. Perhaps its an easy target because the arts is the realm of some fantastical untouchable place where "natural" ability reigns. Those artists gained their "natural" abilities after practicing about 10 years. That "noodling" around is pure illusion. Think of it as "focused screwing up."

Sort of reminds me of an answer I heard to someone sampling a drum hit. The sampler believes that there should be no payment because it is only one strike on the skins. The drummer responds: "It took me 34 years to learn how to hit a drum like that and you want it for free?"

You were perfectly clear, I think we are talking past each at this point. I really feel like you are not getting a lot of what he's saying, rather than disagreeing with it. But who knows. In any case I'll try once more and then I think we'll have to hang this up -- at least you're getting food for thought, as you said, even if you're not convinced by all the arguments.



Again, your examples are clear and good relative to a lot of beginner tutorials out there. But you are still stuck in the same fundamental model that they are (it is after all the model that both you and I learned in), and you are missing the visualizations that he has, specifically visualizations which abstract over time and algorithm itself.

I just flat disagree that it's too much. His visualizations are intuitively and immediately obvious, despite containing all the info -- that is the hallmark of a great visualization, that is its power. Even with my years of programming experience, those visualization make the concepts clearer to me, clearer than any code ever could, and I have internalized that code.


i believe he's an amateur musician, yes, but i'm not sure. your second sentence is just a flat misinterpretation. he's not saying you can learn to play an instrument in a few days, of course not. he's saying that even professional musicians compose incrementally, getting feedback and reacting to it. That may be messing around a guitar and a tape recorder, or a piano and pen and paper, or whatever. What they don't do (at least not many of them) is sit in a silent room with no instrument and hear the music in their head and write down this abstract music in their head. He's not claiming this is impossible, but he's claiming this is very difficult for most people and not a natural way to create. The equivalent of messing around with that guitar for the programmer would be having tools that let him immediately see the program he's creating -- these tools mostly don't exist except in his demos and some projects like lightpad. An in any case the vast majority of programmers do not work with such tools, even though such tools would vastly improve their lives (see the excitement around lightpad and it's kickstarter success, eg). Everyone I've talked to who sees Bret Victor's prototypes thinks it would be awesome to have them.


again, see above. ofc he's not saying that becoming an expert at anything is easy. but that is no reason to just accept a sub-optimal learning process. Basically you are arguing against a straw man here.


Beethoven is just a bad example. It would be like me bringing up the accomplishments of Terrence Tao and saying "See. Look at what he can do. He learned math in our current school system. There's no reason to change anything"

The last 50 years.... you should be looking at new musical forms like rock and rap or whatever, and successful musicians who could not have even been called musicians in the classical era. Comparing the output of classical music would be like saying how come no painters today are making things like Rembrandt did. But anyway, this is a silly argument, I think you're taking this much too literally. Really his only point is that most working musicians compose with constant feedback.

again, this is all straw man stuff. i highly doubt he'd disagree with any of it. and none of it is an excuse to never criticize or try to improve current teaching methodology or tools. that's all he wants to do.

i won't comment on the rest because i'll just be repeating the above paragraph.

ok, well in any case this was fun. i'm glad it's at least provoking to you.

Massive hijack incoming, but the answer to the second question is that people aren't trying to.

Well yeah, that was my point....

I enjoyed that. The thing about how true competence comes with losing the impulses (the emphasis) on every X note really struck a chord with me. It seems like there is a deeper lesson there.

I tend to think of programming as solution oriented learning. And just like with learning there are different types. Some like to draw stuff on paper and "think on paper" some juggle stuff in their head while taking a walk and some just hack away in an editor and adapt as they go along.

I really don't think there's one way to do it. It's knowledge work a bazillion of approaches can work. I'd guess the "do stuff in their head" thing dates back to the ancient days of shared processing time when you kind of had one shot at getting it right before waiting for days to get another one.

What divide? Us Prolog folks just describe what the solution should look like and the computer finds it

[kidding aside at least according to one book I read this is the raison d'être for Prolog. Get rid of the divide between problem solving and encoding]

Should be free if I can recreate it (should be free to use for sampling if I bought the sound once regardless, lol IP)

anyone here have experience with both compass and intuit.css? i am a bit confused if they are solving the same problem or different problems. that is, do you use one or the other (as with bootstrap vs foundation) or would you potentially use both? so would your CSS stack look like

(sass or less) + (compass or intuit) + (bootstrap or foundation)

i know there are other optionss for all categories, but i just want to make sure i understand the full ecosystem, and it's only the compass/intuit part that confuses me

The point I was trying to make wasn't about the quality of the output or other intangibles, but how mastering the fundamentals allows one to think about complexity and create complexity. The so-called natural talent is really nothing more than mindless mastery of basics for years on end. An artist draws circles and squares before drawing a human on top of it. A musician plays scales before creating a Jazz piece, etc...

If your point is that we could be teaching better, then yes, I wholly agree with that assessment. I for one, do not do well with rote learning and data regurgitation, but others excel at it. It doesn't mean that rote learning is bad simply because I disagree with it and I think it is LCD teaching.

I simply don't know, but I feel like introducing complexity too early is more harmful. It's a rough question and I'm certainly glad people are thinking about it, but you see where the issue is. Ask five people the best way to learn and you get five different answers. Ultimately, at least in programming and the arts, you have to be driven to teach yourself and find reputable guides that work best for you, but guides that are challenging as well. The main reason we don't have more artists in our world is that it is very hard and the basics are mind-numbing to learn, much less master. With that said, I sort of recall myself two years ago thinking that the only point of programming is to create images and move them about, and I ultimately think that was a destructive road to go down. when we focus too much on the fun, we are doing ourselves a massive disservice.

I've been using foundation with a simple over ride file. There's so much in foundation that the over ride file is only about 10 lines, which only changes the font colors and stuff like that.

I don't have the ability to think in such complicated layouts that I felt I needed to use Sass on top of Foundation. For certain, the layouts I can create now are more complex than what I can create with plain CSS. Mind that I'm not creating awesome stuff or using mixins right now.

Foundation is seriously huge and I'd suggest just starting with the basic before piling on other stuff. You really don't need Compass, or at least, I haven't needed it and I don't think foundation says you need it anymore. Sass seems pretty mandatory if you want to become the e-Da Vinci though.

How close to Boston??

you all write lots of words

pm'ed

I think I'd like my next job to be as some sort of javascript developer. I have way more fun working with stuff like backbone and angular than I ever have with rails.

Javascript devs still in super high demand?

I started working through that Definitive Guide of JavaScript book and came into this thread to say JS is the most bizarre language I ever seen in my life. Someone posted here a while back about what is so great about "JavaScript: The Good Parts": the author still spends 1/3 of the book discussing the bad parts in the world's slimmest programming book in existence.

@Nchabazam:
Why would you think that there are no js jobs anymore? Why would any sane company hire an angular / jquery / dojo / backbone / node / clojurescript / coffescript / [flavo-flav in the house] dev that doesn't know js?

What is set to replace js on the web?

I've been contacted and asked if I know js and I am only a marketer.

Tons of work out there for good JS developers, I need to hire one asap in fact, and they aren't exactly growing on trees.

I do hope and pray something like asm.js gets traction though, would love to have non-javascript options.

Some thoughts on that Definitive Guide on JavaScript. I'm about 300 pages in right now.

I've read reviews like "very good" to "meh." The title most certainly is not joking when it says "Definitive," but I feel like this book is a little much yet incomplete. The main failure is that it is over-complete on rum-dum syntax mixed with eye-glazing bloviation on every little topic.

It's lacking two critical aspects. The first is the lack of talk about what not to use. I guess I have enough experience now to look at something and have alarm bells go off in my head, but I'd be pretty concerned if someone picked up this book to early in their learning curve. I could be jumping to judgment here as these topics may be touched on later.

The second is the utter lack of usable, working code. I guess this isn't meant to be a tutorial or cook-book, but I would have thought a book of this size would be loaded with awesome code.

The final thing is the organization. While slow-sloping up with topics is fine for a shorter book, this book will introduce strings on page 45 and then say "we'll discuss this in more detail in chapter 17."

I imagine there are superior resources out there and with that thought, I wouldn't suggest this book.

JavaScript itself... I have a weird relationship with this language. I first attempted to learn how to program in js and fell out in spectacular flames. I've always wanted to learn the language because, well, I've always been interested in creating cool stuff you can see. Look at any JS demo site and you know what I'm talking about.

I find it interesting that Brandon Eich read SICP and fell so in love with the book and concepts that he wanted to originally create a Scheme. Knowing that this book was such a strong influence on JS makes the language so frustrating to learn about at this point. There are truly brilliant things going on and then all of a sudden I am hit with such a heavy what-the-**** that my heart wants to break, as this betrays it's rushed birth.

JavaScript is a brilliant train wreck: I want to exam more parts but I know I'm going to find a severed hand along the way. It was rushed off at the 11th hour and someone forgot to tie the chains before it left the station.

Dave, why not read js the good parts? Or have you already?

Better writing == better programming, didn't you hear? All the cool kids say so.

daveT: How old are you again? Iirc oldish that got into programming late, right? I kind of feel like you're some sort of bizzaro-me. I come to post that V 0.1 of my poker code is done...book you already created the thread. I come in here wanting to ask about JS-DG because I'm considering rewriting some/all of the Java part of the framework in JS eventually boom you post about the book (I'll stay away for now).

If you post about Elixir next week or something I'll check my security settings :P

I'm sure nothing is replacing JS, I'm just getting bored writing straight backend rails with very little on the front end. I know there are tons of people who want sweet rich 1 page apps.

That's why I have side projects!

Also, the site I'm working on would really be a good candidate for a mostly 1 page experience, but since I'm basically the only dev it doesn't make sense now. Down the road I might try and spearhead that initiative.

EMCAScript v6 is next for the web with regards to Javascript.
http://en.wikipedia.org/wiki/ECMAScript

v6 has support for classes which will be great. Browsers should start adopting it later on this year iirc

hey guys, wondering if anyone can help me with this code. The line 92 doesn't work (as well as similar lines later in the function, I am getting this error:

What I think I'm doing in line 92 is the following: ptrFlights is an array of pointers to Flight objects. So, *(ptrFlights[i]) will dereference the pointer that is in the i element of the array, and *(ptrFlights[i]).getDeparture() will call the getDeparture function of the Flight object. What is my mistake here?

Oh, and is this the right place/thread for these kind of questions?

just a guess, as my C++ is very rusty, but I think "." has higher precedence than "*" dereference operator, so maybe try:

if ( (*prtFlights[i]).getDestination==myDesination )

Two other things:
- are you sure getDestination is not a method, ie, did you forget parens? getDestination()
- this code needs to be cleaned up badly