This is good observation. There is correlation between knowing esoteric languages and being exceptional (See Paul Graham's Python Paradox) as a programmer, but knowing Haskell or Lisp or OCaml doesn't make one an exceptional programmer any more than learning a bit of Latin makes you well-rounded and cultured. Knowing esoteric languages is a byproduct of being driven by the internal drive to learn and not caring about what other people use. Unfortunately, just like in hipster fashion, unpopular things that initially attracted secure people who didn't care about what others think, later end up attracting insecure status-seeking people who self-consciously embrace.

Exacerbating this is the notion that some super programmers are orders of magnitude more productive than average programmers. While there's some truth to this, the notion has made legions of programmers intensely insecure about where they belong. And insecure programmers tend to look for ways to cheat for status, approval and security, not engage in the kind of natural exploration and learning that makes the great programmers great. To become great, it takes a lot more than taking a week to learn a new programming language.

I think there's more than some truth to this. I think its just true.

Edit: Not trying to make any deep point or anything. Just sayin'.

And I'd be your competitor that wouldn't waste time on pointless compiler pedantry.

I would be doing this in an area where writing super safe/performent C code actually makes sense. Probably in triple AAA game development.

It wouldn't matter if you were my competitor. Stability is really important when it comes to games no matter what and competition is a bit subjective because people prefer different games.

You can still do fairly fast iterations in game development even with strict guidelines.

If you don't want to waste time on "pointless compiler pedantry" then don't use C...duh
It's precisely the areas where you do need C where this pedantry is very much worth the time

There are many, many, many stories where companies spend a lot of unneeded time/money fixing stuff later when the writing was on the -Wall
Also warnings are sometimes escalated to errors on future versions of compilers.

I mean this is almost close to saying "I'll be the competitor that doesn't waste time writing all those tests" for dynamic languages.

"performent?"

And I love how you're saying this when you couldn't even correctly read the compiler output pasted in thread.

CTR-F sng_jason

Simplest Monad overview that sort of works for me:
- lower and higher datatype, allow functions that are defined for the lower datatype to be called on the higher datatype

Also researching this a bit finally makes me want to refactor the old "pass the world around" Prologcode I'm using which has been bothering me from the getgo. Seems like a variation of the State Monad in Prolog should be doable with DCGs. At least a fun little subproject for when I get back to work

I didn't even read dave's compiler output before posting. I only develop with strict error/warning reporting and I've only been taught that in order to use malloc you have to reference the stdlib.

I quickly glanced at your code snippet and noticed it wasn't there. Don't be mad because you're an outlier when it comes to compiler settings.

I spent the morning writing out the rant. 12 pages. Do you think anyone will bother to read it?

I feel better now.

Depends on the topic

ok wtf this is just dongs

Exercise: Make the text in the spans true

ez http://livecoding.io/4507702

hrm. I already got it, but in a slightly different way, and am now trying to figure out how that third line works. Does "son father" mean it can be referenced as either son or father? I was thinking it was an "and".

it's a hierarchy tree, so if you write ".son .father" any node of class father that has some ancestor of type class son will be matched

EDIT: in the html, when you write "son father" that means that that specific node matches both classes, that is what may be confusing you

I should probably not do it since I don't have loads of time but whatever. Ordered "The Elements of Computing Systems" and will start with the Intro/Chapter 1 this weekend.

For those who don't know what it is:
You use a harware simulation suit to build a computer from a single logical gate (NAND). You work your way up from low level stuff untill you can write code for your architecture. The corresponding course is called "NANDtoTetris" because what you'll do in the end is write a working Tetris clone for you self built (virtual) computer.

I did some hacking around building a bootloader and hello world OS (x86 yikes) a bit back but I pretty much didn't know what I was doing. I had some relevant classes during university but can't really remember all that much. I feel like I'm lacking a broader understanding and thus I think it's a neat idea.

If I actually finish it there will probably be all sorts of interesting follow up ideas (I see there's a VM...what if I'd want to replace that with the Erlang VM or a Warren machine). Might also be interesting to build a homebrew OS for the RaspberryPi with the understanding gained from this and some blood and swat etc. pp

If anyone is interested in a study group etc. it's always more fun to do it in a group (starting date can be pushed back). From browsing the book you don't really need any pre-knowledge (basic coding in any language maybe). I think it's actually designed to be used by non CS folks as well.

Corresponding website (with PDFs etc):
http://www.nand2tetris.org/course.php

I loved my electronics class in college. Building a little mini-computer from simple logical gates was really cool.

The one exercise I remember is we had to figure out how to light up a 5x7 led panel where you could only turn on or off entire rows or columns (the intersection of an "on" row/column was off). So you could make something that's symmetrical like an X. But a G was a challenge. I had an epiphany in History class that you had to create the letter out of as few states as you could - then use the logical gates combined with an phase oscillator to cycle through them faster than the eye could see.

I ended up making a smiley face that used two phase oscillators - one to make the faces and one to slowly cycle through happy, neutral, frowning. I was very proud of that.

Electronics and Fortran were the only two classes I really enjoyed in college (BS in Physics). So it only makes sense I wound up doing computer programming. Even if I didn't start until age 29.

Thats the book this guy used to build a 16 bit computer within minecraft,

http://www.youtube.com/watch?v=LGkkyKZVzug

lol

That's ****ing awesome.

can someone explain it a bit? what are the red lines? what is the fire? how does it correspond to hardware, etc?

I'm no expert but I believe you can consider the lines of fire as voltage. Black means off. Red means on. Where he sets the input and views the output are 16 bit binary numbers (black = 0, red=1).

http://www.minecraftwiki.net/wiki/Redstone_Circuits

This is the only reason I ever considered playing minecraft, lol. Still haven't though.

Wow. That looks pretty cool. What programming language are they using?

Unfortunately, I would not be able to dedicate to this, but I have it bookmarked.

I dunno yet but I think they use HDL and a selfbrewed scripting language for the hardware definition and then their own Java like language (Jack) which they build up to. I think in the end it's Java.
I think the ASM is also their own.

From what I read it should be fairly easy to adapt to any language though (decent idea for a followup project)