|
Yeah it's a totally dumb assertion, if people dont get it first time round either they have a lovely teacher, bad source material or both. I did not really "get" C until my third attempt at learning it. The first two times I was using some books that turn out to be embarrassingly bad in hindsight, but as a neophyte I had no way to tell. It went a lot better once I had a copy of K&R and APUE.
|
#
?
Sep 12, 2013 03:20
|
|
|
|
| # ? May 31, 2024 18:29 |
|
|
Hard NOP Life posted:People just spout poo poo that conforms to their views and selectively ignore everything else. I posit that the notion that programming is an innate ability strokes the egos of those that, so they ignore that redaction. "I am special because we have some ability that makes us able to understand X, and you will never be able to, ever" is prevalent in every industry, and in humans and psychology in general. "It's all luck and chance" is more satisfying than "it's a lot of hard work and practice". It's the same thing with that "rockstar programmers" paper from the 60s, using a sample set of 11 and 20-minute interviews about interactive shell sessions to falsely determine that the best programmer is 10x more efficient than an average. It's of course an extremely dangerous attitude to have: people quit early because they think they don't have the brains (or the body parts), and people who are successful feel entitled because they are gifted.
|
|
#
?
Sep 12, 2013 03:40
|
|
|
Blunt Force Trauma posted:Hello, $20/hr is low, $200/hr is high (but not that high). Now, the $20/hr guy could produce great work and the $200/hr guy could be terrible. Judging cost by language is also a pretty terrible idea, the cost is more about the complexity of the project and the working conditions. Also, why are you using PHP/MySQL?
|
|
#
?
Sep 12, 2013 04:02
|
|
|
Seashell Salesman posted:No way the third one isn't mostly experience. Complete inexperience goes for a premium as well.
|
|
#
?
Sep 12, 2013 04:41
|
|
|
The retraction paper no more proves that there isn't a binary "getting it" than the original "proved" that there is. The two papers tested, and disproved, a possible test for programming aptitude (and if you're trying to argue that there is no aptitude for programming at all, then I'd argue that you have a strong aptitude for being a dumbass). Unfortunately, I'm not aware of any papers that have successfully established an aptitude test, so despite our speculation about what is important or hard to learn and what isn't, we really don't know.
|
|
#
?
Sep 12, 2013 06:12
|
|
|
In my opinion if you have the intellectual capacity to solve something like 3x + y = 5 -5x + 4y = 3 you are quite capable of learning to program, and how well you do is mostly a function of how hard you work at it, which itself is mostly a function of time (like most things). It's a relatively long route going from "hello world" to independently developing stuff that other people think is good and want to use, but subject to a reasonably low bar I think most people can get there if they work at it long enough. If they dont it is more likely due to Other poo poo than some inherent inability. unixbeard fucked around with this message at 08:04 on Sep 12, 2013 |
|
#
?
Sep 12, 2013 07:59
|
|
|
Also if you think code is art I suggest you go look at more art
|
|
#
?
Sep 12, 2013 08:00
|
|
|
You can make programming as hard as you want it to be. Easy is a bit more hit and miss.
|
|
#
?
Sep 12, 2013 08:32
|
|
|
baquerd posted:Fairly non-scientific studies have shown that either you "get it" out of the gate or you will never really get it. As has been beaten to death, there isn't such a thing as a natural born programmer. To co-opt Thomas Paine: "the idea of hereditary legislators is as inconsistent as that of hereditary judges, or hereditary juries; and as absurd as an hereditary mathematician, or an hereditary wise man; and as ridiculous as an hereditary poet laureate." nielsm posted:While programming theory is heavily bound to mathematics (many early computer science departments at universities were offshoots from the maths ones), the actual act of designing and writing software can really be closer to the arts, like composing music or painting. Not really, because in the arts you're often forced to appreciate and have a knowledge of the history behind the art and culture. There is a memory hole in technology. Meanwhile Dabblers and Blowhards is a wonderful rant against pg's "I'mma painter but with s-expressions, honest". quote:Give it a shot, you might have it. More accurately, give it a shot, you might enjoy it, and practice enough to be good at it. Hard NOP Life posted:People just spout poo poo that conforms to their views and selectively ignore everything else. I posit that the notion that programming is an innate ability strokes the egos of those that, so they ignore that redaction. Just quoting this and moving on. JawnV6 posted:I wholeheartedly disagree with most of that post (programming is math, not art), but this bit in particular I'd like some clarification on: unixbeard posted:Also if you think code is art I suggest you go look at more art Yes, computer science is mathematical, and yes, proofs and programs, but we can write poetry in english and legal contracts in english. The same is true for code, code is about communication, viz "Programs must be written for people to read, and only incidentally for machines to execute." and has all the opportunities other mediums have. Code is a opportunity to express design, art, engineering, as well as business logic. It is not to say that programming is art, but you can create art using code. Just look at the screenshot thread some time (or if you like, MOMA, which has a collection of computer art).
|
|
#
?
Sep 12, 2013 09:52
|
|
|
Zhentar posted:The retraction paper no more proves that there isn't a binary "getting it" than the original "proved" that there is. Zhentar posted:(and if you're trying to argue that there is no aptitude for programming at all, then I'd argue that you have a strong aptitude for being a dumbass) Which brings me on to quote:programming is math quote:programming is art quote:programming is X I'm not saying that programming is impossible to describe or anything, I'm just saying if someone asks or is curious, don't boil it down to a one-liner out of laziness or a desire to turn a phrase. It's doing both them and yourself a disservice. coffeetable fucked around with this message at 12:37 on Sep 12, 2013 |
|
#
?
Sep 12, 2013 12:34
|
|
|
So how would someone without a math background obtain the necessary skills? In two years I'll be a civilian again, so I want to prepare myself for school.
|
|
#
?
Sep 12, 2013 16:43
|
|
|
Elderbean posted:So how would someone without a math background obtain the necessary skills? In two years I'll be a civilian again, so I want to prepare myself for school. If the concern is that you'll arrive at school without the math background then two years is plenty of time to learn all the math you will need. Get intro books on algebra, calculus, and discrete mathematics.
|
|
#
?
Sep 12, 2013 16:58
|
|
|
Elderbean posted:So how would someone without a math background obtain the necessary skills? In two years I'll be a civilian again, so I want to prepare myself for school. Well, first you should work out just how valuable it'd be to you. There are two ways in which my maths education has helped me with programming. The first is as domain knowledge: sometimes I'll run across a mathematical problem that I was explicitly taught how to solve by my textbooks. How often this happens depends largely on where you're interests lie though, and I could imagine that many (possibly a majority of) very good programmers never have a use for anything more advanced than counting. The second is as a thought process: by the end my maths undergrad of it I was far, far better at reasoning about abstract objects than I was when I started. Maths, being almost exclusively composed of abstract problems, is a great way to pick up these skills. But I don't think it's the only one. Spending several years solving any kind of problem that you can't rely on your intuition for would develop the same processes. Having decided whether you want to learn maths as it's own subject or instead just learning what you need as you learn to program, the next questions are: how much maths do you know already? And: what's your preferred way of learning? (Textbooks/lectures/problem sets/etc)? coffeetable fucked around with this message at 17:06 on Sep 12, 2013 |
|
#
?
Sep 12, 2013 17:04
|
|
|
coffeetable posted:Having decided whether you want to learn maths as it's own subject or instead just learning what you need as you learn to program, the next questions are: how much maths do you know already? And: what's your preferred way of learning? (Textbooks/lectures/problem sets/etc)? I haven't used anything beyond basic math in years; by the time I get out it will have been nine years since I graduated high school. I don’t really have a preferred method of learning.
|
|
#
?
Sep 12, 2013 17:10
|
|
|
In that case, I'm going to make the mistake of assuming your methods of learning are similar to my own, and recommend Stroud's Engineering Mathematics. Your local library should have a copy (edition doesn't really matter). It starts with arithmetic and works its way up to multivariable calculus, and what makes it special is that it's self-contained and takes a very interactive approach: it walks you through an example of every single new concept it introduces, then gives you a few exercises on it: (This is about the third page of the book).
|
|
#
?
Sep 12, 2013 17:27
|
|
|
Elderbean posted:So how would someone without a math background obtain the necessary skills? In two years I'll be a civilian again, so I want to prepare myself for school. A lot of programmers overemphasize formal math as a prerequisite for programming. Probably because the type of programming they do a lot of (or at least the type of programming they find interesting) requires math. There's vast swaths of interesting (depending on your interests) programming projects that require very little knowledge of maths.
|
|
#
?
Sep 12, 2013 17:34
|
|
|
I took "prepare myself for school" to mean studying CS or something similar at university. I think what kind of study you are planning to do will determine what math fundamentals you need.
|
|
#
?
Sep 12, 2013 17:45
|
|
|
There is a somewhat dumb but nonetheless solid reason to have some math background before trying to learn programming: a lot of the lessons, assignments, and sample programs out there are centered on various problems from math, and if you don't know the problem, you'll find yourself trying to learn two things instead of one. For example, almost everybody teaching recursion brings up the Fibonacci sequence at some point. If you don't already know (or have forgotten) the Fibonacci sequence, that example is not going to feel natural to you, and it's easy to lose track of what's going on when the professor plunges forward without justifying it.
|
|
#
?
Sep 14, 2013 08:15
|
|
|
Just make sure you know what the hell a function is, i.e. not a wiggly line and two axis'. Apart from that the only math skill you're going to need as an absolute prerequisite is "don't be afraid to tackle hard problems". As in, don't give up because you don't know how to do it. Otherwise try to find your high-school calculus and algebra books.
|
|
#
?
Sep 14, 2013 14:58
|
|
|
I’ve heard that some schools focus on math and theory, while others lean toward programming. Which schools should I be looking at? I live outside of Seattle right now, I’d like to stay in the area if I can. Elderbean fucked around with this message at 20:03 on Sep 14, 2013 |
|
#
?
Sep 14, 2013 19:58
|
|
|
I can't speak from personal experience, but the University of Washington is a large and well-respected university. Their CS department probably runs everything on Windows for obvious reasons, but I'd imagine they have plenty of funding.
|
|
#
?
Sep 14, 2013 20:38
|
|
|
Elderbean posted:I’ve heard that some schools focus on math and theory, while others lean toward programming. Which schools should I be looking at?
|
|
#
?
Sep 14, 2013 20:52
|
|
|
Are there any good plain C libraries for JSON? json-c is the one that keeps coming up in my searches.
|
|
#
?
Sep 14, 2013 23:47
|
|
|
more like dICK posted:Are there any good plain C libraries for JSON? json-c is the one that keeps coming up in my searches.
|
|
#
?
Sep 15, 2013 15:06
|
|
|
I get two errors from rvm, when I try to update my version of Ruby: an Xcode version error and a permission error. Since my Xcode version is newer than the log suggests, and since I am logged in as the main admin account, I can't make sense of the error message. The full message is available here with some added commands thrown in to highlight the problem.
|
|
#
?
Sep 15, 2013 15:53
|
|
|
Make sure you have the latest command line tools installed (Xcode Preferences > Downloads > Components), then run `brew doctor`. Otherwise read homebrew's troubleshooting wiki.
wolffenstein fucked around with this message at 16:02 on Sep 15, 2013 |
|
#
?
Sep 15, 2013 15:59
|
|
|
wolffenstein posted:Make sure you have the latest command line tools installed (Xcode Preferences > Downloads > Components), then run `brew doctor`. Otherwise read homebrew's troubleshooting wiki.
|
|
#
?
Sep 15, 2013 16:15
|
|
|
I missed the part where you said you were using RVM. You should ask the Ruby thread. You should also learn CLI.
wolffenstein fucked around with this message at 16:27 on Sep 15, 2013 |
|
#
?
Sep 15, 2013 16:23
|
|
|
more like dICK posted:Are there any good plain C libraries for JSON? json-c is the one that keeps coming up in my searches. nvm, shrughes reminded me why it's not good Scaevolus fucked around with this message at 05:26 on Sep 16, 2013 |
|
#
?
Sep 15, 2013 18:14
|
|
|
No, cJSON is ABSOLUTE poo poo DO NOT USE CJSON.
|
|
#
?
Sep 15, 2013 23:19
|
|
|
Scaevolus posted:
shrughes posted:No, cJSON is ABSOLUTE poo poo DO NOT USE CJSON. I'm curious to the reasons so I know what to look out for in the future!
|
|
#
?
Sep 16, 2013 10:06
|
|
|
Adding elements to the end of an array takes linear time, adding attributes to an object takes linear time. Numeric parsing gets the rounding incorrect. Crazy logic with valueint and valuedouble fields. An insane API, insane internal APIs, absurd coding style. Printing a number has a buffer overflow: code:
|
|
#
?
Sep 16, 2013 10:41
|
|
|
When is that gonna overflow?
|
|
#
?
Sep 16, 2013 19:18
|
|
|
A 21 character string needs 22 bytes.
|
|
#
?
Sep 16, 2013 20:19
|
|
|
It doesn't outright say it (the comment is misleading and dumb), but the maximum representable integer value requires 20 characters to store, so 21 bytes.
|
|
#
?
Sep 16, 2013 20:32
|
|
|
(Negative integers use an extra character.)
|
|
#
?
Sep 16, 2013 23:33
|
|
|
shrughes posted:(Negative integers use an extra character.) That was my guess but it's been so long since I used the printf formatting I'd forgotten what %d did with negative numbers and was afraid to out myself as an idiot  vvv upon actually checking the smallest 64 bit signed int you are correct, 19 digits + negative sign + end of string character = 21. Seashell Salesman fucked around with this message at 00:20 on Sep 17, 2013 |
|
#
?
Sep 16, 2013 23:48
|
|
|
shrughes posted:(Negative integers use an extra character.) The smallest 64-bit signed integer takes up 20 characters, including the negative sign. So unless you have some magical 65-bit signed integers, I'm not seeing the overflow.
|
|
#
?
Sep 16, 2013 23:58
|
|
|
Okay I have a pretty specific question. Say you had a huge amount of data with a tree-like relationship, maybe for example a list of all species names (and their respective genera, families, etc.). And say you wanted to make queries like, "Output all children (and grandchildren and so on) of the class Mammalia" (or phylum Arthropoda or whatever the hell). What kind of data structure would you want for that? I'm completely inexperienced with hash tables but I'm pretty sure that's the right answer. A straight-up tree data structure seems like it might make traversing and outputting the data simpler/faster but after doing some research it looks like it might be better to do some kind of hash list or even a hash tree to have pointers(?) to a child's parents or adjacent siblings (assuming there is ordering between siblings) with an added bonus of finding the starting node I want to begin with in a reasonable amount time. If there aren't any changes made to the data set whatsoever it makes me think that some augmented hash table is probably the way to go but I'm still unsure. If hash tables do the trick I'm open for any book recommendations or other references that might put me on the right track. 
|
|
#
?
Sep 17, 2013 03:22
|
|
|
|
| # ? May 31, 2024 18:29 |
|
|
You wouldn't want to grab all the data and then futz with it. Use a search algorithm to find the item you are starting from, then recurse over the children of the item.
|
|
#
?
Sep 17, 2013 03:30
|
|
