15 ms·
The Mathematical Hacker
- kenko 14y ago" long int fac(unsigned long int n) { return lround(exp(lgamma(n+1))); } " Sounds great! What's the factorial of 40? Let's use some python: >>> int(math.exp(math.lgamma(41))) 815915283247882431423526575245034982027017846784L Huh, weird. I would've thought the factorial of 40 would be divisible by ten. Let's try it another way: Prelude> let fac n = foldr1 (*) [1..n] Prelude> fac 40 815915283247897734345611269596115894272000000000 The author's disdain for the lambda calculus, as if it is somehow not real math (because it doesn't involve numbers in the familiar sense?), is bizarre.
- sbi 14y agoWell, lround(exp(lgamma(...))) is just about the worst way to use lgamma. If long ints are signed 64 bit integers, then the largest factorial you can store is 20!. You might as well look it up in an array. But you frequently need ratios of factorials (and values of the gamma function more generally) when doing statistical computing and lgamma is invaluable.
- ninetax 14y ago>After coding his recursive solution, the Lisp hacker is more likely to ask the irrelevant question: how can I reduce these two functions down to one function? Why would the Lisp hacker ask that? That makes no sense. There is no reason to combine those two functions.
- jrajav 14y agoI think this article raises a great point overall, but the Fibonacci example is a bit weak and smacks of a strawman. Functional programming tutorials that teach you to write a recursive Fibonacci or factorial calculator are not doing so with the purpose of teaching a useful function, or even a workable one - in fact, more often than not this is called out by showing its performance for moderately large N. They only use those examples because they're simple, abstract problems that can be used to elegantly demonstrate recursion in a relatable way. If we're going to talk about real world, practical applications, then it might be more relevant to bring up a problem for which there isn't a trivial O(1) solution. A dynamic programming problem, perhaps? It's glossed over with a terse dismissal: > “Advanced” discussions might consider engineering considerations such as tail-call behavior, or the possibility of memoizing results. ... But this is exactly why a functional programming language might be better than an imperative for certain applied mathematics problems. Note, I'm not saying that it is better, just that it's not such a simple argument. With a different set of examples, this article could have been written in favor of Lisp just as easily. I still think it's a good article overall, just that the point it tries to sneak in about Lisp and functional programming is a little shaky.
- norswap 14y agoThe article casually confounds what it is programs do and how programs are written. Programs operate on data, and said operations can be more or less mathematical in nature. Writing programs now, that's not mathematical in nature at all. There are an infinite number of programs that compute the same result. A program is good because it meets a number of criteria, among which efficiency and math soundness are not always the most important. Many of the criteria have to do with how the human mind works and how it intuits.
- frozenport 14y agoYou don't need to be good to program. You need to be exceptional to do math. Most programmers aren't that good at what they do, and I think they should focus on their craft. I struggle to explain to my colleagues who make 100k that they need to do memory tiling: or that their loop doesn't vectorized. Changing a float from a double in a BFGS is 2 days of work for these people. Let the chosen few do math.
- 11001 14y ago>Let the chosen few do math And then you wonder why an average person hates/is scared of math.
- new_test 14y agoAs a programmer you may not need to use algebra or analysis on a daily basis, but as a rule I've been noticing that the solutions that are more mathematically elegant also tend to be more pragmatic.
- p4bl0 14y agoI find it disturbing that this article has been written in 2012. While I was reading it I really thought that it was at least 10 years old. Computer science researchers are doing actual mathematics, and they are clearly more in what the article calls "Lisp school" than the "Fortran school". Research in functional programming is mostly mathematics. Lambda calculi (which was originally not developed to be a programming language at all), types, logic, category theory… Most of the tools used in programming languages research are mathematical tools, and they always have been. The article is great and I mostly agree with its conclusion, but it seems to miss a lot about what is actually going on in computer science. I understand that it mainly talks about programmers and not computer scientists, but it's less and less true that programmers are not interested in what is happening in computer science research. Lisp invented a lot of the stuff that are now in almost every languages (conditionals, garbage collector, first class function, everything is an expression, recursion…) and it took many years to arrive here. But nowadays I see a lot of interests from the programmer community for what is happening in the programming languages research around Haskell, Scala, OCaml, and even stuff like Coq, Isabelle… ! In order to understand even a little what is actually going on with these research project one have to study (even if indirectly) mathematics. Now that I have written this comment, I'm thinking that may be I have a skewed/distorded vision and that I'm not really talking about the same people as the article does. I don't know.
- tel 14y agoI feel somewhat similarly to you about the wonderful influx of math via the functional languages these days. I think this article misses the boat on this one when the author considers that lisp hackers avoid math by seeking abstraction---when really there's a lot of math there too.
- BruceIV 14y agoYes, there's lots of great math in programming languages research (it's my field, and I quite enjoy it), but I think the point was that it's only being used to solve "our" problems, not the domain problems that people fund CS to solve.
- 11001 14y agoDidn't go over so well the first time: http://news.ycombinator.com/item?id=4796586 http://news.ycombinator.com/item?id=4796586
- coolSCV 14y agoCertainly it was the missing "the" in the title.
- S4M 14y agoI don't really see the point of this article. You don't need Fortran or C to implement the calculations of the Fibonacci as described in the article. On top of that the article is missing the point that some elements functional programming (map, lambda) are actually making a numerical implementation neater. It is not an accident that the original authors of R were themselves lispers and admitted having been inspired by lisp when creating R. I think among others, those links will illustrate the previous point: http://cran.r-project.org/doc/html/interface98-paper/paper.html http://cran.r-project.org/doc/html/interface98-paper/paper.h... http://www.stat.auckland.ac.nz/~ihaka/downloads/Compstat-2008.pdf http://www.stat.auckland.ac.nz/~ihaka/downloads/Compstat-200...
- jpitz 14y agoThe article didn't claim that FORTRAN or C are required for these calculations. Rather, It lamented that the prevailing attitude of Lisp practitioners was to eschew the knowledge behind those calculations.
- erichocean 14y agoThey seem to agree on one thing: from a workaday perspective, math is essentially useless. I would love it if someone had resources on how mathematics could actually be used as a tool by programmers. What kind of problems would mathematics as a tool help me solve better/faster/more efficiently than the other tools in my programmer toolbox?
- Jacobi 14y agoThere are plenty. The analysis of basic algorithms involve math (e.g. QuickSort, binary search, etc). Programming a user friendly website can involve advanced mathematics, for example you want to recommend to your users a product that might interest them (Collaborative Filtering)
- meaty 14y agoAlgebra is required all the time for me for pretty much any non trivial problem that isn't CRUD (most apps seem to have no functional complexity past CRUD). The only time I had to delve into deep mathematics was implementing CORDIC algorithm for microcontroller powered floating point ops (sin/cos/ln) on a 68HC11 because we couldn't buy an implementation in that had source code so we could verify it. Even then it wasn't all that hard. Took about a week to wrap my head round the maths involved. I do find that mathematical literature is all theory and no application, even with my engineering background. If we had some applications, people would use it more and take it seriously as well and therefore there would be more mathematical programmers.
- wolfgke 14y ago> I do find that mathematical literature is all theory and no application, even with my engineering background. If we had some applications, people would use it more and take it seriously as well and therefore there would be more mathematical programmers. I have a mathematics background and I see immediately thousands of not-implemented applications when I read mathematical literature (especially the highly abstract one). Thus applications are there. But getting the abilities to see these applications takes time and dedication. I try to hint lots of colleagues to these connections and applications and get ignored (I could talk to a wall instead - it wouldn't make a difference). The colleagues only want to get their job done somehow and nothing more...
- krutulis 14y agoAs I see it, an essential point of "Data Science" is to finally introduce mathematical awareness to the world of programming. I welcome our newly arriving & refreshingly numerate overlords! "One way to read the history of business in the twentieth century is a series of transformations whereby industries that 'didn't need math' suddenly found themselves critically depending on it." This is a significant understatement. It also summarizes the history of business in the 19th century and even earlier. See, for example: steam power, metallurgy, bridge building, chemical industries, and electricity. Not all hackers miss Miller's point. One of Zed Shaw's older rants "Programmers Need To Learn Statistics Or I Will Kill Them All" comes to mind[1]. Perhaps I'm missing some intended irony, but Miller's polemics on Lisp and his reverence for Fortran feel overdone & unnecessarily narrow. Coders who buy into Miller's notions can join the "Fortran School" by learning something along the lines of R programming (lots of fun, functional, array-based, and adept at statistics) or J (compression and austere beauty that only a mathematician could love) or any number of other languages that have not shunned math. Of course, calculus, statistics, and linear algebra also required. For the university trained, brush up (or curse your university for not requiring you to learn these subjects). For the autodidacts among us, the likes of Coursera and Sal Khan are indispensable. [1] http://zedshaw.com/essays/programmer_stats.html http://zedshaw.com/essays/programmer_stats.html
- pdmccormick 14y agoI've taken some mathematics and logic courses at a university level, and the thing I took from those experiences and apply everyday is the reasoning and proof structuring. Struggling through and coming to something of an understanding of the concepts of analysis (calculus) and algebra were a pivotal point in my intellectual development. As somebody for whom math has never come easily (and still doesn't), but for whom the beauty and wonder of it all was still enchanting, I'd heartily encourage any and everyone to learn some higher maths in detail. The specific results may not always be applicable day to day, but the way of thinking will stay with you. Good luck!!
- mgnagy 14y agoWhile an interesting read, it struck me as an odd, we've taken the wrong fork, nostalgia piece. If you look at the very early days of computing, you tended to find three types of degree/backgrounds: Electrical Engineering, Mathematics, and Philosophy. The double E is obvious: building the kit. The other two, the potentially Left Brained or Right Brained approach to programming. All three of them have one thing in common: A structured, _logical_ approach to solving problems. If there is anything to lament, it is that we still have the tug of war between left and right brain approaches. The reality is that one size, indeed, does not fit all. We should accept that and move on to getting the work appropriately done.
- abrezas 14y ago"Despite the aesthetic virtues ascribed to functional programming, I find the preceding solutions to be more beautiful than their recursive counterparts. They run in constant (rather than linear) time, and they are easily adapted to work with non-integer inputs." Isn't this wrong? I don't think pow is computed in constant time.
- Jacobi 14y agoI think that it can be implemented in constant time for floating point numbers. For integers I know that the exponentiation by squaring takes O(log(n))
- Patient0 14y agoI think many floating point functions (e.g. sin, cosine) are implemented using Pade rational approximations - basically, the ratio of two polynomials. (http://www.dattalo.com/technical/theory/sinewave.html http://www.dattalo.com/technical/theory/sinewave.html) This usually gives enough accuracy for the purposes of floating point. However, I'm not sure if "pow" can be usefully implemented this way. I am guessing no, because pow grows faster than any polynomial eventually... edit: Hmm, pow at least looks linear here: http://www.netlib.org/fdlibm/e_pow.c http://www.netlib.org/fdlibm/e_pow.c
- rartichoke 14y agoMath seems extremely important to me as a programmer, even for CRUD apps. Let's say you graduated HS and slept through your algebra classes. Now a few years later you're a programmer with pretty much no math skills other than what you've learned in elementary school. Math seems to teach you one of the most fundamental and useful things about programming. The ability to reduce a problem from something that is complex into something that is not complex. I've only gotten a taste of some basic algebra after taking some online CS courses and really, coming into it with a background of "embarrassingly poor math skills" I can really say that it has changed my life for the better. I'm still clueless when it comes to some algebra but I find myself looking at problems and being able to solve them much easier now and this is only after a few weeks of programming related courses that happen to use algebra on some occasions. The math isn't what made it easier. It's applying the same things to solve math problems to programming.
- irahul 14y agoPicking on Fibonacci of all things? The goal of fibonacci and factorial examples are to teach recursion. Both fibonacci and factorial are good starting points for a beginner. It can be followed by discussions of dynamic programming where the student can be introduced to recurrence relations and solving them top-down and bottom-up. EDIT: Adding some background on dynamic programming For dynamic programming, the problem should be breakable in terms of overlapping smaller problems, and the base case should be recognized. If the problems don't overlap, they fall within broader divide and conquer category(mergesort, quicksort etc are famous examples). fib(n) is defined as fib(n-1) + fib(n-2) (overlapping smaller subproblems) and fib(0) = 1 and fib(1) = 1 (base cases) fib(n) = fib(n-1) + fib(n-2) fib(0) = 0 fib(1) = 1 A relation defined as above(recursively) is known as recurrence relation. Discrete math courses deal with finding closed form expression - a non-recursive function of n. But in programming, we are fine with solving the recurrence relation without finding a closed form expression. Recurrence relations form the basis of dynamic programming and they can be solved either top down or bottom up. The top down approach is the traditional recursive solution. def fib(n): if n == 0 or n == 1: return n return fib(n-1) + fib(n-2) And then the student is to realize fib(n-1) is recalculating fib(n-2) and memoization is in order. def fib(n): cache = {0: 0, 1: 1} def _fib(n): if cache.has_key(n): return cache[n] cache[n] = _fib(n-1) + _fib(n-2) return cache[n] return _fib(n) Then the student should realize modifying every function isn't apt, and should implement a general memoize decorator. EDIT: Adding table based bottom up fibonacci. Now once the student understands top down dynamic programming, as in he can find the recurrence relations and base cases, it's time for bottom up. As the name suggests, bottom up starts from the bottom and calculates n compared to top down which starts from n and boils down to base cases. def fib(n): vals = {0: 0, 1: 1} for i in range(2, n+1): vals[i] = vals[i-1] + vals[i-2] return vals[n] Student should recognize how top down and bottom up are calculating the same recurrence relation, but in a different order. The table vals here is the same as cache above in top down. Top down is recursive and might trigger the recursion limit. Bottom up doesn't have the recursion problem. Sometimes in case of bottom up, table can be eliminated depending on the overlap. But the important thing is, once the recurrence and base cases are known, it can be implemented quite easily. In fibonacci's case, nth number depends only on n-1 and n-2 and maintaining the whole table is wasteful. The bottom up approach will be better. def fib(n): f0, f1 = 0, 1 for i in range(n-1): f2 = f0 + f1 f0, f1 = f1, f2 return f2 Fibonacci just happens to be one of the problems used to demonstrate recursion and dynamic programming. It's small enough for a beginner to comprehend, and big enough to explain recursion and dynamic programming. The article picks one recurrence which has a closed form expression. The dynamic programming problems which I have encountered aren't that easily reduced to closed form expressions. Also, I don't know about Graham or Raymonds, but Yegge advocates maths for programmers. http://steve-yegge.blogspot.in/2006/03/math-for-programmers.html http://steve-yegge.blogspot.in/2006/03/math-for-programmers....
- Dn_Ab 14y agoThe problem with the article is that it dreams up a false conflict and then buries its thesis, a very valid and solid one, by trying to force a contentious narrative onto that just isn't there. Very few people these days are ignorant of the points he makes and neither Yegge nor Graham are posts of some imagined counter applied math camp. I would tend to think they are pro - Yegge's interested in bio, Graham did spam stuff. Here is my attempt at a summary that captures all the detail: Programmer math is not only restricted to type theory and formal logic. Applied Math is also a very important aspect. Eric S Raymond is wrong for trying to make it look like the 'Except for X' is negligible. Plus, machine learning, bio, geo and so on are increasing in importance and growing rapidly as fields. Even in the past, applied Math Programs had a massive impact on the economy by enabling engineers with specialized software. Learn Math. It is important for you as a programmer going forward. --------- I am not sure why he restricts functional programming to recursion with lisp. There is no need to denigrate functional programming, it fits extremely well with these applied math problems and helps a lot with reducing complexity of implementation, in my opinion. They also tend to be at least as fast as Java with OCaml posting incredible single core speeds. Another point is math algorithms is distinct from mathematical programming. Math helps one calculate and choose faster algorithms or find better bounds. But except for some portions of the haskell compiler there are few uses of direct calculation to derive programs mathematically. In engineering, math allows you to precisely calculate behaviour and properties but in programming you have to debug. The key reason for this divide is that other engineers have lots of components with well defined properties to work with. This is another reason why functional programming is more mathematical. Although the idea of composition is not inherent to FP, in FP it is the default paradigm. The ideas of combinators with well defined properties, and theorems proven on them, that only go together in a certain way and that you can sit down and have a pretty good idea of how your program would behave theoretically, this idea is what makes FP so close to doing applied math. And Haskellers really shine at that. Haskellers like to talk about monoids and categories but really most of haskell programming is closer to what a reguler engineer does with the category theoriests serving the same function as physicists.
- sirclueless 14y agoTo add to your claim that math is fundamental to computer science, I think there's a big factor that isn't being mentioned: knowing math future-proofs your career better than knowledge of any technology or language. Mathematical depth is the irreducible core of a computer scientist's value. Fundamentally, what computers provide is the mechanization of knowledge. In the same way that robotics came along and mechanized all of the entry-level manufacturing jobs (the root of the current "skills gap" in manufacturing, more so than outsourcing in my opinion), over the coming few decades I expect that computers will mechanize most entry-level knowledge workers' jobs, if they haven't already. And the wacky thing about comp. sci. is that our entry level positions are in fact basic knowledge-work. As the field matures, the fields that are the bread-and-butter of basic vocational computer science are increasingly swallowed by more sophisticated technical solutions that are more mechanized and productized. Comp. Sci. is an ouroboros, eating its own tail of unskilled positions. The most obvious example currently is sysadmins: computing infrastructure is being productized and sold under the umbrella of "cloud computing," and its main value proposition is that you can cut your sysadmin budget by a huge margin. Similarly, QA and software testing departments are having their lunch eaten by better testing practices and automated solutions such as CI: the jobs aren't disappearing entirely, they are just being consolidated into a tiny department (maybe even one dev, part time) that can maintain an automated infrastructure. At a more basic level, data entry positions started drying up years ago with the advent of OCR, and now most of the big easy targets are entirely mechanical (ex. the post office/shipping). In short, the best way to provide enduring value is to provide mathematical insight that can't be mechanized. If you can recognize optimizations and opportunities that take advantage of the unique structure of your domain, you are unlikely to be swallowed by someone with a productized version of your livelihood. If all you can manage is CRUD apps for enterprise, then when someone comes along with a general product that reduces your department from many guys writing software to one guy managing the product, you aren't gonna convince anyone of your value. At that point your only defense is the behemoth of bureaucracy, but the competitive nature of business suggests that it won't be a great defense for long.
- louischatriot 14y agoVery interesting thesis. But why did you use the two very specific examples of the factorial and fibonacci sequence? They don't strengthen your (again, very interesting) point but rather dilute it.
- michaelochurch 14y agoI'm in the Yegge school, so far as I think that the mathematical ignorance and, more generally, anti-intellectualism of our industry is its downfall. We see it in the lack of design sense and the awful code that is produced. VisitorFactory nonsense is something that was invented by people who hated math and wanted to tear programming away from its mathematical/problem-solving roots with a bunch of junk complexity that adds nothing. If you're doing simple things over and over, with the objective function being minimal rejection/loss rate, the anti-intellectual/industrial get-your-boss-off-your-back-oriented development works. Complex work is different. You need to invest a lot of additional energy into elegance and quality that most companies won't pay for. However, what we do is so intensely collaborative if it's done right that we are teachers more than anything else. Not engineers or programmers. Our job doesn't end when we build the thing: we have to teach people how to use it, so that it's not mindless complexity being inflicted upon them (as is true of many enterprise products) but something that actually makes their lives better (most software doesn't, because bosses have a tendency to force people to use bad software and this makes engineers' lives, and thus products, worse). It's sad to me that this is an extreme minority viewpoint, and that almost no software manager on earth will actually pay for things to be this way.
- oinksoft 14y agoI'm not sure why you think that math is somehow divorced from design patterns (in the abstract sense, not GoF). Some systems just need more moving parts than others, and math won't save you from that. Design sense and mathematical aptitude are complementary, not opposed, to one another. I think that "industrial get-your-boss-off-your-back-oriented development works" and "Complex work" is a false dichotomy.
- michaelochurch 14y agoMathematics isn't about numbers or even proofs. It's about simplifying until you get that class of problems down to a conceptual "nub", while thinking in a very rigorous way. Instead of five oranges plus nine oranges being 14 oranges, you just say 5 + 9 = 14, and that applies to any countable quantity. It's because of these abstractions that we can predict the motions of planets or build computers. This also explains the "why" of functional programming. Referentially transparent functions are just simpler than methods that are constantly changing. Of course, in the real world, it's neither desirable nor possible to eliminate mutable state: you sequester and manage it. The opposite of mathematical rigor (defined semantics, reproducible behavior) is the sloppy, ad-hoc, corporate FactoryVisitor stuff that is tolerated under the assumption that programming has nothing to do with math.
- szany 14y agoThere's also the simple fact that programming literally is math. http://ncatlab.org/nlab/show/computational+trinitarianism http://ncatlab.org/nlab/show/computational+trinitarianism
- btilly 14y agoProgramming is math, in the same way that accounting is. Doing programming all day will not make you a mathematician.
- szany 14y agoProgramming is math, in the same way that accounting is. If you read the link that's not what it says. Doing programming all day will not make you a mathematician. Sad truth. For now!
- ekm2 14y agoThe irony is that the best way to find out if mathematics is actually useless to either programming or real world problem solving is to know a lot of it.Just staring at it from a safe window and then making catholic pronouncements does not help.
- alexkus 14y agoAs someone with both a Bachelors degree in Computer Science and a Bachelors degree in Mathematics I'd agree. My CS degree was quite theoretical (completed in 1999) so it wasn't just being taught the language of the day (we did x86 assembly, Modula-2, Prolog, Lisp, C++ and Occam). The Mathematics degree has definitely been useful for programming; specifically set theory, matrix manipulation and linear algebra, number theory, and graphs/networks/design, but a lot of it is stuff I'm unlikely to use for usual programming areas. Number theory (and implications within assymetric encryption) is where I'll probably look to continue learning (as part of a Masters degree) if I find the time.
- killnine 14y ago"advanced mathematics seems completely unnecessary to existing practitioners, but only until someone figures out that a particular mathematical concept is the right way to think about the problem at hand."
- Someone 14y ago"If you are a systems and network programmer like Raymond, you can do your job just fine without anything more than multiplication and an occasional modulus." A good network programmer should have at least passing knowledge about graph theory (the network is a graph), queueing theory (how else are you going to size your buffers?), and also some statistics (given these numbers, how much bandwidth will be taken by resends? How much data can we send per hour? How many 9s do we get for 'probability that a messages is handled within 10 seconds'?) Good systems programmers should have some knowledge of all of these for essentially the same reasons. They also will need some knowledge of Petri nets (for deadlock avoidance), formal languages (how else are you going to spec a language or some complex protocol?) Also, an article on this subject that mentions neither Knuth nor Dijkstra? Weird. Is that because they do not have a blog (yes, Dijkstra does not live anymore, but I doubt he would have a blog if he did)?
- tzs 14y ago> A good network programmer should have at least passing knowledge about graph theory (the network is a graph), queueing theory (how else are you going to size your buffers?), and also some statistics (given these numbers, how much bandwidth will be taken by resends? How much data can we send per hour? How many 9s do we get for 'probability that a messages is handled within 10 seconds'?) Partial differential equations can also be useful there. From Danny Hillis' TEDxCaltech talk about Feynman's work on the Connection Machine, where Feynman was figuring out how many buffers they needed in the routers: --------------------------- By the end of that summer of 1983, Richard had completed his analysis of the behavior of the router, and much to our surprise and amusement, he presented his answer in the form of a set of partial differential equations. To a physicist this may seem natural, but to a computer designer, treating a set of boolean circuits as a continuous, differentiable system is a bit strange. Feynman's router equations were in terms of variables representing continuous quantities such as "the average number of 1 bits in a message address." I was much more accustomed to seeing analysis in terms of inductive proof and case analysis than taking the derivative of "the number of 1's" with respect to time. Our discrete analysis said we needed seven buffers per chip; Feynman's equations suggested that we only needed five. We decided to play it safe and ignore Feynman. The decision to ignore Feynman's analysis was made in September, but by next spring we were up against a wall. The chips that we had designed were slightly too big to manufacture and the only way to solve the problem was to cut the number of buffers per chip back to five. Since Feynman's equations claimed we could do this safely, his unconventional methods of analysis started looking better and better to us. We decided to go ahead and make the chips with the smaller number of buffers. Fortunately, he was right. When we put together the chips the machine worked. The first program run on the machine in April of 1985 was Conway's game of Life. --------------------------- Source: http://longnow.org/essays/richard-feynman-connection-machine/ http://longnow.org/essays/richard-feynman-connection-machine...
- sdfqwer 14y agoThe mere fact that SICP (Structure and Interpretation of Computer Programs) and SICM (Structure and Interpretation of Classical Mechanics) were written by the same person should be enough to give the lie to Miller's thesis. Although willful ignorance of math is clearly a problem among certain segments of industry, it's just plainly not true in general and certainly doesn't correlate with lisp.
- cs702 14y agoThis rings True to me. It also represents potentially really good news, because it means that we are probably at the beginning of a huge wave of wealth creation as we figure out how to apply mathematical knowledge to the massive growing mountains of data generated by everyone and everything everywhere. Think about the historical impact that statistical-quality-control software has had in manufacturing, or the impact that economic-forecasting-and-optimization software has had in farming, or the impact that 'linear programming' software has had in supply-chain management and logistics; and then extrapolate these past achievements to get a sense of the potential impact that applications of mathematical knowledge could have in an everyone-connected, everything-networked, everything-measured world. Exciting times.
- b_emery 14y agoYes! This is what had me thinking that this article is the best one I've read all week. It's not about language wars, it's about finding the new applications of math. With the data and computational power at our disposal, big questions are going to fall and big industries will be disrupted.
- pmelendez 14y ago> "If you are a systems and network programmer like Raymond, you can do your job just fine without anything more than multiplication and an occasional modulus. " So a systems programmer would never use a tree, a heap or any not trivial data structure? Since when graph theory is not part of math? I am with the author on this, those statements are fairly simplistic. In my experience, a strong math background is the difference between an A and a B/C programmer.
- darec1 14y agoPointless rant by some math student, or some programmer who feels he is missing out on math. Unclear on demands, probably wants to teach more math to kids.
- papsosouid 14y agoAm I the only one who was put-off at the first paragraph calling Yegge and ESR "gifted essayists"? Some people genuinely are gifted essayists. People who write incredibly long-winded essays that jerk wildly to and from seemingly random topics are not among them. I'd say PG actually does qualify, his essays are generally well thought-out, and able to convey a concise, clear point. How does one possibly group PG and Yegge into the same category as writers?
- juiceandjuice 14y agoThe first real library I had to learn to use was GSL. The first framework I ever learned to use was ROOT. I also had to learn how to use FFTW and all sorts of other math libraries. I've had to hack fortran a bit, but mostly just know how to read it and compile it with g77 and stuff because that's what my professors would program in. In my only real course in programming at a university I learned how to link C to fortran libraries. I used Maple way more than matlab, and all of this was the majority of my programming training before I graduated from college. If there is a divide between fortran and LISP hackers, I'd have to say that Python is going to be the adopted parent/mentor, and it's the only programming language truly situated to bring them together. This is why projects like PyPy get me excited. Other projects like Julia are really interesting as well. I have to agree with the premise of the article though. There is a divide between mathematical hackers and non-mathematical hackers. I don't think it's anti-intellectualism per se, it's possibly more of a result of the increasing stratification of backgrounds in an industry which has increasingly focused on products, services, and most importantly consumers and "social". I went to a talk yesterday by an veteran applied mathematician who does research at HP labs. He mentioned something that really resonated with this notion: The days of academic style research, of Bell Labs and the like, has disappeared. This isn't to say that research and math has disappeared from the industries, but as a researcher you have to justify the research by explaining how it can contribute to a bottom line for the company. There's a few companies with looser restrictions on this (Microsoft, Google), but typically this kind of research is best done at companies that are monopolies with high profit margins. In that case, a company like Apple would be best situated to conduct research on the level of the Bell Labs of the past, but it's evident that they don't. The mathematical hacker started falling out of vogue sometime in the 70s. If the mathematical hacker ever becomes vogue again, it won't be because of Apple. It will be because of big data, autonomous cars, medicine and bioinformatics.
- agentultra 14y agoI found the thesis of this essay to be a foregone conclusion. We know that math is a universal language that, as far as we know, is capable of describing every phenomenon in the universe. And while Computer Science (if you can even call it a science) has had a storied history with mathematics it's not unusual to me to see maths used as a tool rather than the foundation of study. It's the same in many other fields such as physics, engineering, chemistry and so on -- mathematics is a useful tool and nothing more. In Computer Science mathematics is a useful language for discovering and discussing the properties and consequences of algorithms and their computation. It stands to reason then that if one increases ones knowledge of mathematics then one will have a larger vocabulary to work with in their area of study. If you're studying physics then the more math you know the more complex phenomenon you will be able to model and discuss. In Computer Science you will be able to find efficient applications and optimizations of various algorithms and model physical and ephemeral phenomenon with more rigour and precision. However there is a maximum lower bound on the amount of mathematical knowledge required in order to effectively practice programming. That is where I think the meme, "you don't need to know math," comes from in our field. More often than not the primary concern of a software developer is to balance two primary concerns: "does it work as it is expected to?" and "can another human being understand and maintain this?" Having a broader understanding of mathematics will certainly open your horizons but there's only so much you need to know to get by. Most of your "work-a-day" programming won't involve much math at all and is primarily concerned with APIs and implementation issues. I personally believe that Computer Science certainly needs more rigour and mathematics should be stressed a little harder in our curriculum. However in a "hacker" culture rigour is the least of your concerns. The hacker is a pragmatic creature and heavily leans towards asking the question, "Does it work?" more often than, "is this the right way to do it?" See the "New Jersey" school of development (or "Worse is Better" philosophy). As a field I think we'd benefit from a stronger approach to rigour. We are currently mired in a populist culture that constantly re-invents the wheel every decade or so. It'd be nice if we had a common literature and history from which to draw upon as part of our standard curriculum. Interesting article... but I don't think the treatment of lispers is quite fair. :p
- minton 14y ago"Although braggadocio doesn't come naturally to most computer programmers..." Ha! Made me laugh.
- nerdfiles 14y agoMathematics, Programming, Logic: We are looking at, like with the philosophical beginnings of the 20th century a question found in Hilbert, Russell, et al. Is mathematics reducible to a base system of logic. Today, we might see the cleanest expression of a theoretical framework answering to this question: free logic. At the same time, programming itself, like mathematics, is not a world of numbers or any of this, and is in fact the expression of complexity: the complexity of the sign as written or drafted by the mathematician, itself expressed in a non-denumerable set. Programming itself depends on the conventional expressions and idioms of programmers who write actual code. As Wittgenstein might say “Mathematics consists of calculations, not of propositions.” Like so: "Programming consists of executions, not of functions."
- B-Con 14y agoJust as an aside, I dual-majored in both pure math and C.S. I found the types of thinking for both subjects to be very, very similar. In fact, finding a constructive proof for a math theorem was almost like writing code. I started programming fairly young, about 12, so the "constructive" mathematical approach was the most natural to me. By my third year at university it had worn off and I was much more comfortable with things like existence proofs and thinking in terms of relations rather than constructions (a very vague description, but it will suffice for now), but programming always felt like it emphasized the constructive part of mathematical thinking. It was like I went to a math class and thought one way, then a C.S. class and emphasized a large subset of that thinking.
- zem 14y agoi was amused by "calculus (the real kind)" to distinguish it from usages like "lambda calculus". don't think i've ever seen that one before, though it parallels usages like "a real doctor (not a phd)" and "anything with a science in its name isn't".
- bitanarch 14y agoWhat this article essentially says is one kind of math is better than the other. No.
- KirinDave 14y agoA fine essay, except that it ignores that the "workaday" programmer actually hates high degrees of abstraction and ignores it. No, seriously. We're taught in mainstream programming languages to temper our desire for abstraction with mechanical sympathy; being careful not to build too high a tower of abstraction lest our CPU god grows angry and knocks us down to earth with miserable runspeeds. Even if your algorithm is right, failing to fit in cache or having too many memory fetches can increase your code runtime by multiple orders of magnitude from where an optimal, machine-aware piece of code should be. And the languages that do try to bring more abstraction to the table are ironically dismissed as "too academic" or "too strange" or "not practical" by most people in the industry.
- ianstallings 14y agoCan we at least restrict the word "hack" to using something in a way it was not intended? It's being thrown around for everything and it's downright confusing.
- mcgwiz 14y agoThe case for relying on math more when programming is one pillar of a more general aim: for programmers to be more creative in their approach to solving problems programatically. IMHO, creativity can be learned, but not over night. Creativity requires comfort with risk and can therefore be costly, which goes against the contemporary programmer's drive toward efficiency and optimization. Creativity requires bring a fresh perspective to each problem encountered, and can therefore be exhausting, which eschews the programmer's drive toward re-use and generalization. Creativity requires leaning on other disciplines for inspiration, which runs counter to the demands of being a language or technology expert. With time, the grip of these convictions can be loosened and contextualized, allowing room to explore curiosities and creative impulses, many of which may be mathematical (or psychological, or involve re-framing), that arise naturally during the course of a project.
- laichzeit0 14y agoI don't understand what the apprehension is towards learning mathematics. Take it at college (university). Take as much as you can. Take it instead of silly courses like philosophy (for which you can sit in a lecture in your spare time, or look at online videos, or take out a book from the library and read at your own leisure). Just learn the freggin' math and stop bitching about "is it necessary, do I really need it, blah blah blah". It's an IMMENSELY useful subject in just about every branch of human knowledge. It's never a waste of time, perhaps even more so if you're a computer programmer.
- edanm 14y agoYou can talk and argue all you want. But the fact is, many many successful systems have been built, and are being built, by programmers who couldn't care less about mathematics. Many of those couldn't care less about programming, either, to be honest. We here who are reading such discussions are the elite of programming, and also the small percentage of programmers who actually care about our craft. Let's not forget that most don't - and it's not a problem, the world still works.
- sunkencity 14y agoI think it's possible to come into programming from lots of other disciplines, one example is Philosophy. Reading code is quite similar to doing argument analysis. I think it's also easy to stare oneself blind on low-level stuff. We need people that have higher skills in the domain that the code should solve that can code. Having coders in one camp and then the "product owners" in another camp leads to solutions that doesn't work for either group.
- Crake 14y ago"Rather, mathematics is a tool for understanding phenomena in the world: the motion of the planets, the patterns in data, the perception of color, or any of a myriad things in the world that might be understood better by manipulating equations." ^---So true.
- peripetylabs 14y agoIt would be nice if every programmer owned a copy of Abramowicz and Stegun, and every mathematician knew at least one programming language.