Normal Humans Effectively Excluded From Developing Software 608
theodp (442580) writes Over at Alarming Development, Jonathan Edwards has an interesting rant entitled Developer Inequality and the Technical Debt Crisis. The heated complaints that the culture of programming unfairly excludes some groups, Edwards feels, is a distraction from a bigger issue with far greater importance to society.
"The bigger injustice," Edwards writes, "is that programming has become an elite: a vocation requiring rare talents, grueling training, and total dedication. The way things are today if you want to be a programmer you had best be someone like me on the autism spectrum who has spent their entire life mastering vast realms of arcane knowledge — and enjoys it. Normal humans are effectively excluded from developing software. The real injustice of developer inequality is that it doesn't have to be this way." Edwards concludes with a call to action, "The web triumphalists love to talk about changing the world. Well if you really want to change the world, empower regular people to build web apps. Disrupt web programming! Who's with me?" Ed Finkler, who worries about his own future as a developer in The Developer's Dystopian Future, seconds that emotion. "I think about how I used to fill my time with coding," Finkler writes. "So much coding. I was willing to dive so deep into a library or framework or technology to learn it. My tolerance for learning curves grows smaller every day. New technologies, once exciting for the sake of newness, now seem like hassles. I'm less and less tolerant of hokey marketing filled with superlatives. I value stability and clarity."
"The bigger injustice," Edwards writes, "is that programming has become an elite: a vocation requiring rare talents, grueling training, and total dedication. The way things are today if you want to be a programmer you had best be someone like me on the autism spectrum who has spent their entire life mastering vast realms of arcane knowledge — and enjoys it. Normal humans are effectively excluded from developing software. The real injustice of developer inequality is that it doesn't have to be this way." Edwards concludes with a call to action, "The web triumphalists love to talk about changing the world. Well if you really want to change the world, empower regular people to build web apps. Disrupt web programming! Who's with me?" Ed Finkler, who worries about his own future as a developer in The Developer's Dystopian Future, seconds that emotion. "I think about how I used to fill my time with coding," Finkler writes. "So much coding. I was willing to dive so deep into a library or framework or technology to learn it. My tolerance for learning curves grows smaller every day. New technologies, once exciting for the sake of newness, now seem like hassles. I'm less and less tolerant of hokey marketing filled with superlatives. I value stability and clarity."
Cry Me A River (Score:5, Insightful)
Normal humans are excluded from a lot of things.
1. Olympic Gold Medal
2. 5x Jeopardy Champion
3. Professional Concert Pianist
4. Bolshoi Ballet
5. Supermodel
etc.
The idea is to find your niche in life and exploit it. Not call the whaaambulance.
Apparently dedication = autism (Score:5, Insightful)
just like EE, ME, CE, finance...... (Score:5, Insightful)
just about every field above burger flipping requires specialization. Are you going to ask Joe Blow about your corporate tax accounting needs? Or are you just going to drop by Intel and see if you can lend a hand with some microcode? Work is becoming increasingly specialized across all fronts, time to get used to it.
Or Some People are Finally Employable (Score:5, Insightful)
Well duh (Score:5, Insightful)
There's a good reason I'm not a brain surgeon or a professional hockey player. Have you ever tried explaining even a 'hello world' example to someone who can't handle strict logic and math?
normal people (Score:4, Insightful)
Have this guy seen "normal people" use a computer? There are some people so uninterested in the thing (even when is their primary work tool) that they can't be bothered to learn so simple stuff as mouse dragging or keyboard shortcuts.
Hell, I've seen people using Spreadsheet software for 10 years without learning how to use formulas. Don't even try to show them what all that HTML gibberish is.
And Spreadsheet software is a pretty good introduction tool for programming.
Re:Cry Me A River (Score:5, Insightful)
They want to be able to be a programming superstar by reading a book such as:
* Learn Programming in 24 Hours!
* Learn Brain Surgery in 24 Hours!
* Learn Rocket Science in 24 Hours!
* Learn To Be A Concert Pianist in 10 EASY Lessons!
Various programming boards are flooded with people who want to know how to break into programming for big bucks, quick, overnight, but don't want to actually do the hard learning [norvig.com].
Crazy rant full of BS except for one thing (Score:5, Insightful)
I will mod that one way up.
Maybe because normal humans can't code (Score:5, Insightful)
Seriously, does everyone think programming is a spatial relationships problem or something?
Let's put this on the table right now: Normal humans can build houses. Oh, you might not have any construction knowledge, and you'll build a horrendous little shitheap that falls over when the wind blows, but that's not the point. I can put construction knowledge in your head and, in a few months, you'll be able to properly select foundation for a site, properly frame a house, and properly build out the sheathing and siding and insulation and walls. You won't be a master craftsman, but you'll be able to do it right.
Humans are good with spatial things. Humans can look at a two-by-four and understand what a two-by-four is. The engineering concepts behind building a workable shed are a little different, but easily transferred. Given a little time and guidance, a human can learn to relate building materials spatially, measuring and cutting and nailing or screwing or gluing as needed, planning and building a proper structure.
Humans are terrible at numbers and algorithms.
Humans are so terrible at numbers and algorithms that they become *extremely* proficient at math if you teach them with a soroban--a machine that converts numerical problems into spatial procedures--and can't be taught algorithms without visual diagrams of trees and boxes and other shit to show sorting and transformation algorithms. Have you ever looked at textbooks or Wikipedia pages for stuff like PKI, red-black trees, or AES encryption? There's pictures of the simplest shit! Why? Because HUMANS CAN'T PROCESS ALGORITHMS!
The easiest process for a human programmer implementing an algorithm like a quick sort is to associate variables with objects in the visual diagram, associate their state changes with the movements in the visual diagram, and write code that carries out the analogous behavior. By comparison, BUBBLE SORT IS FUCKING HARD TO IMPLEMENT when your only guidance is: "iterate through each list element. Compare each element to the previous. If the previous element is larger, swap them." You actually have to think about how to do the comparison (greater than, less than? Wait, which am I comparing to which?), and how to swap them--usually with a temporary variable, although "A ^= B; B ^= A; A ^= B;" works. Most people visualize some kind of diagram while trying to understand the algorithm.
The real world requires interaction with space, mainly to avoid hungry tigers, kill tasty deer, and avoid driving your car into trees like you're fucking drunk. It doesn't involve shift accumulator left and XOR with memory at address $FC. It doesn't involve explicit semaphore locking and deadlocks if you fail to unlock the semaphore in a loop with multiple function calls and thread branching during the loop. It requires things you can put your fist through if they don't work right, and then continue with successfully.
We can't all be rocket surgeons.
Normal humans exlcuded from practicing law/medicin (Score:5, Insightful)
"The bigger injustice," Edwards writes, "is that being a doctor or lawyer has become an elite: a vocation requiring rare talents, grueling training, and total dedication. The way things are today if you want to be a lawyer or doctor you had best be someone like me on the autism spectrum who has spent their entire life mastering vast realms of arcane knowledge â" and enjoys it. Normal humans are effectively excluded from performing surgery or arguing cases before a judge. The real injustice of legal or medical inequality is that it doesn't have to be this way." Edwards concludes with a call to action, "The web triumphalists love to talk about changing the world. Well if you really want to change the world, empower regular people to perform open heart surgery and argue cases before the supreme court. Disrupt specialist knowledge and training! Who's with me?"
Re:Apparently dedication = autism (Score:5, Insightful)
Generally when people say autistic, they don't mean a mildly afflicted, high functioning person, but someone who never speaks, rocks in a corner, and screams if their normal routine is changed. You DON'T have to be autistic, or be anywhere on the autistic spectrum to be a great programmer.
Like becoming good, even excellent, at anything it requires hard work, dedication, and practice. Any normal person can do it. Programming, and I've been doing it since an assembler was a real cool tool, can be mastered by normal people. Sure, I've seen a few odd balls in the field, but no more so than in other fields.
As far as making programming easy for the masses: that is fine for little toy systems, but if you want a large system built, you want properly trained professionals working on it.
Re:Apparently dedication = autism (Score:5, Insightful)
Because it has become a meaningless buzzword used to describe every introverted snowflake on the planet?
The GP responded more-or-less appropriately to the TFA's nonsense. You have simply said "nuh-uh!". Substantiate, please.
Today, I would never have learned programming (Score:5, Insightful)
I got my first computer in 1986; I was 13, and it was a ZX Spectrum with a build-in BASIC interpreter. When you switched on, you could start away programming. In fact, the computer came with a little book with programming examples and little games. I spend countless hours typing in listings that I found in newspapers. To even load a simple game you had to enter a command.
Since then, I learned C, tcsh, C++, bash, Perl, much later also Python and R. It was a step by step process, and I would never have started it (and became what I am now, that is, computational biologist) if not for this one computer with the BASIC interpreter.
I have kids now, and they have Android tablets. The sheer power, their parameters and their capabilities are overwhelming. I don't know how many instances of a ZX Spectrum emulator I could run on one of these, a thousand?
But even though they run on a system that is related to the system I am using every day, I would not know how to write a program for them to save my life. In theory, I know how I would approach it, I even set up once an Eclipse environment once, but I never got to even start a Hello world program. If I were 13, I would not even know that I can write a program myself.
It is amazing, but I think that actually, my kids will have a much harder time to learn programming than I had, and they will get much less fun in return...
Re:Cry Me A River (Score:3, Insightful)
I have seen plenty of these "tools" and they are worthless for anything complex. If you need to put a nail in, I can give you a hammer to make your job easier, but what happens when you need to put the nail someplace the hammer doesn't fit?
Professional athletes and "unfair advantage" (Score:5, Insightful)
Some people in life find an "unfair advantage". This is very evident in professional athletes. They must start with natural athletic ability and then hone that through practice and training. And then a select few get paid huge dollars to essentially play a game.
People with natural problem solving and logic skills also have an "unfair advantage". It doesn't generate the quick wealth of the professional athlete but can lead to a promising professional career path. It still takes practice and learning to really take advantage of these skills much like the professional athlete learning their sport.
I will not apologize for taking advantage of my abilities any more than a professional athlete will give back the money they earned playing a game.
Re:Normal? (Score:5, Insightful)
My tolerance for learning curves grows smaller every day. New technologies, once exciting for the sake of newness, now seem like hassles. I'm less and less tolerant of hokey marketing filled with superlatives. I value stability and clarity.
Yeah that's called growing up.
"Rare talents"?! (Score:4, Insightful)
Does this egotistical idiot actually believe that?
Programming is not something that requires grueling training or rare talents. Algebraic topology, cardiothoracic surgery, and competitive chess require those. If you're writing code that requires elite skills, you're doing it wrong - no one is going to be able to understand it, and you will never be able to troubleshoot it. Someone with an IQ of 100 can become a perfectly competent Java or C++ programmer with two years of intensive training. Programming is more akin to a trade skill like plumbing or electrical work, than it is to engineering. And before everyone gets on my case that being a top 1% programmer is incredibly difficult, the same holds for a top 1% electrician.
Brings a tear to this sysadmin's eye... (Score:2, Insightful)
"I'm less and less tolerant of hokey marketing filled with superlatives. I value stability and clarity."
That's what we've been telling the Dev side for decades now. .
Re:Cry Me A River (Score:5, Insightful)
I know, right? We had it so much easier back when we could just write our own interrupt handler (and pray we didn't step on DRAM refresh or vice-versa) to pull bytes directly off the 8250 - And once we had those bytes, mwa-hahaha! We could write our own TCP stack and get the actual data the sender intended, and then do... something... with it that fit on a 40x25 monochrome text screen (yeah, I started late in the game, those bastards working with punchcards spoiled all the really easy stuff for me!).
And now look where we've gone: Anyone using just about any major platform today can fire up a text editor and write a complete moderately sophisticated web app in under an hour. Those poor, poor bastards. I don't know how I can sleep at night, knowing what my brethren have done to the poor wannabe-coders of today. Say, do I hear violins?
Re:Cry Me A River (Score:5, Insightful)
No, he has a point. Back in the day, we had few tools and we learned how to use them.
now, we have a tool for every hour of the week, and as soon as you've mastered one, someone comes along and says "your skills are sooo obsolete, you must learn now or fall behind", so you get to grips with it and start top master it, and then realise its a pile of poop and hunt around for a new, cooler tech to use instead.
Software projects today are littered with the corpses of technology that was the silver bullet to make your life as a developer so much better, easier and productive. Constantly.
That's the problem - we're not productive, we spend all our time learning new crap that is little better than the ancient stuff we used to use and got stuff done with.
The tools, well I know people who swear vim is easier to use than the latest IDE that has full intellisense and refactoring builtin, and they are probably right - in that they have learned their craft using that tool and actually are more productive than the bloated and slow IDE could make them. The trouble is that newbies start with the IDE and don't know anything else, so they stay in the "its easy" camp and never progress to real masters of their art. Which is understandable when you need to re-skill every couple of years, but not beneficial to the software industry.
Re:Cry Me A River (Score:5, Insightful)
But the real problem is this impression that you have to be born 80% as smart as Einstein to get into this field, and that the learning curve is impossible for regular people. That's totally wrong. Average intelligence plus persistence is all you need. You won't be Linus Torvalds tomorrow, you won't be Steve Wozniak next month. But put your time in, try things out, get used to being frustrated as you learn and keep learning anyway, and in a few years you'll understand what's going on and be able to do anything this side of the most advanced work as well as anyone.
That's the lesson we the progressives should be teaching people. And to be clear, it fits all of my original examples too. Few people walk into an automotive engineering program and instantly grasp all of the concepts involved - years of persistence matter more than raw talent if you want to design engines. Few people start building medical equipment and have an instant knack for getting it right - years of persistence matter more than raw talent again. If you were born with an 80 IQ, sorry there's only so far you can go. But the difference between a person with 110 IQ that contributes code to the Linux kernel and one that works at a gas station is their persistence, not raw intellect.
um (Score:3, Insightful)
Hate to break it to you but the Autistic ARE normal humans. It's just another way of thinking that's slightly different that what a lot of people are used to.
Also, where the hell are you working? I know plenty of programmers that are just as stupid as everyone else. Vast reams of arcane knowledge? Are you using the Forgotten realms addition of C#?
Re:Crazy rant full of BS except for one thing (Score:5, Insightful)
This phenomenon is otherwise known as "getting old".
Re:Cry Me A River (Score:5, Insightful)
The problem is that "accessible" usually also means "nerfed" and "limited in scope" and "very annoying to the power user".
Re:Cry Me A River (Score:5, Insightful)
By contrast, a hammer, ranging between $5 at Harbor Freight Tools to $80 for a top-of-the-line deluxe framing hammer forged from olympus steel and quenched in the tears of angels will drive in almost any conventional nail that one needs, and unless abused will probably last as long as the owner will.
I'm working with some web software at the moment. It's the kludgiest amalgomation of crap that I've seen in quite some time. It's got OS library dependencies, but they need to be newer than one stable distribution's version, but older than another stable distribution's version, so one has to use unsigned third-party repositories for those. Then for Ruby on Rails and for Node.js it needs two other sets of proprietary repositories, and it needs specific versions of packages from those repositories too, not default, and it installs some redundant packages that were already covered by OS in slightly different version. Then once you go to put it in it requires MySQL for some of the dependencies but the main program itself only runs on PostgreSQL, so you're stuck with two DBs running, one doing almost nothing but required to be there.
This is sickening. This will make it almost impossible to do OS updates, and will cause all manner of problems if those third-party repositories ever go away, or if the developers for them stop maintaining those specific versions. It's dangerous and stupid to do this.
Re:Cry Me A River (Score:5, Insightful)
No, he has a point. Back in the day, we had few tools and we learned how to use them.
now, we have a tool for every hour of the week, and as soon as you've mastered one, someone comes along and says "your skills are sooo obsolete, you must learn now or fall behind", so you get to grips with it and start top master it, and then realise its a pile of poop and hunt around for a new, cooler tech to use instead.
Apologies, but we still have all those old tools. We just don't use them any longer. Because you can't use Turbo Pascal to make web pages, but you can use jQuery. If you were working on the same problems today as you were working on 20 years ago, you probably would be using many of the same tools. The only reason you're using the new tools is because you'd rather spend 20 hours throwing something together versus 20 weeks writing it from scratch.
Honestly, if you think this is different than it was in the 90s and 80s, then you weren't paying attention in the 90s and 80s. The technical periodicals were FULL of the new stuff that was going to change everything. The only real difference is that it's easier to find stuff and get distracted these days, simply because the industry is much larger. I assume it was similar as you go back further, I just am not old enough to remember it first hand.
Re:Cry Me A River (Score:4, Insightful)
And when I see words like "injustice" and "excluded" I see a typical liberal who views skill, talent, dedication and mastery as bad things. Misapplying words like this, in ways exactly like this, cheapens real injustice, and real exclusion. Normal Humans are excluded from exceptionalism, not because of some "injustice" but rather that is what makes the exceptional so great (mostly hard work and dedication).
How about, instead of deriding the exceptional among us, we inspire others to exceptionalism?
Re:Cry Me A River (Score:5, Insightful)
What these utopian visions of programmatic democracy all lack is any notion that attacks the essential complexity of the problems being solved by code. Problems that have, if anything, grown more complex with increasing memory and CPU power. All the forays into "graphical programming" or other tools to take the programminess out of programming have shown that it doesn't matter whether you're expressing a solution in text or little icons connected by arrows - the essential complexity of the problem remains. The only way we're going to democratize programming is if AI gets to the point where the thoughtwork of breaking down the essential complexity of problems can be offloaded to some other intelligence.
Re:Cry Me A River (Score:2, Insightful)
Most of what was written in COBOL looked very much like any non-AJAX web app. Sure, there were the batch programs you describe, but mostly it was apps for terminals. Send a form, the terminal posts a reply, hit the database, send the result fields. Same-old same-old. But it was all server-side code.
There are a few WYSIWYG web editors out there, but for some reason they were never that popular. People seem to want to muck around with JavaScript and frameworks and otherwise dick around with the client side code, as if the bit that paints the screen were the important bit. That was the difference in the COBOL years - you wrote the server side and let the client take care of itself, instead of trying to do that backwards.
Re:Cry Me A River (Score:5, Insightful)
But the real problem is this impression that you have to be born 80% as smart as Einstein to get into this field, and that the learning curve is impossible for regular people. That's totally wrong. Average intelligence plus persistence is all you need.
What you really need is to deal with this anti-intellectualism that's so popular in the culture today, and replace it with genuine curiosity, a joy of discovery, and a delight at learning new things.
Do that, and the rest will naturally follow, and not just in software development.
Re:Cry Me A River (Score:4, Insightful)
No, it isn't.
Tools are simpler and easier to use than ever, and this guy is mistaking nostalgia and innocence for actual difference.
When websites like TopCoder are offering $100-200 bounties for something as simple as changing how a table is sorted it really shows how it's become easier than ever. It's just layer after layer of needlessly complex code and we don't realize how inefficient, poorly designed/coded, and horrible it really is because the speed of the hardware masks it.
Working with a framework and some spreadsheet code things just didn't "feel" right to me, felt sluggish. No one else noticed it, they said it was fine, I was imagining things. It took me almost a full day of digging to discover that a single 37 character line of code was slowing everything down - no one bothered to do any efficiency testing on it because it was a "low level function". Changing it sped up the application a thousand fold. Had the code been well designed and not set on mounds of anonymous functions, hacks, and bad practices it probably would have taken about 20 minutes to locate and fix.
Just because something can accomplish the tasks you need it to simply/easily doesn't mean it's well designed, simple, or easy from a coding perspective.
Re:Crazy rant full of BS except for one thing (Score:5, Insightful)
Yeah, I was going to say that. He's just facing burnout and age.
But "New technologies, once exciting for the sake of newness"? That phenomenon is known as being young and stupid. The new technologies are exciting because of the additional capabilities they give us. If you thought the technology was exciting just because it was new, then you've been misguided the entire time, and marketers must have loved you. They could slap "NEW" onto an old product and generate some more sales. A fresh coat of paint and it's a top seller again.
Ignore the paint. Cut through the bullshit. Does the new thing work better? If so, it's worth learning.
Or hey, you can stick to the stable and clear COBOL platform that you know so well. Since all your peers are dying off you can charge an arm and a leg for being a master at it. Hopefully you didn't gamble your decades on something like RPGII.
Re:Cry Me A River (Score:5, Insightful)
I wonder if the writer has ever seen the monstrosities programmed in BASIC/VB, COBOL or HyperCard by the resident business manager. People in general have no clue about programming or mathematics. People in general, don't go for higher education. People in general have an IQ of about 100. People in general can't work with a computer when the outline of things changes or the buttons move around. And you want those people to program a math equation that requires 2 years of college math... and they need to place the buttons themselves?
Hell, take things "programmed" in Excel for that matter. I've seen people use 3 columns to do things which could've been written in 1 operation especially when it comes to adding percentages to a value (they'll calculate 4%, then add it's outcome to the source value to get a +4% and then hide the other 2 columns instead of just doing 104%). That will take them 2 hours to complete.
The Web is fine. Plenty of people understand HTML, even without much education. People UNDERSTAND that things within a document need to be described at some point. Plenty of people can even understand basic JavaScript, even without much education.
The reason the web and most of programming in general is so kludgy and broken in many places is because we've let those people that understand HTML and basic JavaScript make websites and entire applications. We have told business managers that they can describe their business in a common and easily understood language and the business manager did describe their business but then they've gotten in way over their head where they themselves can't even understand what they've done. And then those business managers moved on and started claiming they had programming experience and then they went to another company to make ever bigger monstrosities. And REAL programmers get a bad name because programming these days is so easy, anyone can do it.
Re:You Insensitive Clod (Score:3, Insightful)
3D printers are a fundamental game-changing technology. What you're seeing now is basically proof-of-concept stuff, but the technology is developing rapidly and it won't be long before pretty much anything can be built from scratch on demand. 3D printing and associated CNC mills have the potential to revive American manufacturing in a localized, small-batch paradigm rather than wasteful mass production and national/internation distribution chains.
3D printing is as disruptive a technology as the printing press -- that's no hyperbole, I really mean it. If you can't see it, I pity your lack of vision.
Re:Cry Me A River (Score:4, Insightful)
However, if the Dewalt Model XJ-9 nailgun lasts 5yrs you can finish a helluva lot more roofs in that time than you could with a hammer. Perhaps then we should look at Angular, PhoneGap, nodeJS as specific models of nailguns from which we should extract as much 'juice' as we can in the 2-5yrs they might be useful and presume that we'll be using something else after that.
Unfortunately, the roof/nailgun analogy completely falls apart when you realize that if some of the shingles fall off after the XJ-9 has been discontinued you can still use a regular hammer to fix it; whereas if Angular 3 is EOL'd in 2017 then your PhoneGap app built on it might be left with some vulnerability (all geolocation requests are hacked to only report your current location as the nearest strip club) that Google is not going to fix (having sold off their money-losing software biz in 2016 to focus on crowd pacification robots).
And perhaps, instead of waking up every day wondering if today is the day the Yosemite super volcano [goo.gl] or a planet killer comet wipes us all out, we should just dance (and code) while the sun shines and not worry so much about the future.