The Extinction of the Programming Species

Max Goff writes "Given the recent chatter surrounding the extinction of the U.S. programmer, /. readers might also be interested in a series of articles I recently penned for java.net -- the Blacksmith and the Bookkeeper (part 1, part 2 and part 3) -- in which I posit that the postmodern programmer (the entire sub-species, not just those domiciled in the U.S.) shares much with the blacksmith of old, and will become just as extinct in relatively short order. It is not due to work visas or outsourcing, but has much more to do with the evolution of work itself."

Comments

Blacksmith?

[nuclear305]

Ok...without RTFA...since the blacksmith was generally replaced by industrial production (either directly or indirectly) ...does that mean programmers are going to be replaced by machines that can code on their own?

If that's the direction TFA is taking...I highly doubt it. To quote the Borg Queen from ST: First Contact: "You are an imperfect being created by an imperfect being."

Unless humans as a whole somehow reach perfection, we'll never be able to produce something to completely replace the human element.

Re:Blacksmith?

[LiquidCoooled]

No, but we can mechanise the process.

A factory machine is not autonomous, it doesnt just do everything itself, it must still be told what to do.

I see development work adapting and expanding, and yes I see fewer actually code monkey jobs.

In a computerised brassband, there is still a need for a skilled composer.

Re:Blacksmith?

[smallpaul]

Why not RTFA?

Re:Blacksmith?

[drix]

Isnt this pretty obvious? Java and Python are two "machines" that allow me to forget about areas of programming that people had turned into veritable art forms just a decade ago, things like memory management and code portability. People who are still doing that kind of work are the technological equivalent of blacksmiths--still needed for some niche roles (kernel hacking, scientific computing), but by and large a relic. The same may be said for the ever-increasing (quality and quantity) assortment of code libraries available today, and for code reuse in general.

I don't think this is a bad thing; I don't agree with the chicken little tone of this story. If we have a lot of blacksmiths--e.g. high skilled, labor-intensive jobs--dominating any given sector, they should be disappearing at a healthy clip. The US (or any country, for that matter) is left better off as a result. That's just plain and simple Econ. 101: technological advancement drives long term growth and prosperity. Programmers (esp. on Slashdot) are quick to lament the de-commoditization of their profession, because it means lower wages and more competition than that to which they became accustomed in the nineties. But there's no denying that in an overall, macro sense, having a nation full of Java programmers will result in much greater overall well-being than one of 10,000 C haxors and 250 million farmers.

Re:Blacksmith?

[azyuroth]

I can't honestly see this happening any time soon. My main problem with it is games... I wouldn't want to be having to spend X dollars on hardware to run a game that would be able to run on my current machine, were it written in a faster language.

I suppose this has already happened to an extent, with C instead of assembly language. Maybe writing fast code is an obsolete technique...

I don't understand your remark that a nation full of Java programmers results in better well being... Could you explain it?

It's all a fad

[tomstdenis]

Tools come and go. Knowledgeable people are here to stay. The problem is many people pretend to have knowledge in fields they're not experts.

It's the same with diet/exercise. As the years go on new diets/plans come and go yet the old school traditional "just eat less crap, more good and get off your ass" philosophy is still here.

Wanna make sure you're employable as a software developer? Make sure you're actually knowledgeable about the science *and* the field. It's good to know about things like Java/C++/etc but it's equally important to know about design theory, algorithms, math, etc.

So if you signed up for some 6 month "computer science" program well don't blame the industry for your unemployable status. Code monkeys are a dime-a-dozen.

Tom

Re:It's all a fad

[cubicledrone]

Knowledgeable people are here to stay.

Well, at least until the next round of layoffs.

Wanna make sure you're employable as a software developer? Make sure you're actually knowledgeable about the science *and* the field.

You know, the animation industry is a good example of why this doesn't work. Disney laid off hundreds upon hundreds of totally irreplaceable feature animators over the past few years, most recently in Florida. These people were almost directly responsible for hundreds of millions of dollars in revenue to the company. But, they're too expensive.

So to fuck all with their skill, experience and DIRECT INFLUENCE ON TOP-LINE REVENUE, out on their ass they went with their pencils and ink and an EIGHT-DECADE tradition of animation quality right after them.

Companies don't give a fuck about skill or education. It's too expensive. They would much rather sell shit at a 50,000% markup and congratulate themselves at company parties.

Re:It's all a fad

[frenetic3]

Uh, I would imagine many of these people retrained as computer animators, which Disney, Pixar et al still need. The *tools* are different, but the core skills and concepts are the same whether it's ink and paper or Maya.

And if not, hey, they could draw kickass invitations for those company parties. With bunnies and gold stars and shit. It's a win win, dude.

-fren

Re:It's all a fad

[ericspinder]

Sure, knowing what's going on allows you to learn new stuff easily (look at all the C++ programmers who can code up Java once they have a 10 minute tutorial)

I've seen this, and it ain't pretty. Hell, even most of the old C++ programmers will acknowledge their early Java code tries as misguided. What helps an experienced programmer the most is his (or her) 'general' knowledge of the task of programming itself. Moving from C++ to Java (IMHO) is something like moving from a motor cycle to a car. Sure the car's 4 wheels help balance (ie trash collection), but car can't fit though your old short cuts.

One day, I think business apps will be programmed using tools something like Visio - drag and drop your 'web service' 'component' match up the inputs and outputs, and connect a GUI at one end (that provides inputs) and a DB 'component' at the other (to provide outputs).

Bill Gates!, glad to see your here! AKA, they have been predicting this exact situation for the past couple of years. As one who has coded and worked with several web services, Bull Shit. It's not going to happen anytime soon, if ever.

Something like this would be suitable for programming 99% of business applications and would be useable by 'business people', not dedicated programmers.

Then those business people will say that they program. look on some resumes, you will eventually find someone who can 'program' MS Excel.

As the tools improve (ie become more automated) you have to keep up with them, or become as anachronistic as the blacksmith mentioned.

Sure you don't see too many black smiths around these days, but you do see...

• Metallurgical Tech./Technician
• Machinist
• Machine shop workers
• Steel workers
• locksmiths
• and many more...

Just becuase they don't call themselves Black Smiths, doesn't mean that can't trace back to some guy standing over an avil, pounding away with a hammer.

Re:It's all a fad

[Safety Cap]

They did everything right, ~.

No they didn't. They forgot one simple fact:

2. Consider yourself a long-term contractor, not an employee. The company will discard you when it becomes expedient. You should be prepared to do the same.
-- Ripples

Re:It's all a fad

[tolan-b]

Well yeah I basically agree with what you're saying, but perhaps I didn't make myself clear enough.

Yes the wider your skillbase the more adaptable you're going to be, unless of course you spread yourself too thin.

My point about higher level languages wasn't that they're more maintainable by default, far from it. My point was that people using them tend to have a different set of priorities in the working lives.

To go back to schedulers for example, when you write a scheduler, I imagine one of the key aims is for it to be as fast and lightweight as possible, and to that end, you're going to write it to do one job very efficiently. The chances of your boss turning round and saying "yeah, now we want it to integrate with our new CRM system" are probably quite low...

However if you're writing business applications then there is much more emphasis on changing needs. Everything you do has to be with the assumption that at some point, probably sooner rather than later, someone is going to turn round and say "actually, you know that system you wrote? We wan't it to do something it was never designed to do. By wednesday please."

So cards on the table. Yes I get a bit defensive when people start telling me I'm a coding lightweight simply because I can't reel off the major sorting algorithms off the top of my head. Yeah, I've been over them, I think I wrote a bubblesort once for some reason when I was about 15, but you know what? In 8 years writing business applications I've never needed to write a complex sort. It's just not the sort of thing that's important.

What I *do* need to understand includes things like good OO design, so that changes to systems my code interfaces with won't require re-writes of more than a small amount of my code. Understanding where the typical bottlenecks in the types of applications I work on tend to be (the database 9 times out of 10). Again with databases, how to efficiently abstract database logic from the main application logic so that changes in schema for other systems don't require a re-write of the one I'm working on.

None of these things are rocket science, but they're my domain, and in my experience, lower level coders who move into this area often don't really understand these things and I end up inheriting a codebase that's 10% more efficient, but needs major re-architecting for relatively minor changes. That's fine, they're from a different discipline, but not necessarily a 'better' one.

Oh dear that turned into a rant didn't it? :)

Re:It's all a fad

[tomstdenis]

I.T. != software development

Some clarification....

Software Developer - Responsible for design, testing and implementation of software.

Software Programmer - Code Monkey.

I.T. - Someone who manages information [a/k/a SysAdmin and SysOp].

Programmers [while talented in their own right] are often mislabeled as developers when they're not. If you don't design, test and verify software you're not a developer.

Though I agree abut the resume issue. That's why you do your own projects/OSS to stuff your C.V. with accomplishments.

Sadly though even that's overlooked. RiM [those blackberry fruitcakes...] for instance don't consider projects/OSS as "experience" and *only* consider education in the interview.

Tom

Re:It's all a fad

[Anonymous Coward]

Making horseshoes requires at least 2 distinct phases. Design, and implementation. Neither is going away, and in the case of the blacksmith it was only the implementation that was automated. Despite the author's innuendo machines do not do design. Essentially the manual labor market shrank, while the design side exploded (think engineer).

While computers are a similar issue (we see fewer and fewer mainstream programmers needing to know assembly) that does not mean that a machine will ever be making up the business rules, just as an accountant is needed, so is a "programmer". Perhaps his/her technical skills will change over time, the act of giving a machine instructions to carry out a task cannot and will not go away.

Ok

[cubicledrone]

It is not due to work visas or outsourcing, but has much more to do with the evolution of work itself

Nonsense. It has to do with the wholesale re-negotiation of the social contract between business and the society it sells to.

Business is no longer satisfied with the simple model of building a product and selling it. That would require work, investment, long-term planning and respect for the skill, education and loyalty of their employees.

Much better, business says, to sell all of that and simply manufacture brands. The modern workplace is therefore obsolete, because business does not want to pay for it. It's really that simple. People are only useful if they are opening their wallets at a cash register. They are not worthy to be employed and paid a fair wage.

There are voluminous statistics to support this, but a few should be sufficient. In the last few years, corporate assets have increased 288%. Employee compensation has increased 9%.

Over half of the working-age population is either a temp, part-time, unemployed or out of the workforce entirely. Read it again: HALF OF THE WORKING-AGE POPULATION IS NOT EMPLOYED FULL-TIME. Companies have no respect for anything: skills, education, experience are all totally meaningless to these companies. ALL that matters is money. That's why it takes five interviews and a credit-check to get hired, but only an e-mail to be fired.

Unless, of course, we're discussing the price tags of all these "brands." Then, it's "you get what you pay for." Like $2500 for a mattress and $4500 for a television, $175 for a basket of groceries and $50,000 for a car. It's nonsense, of course, but everyone's too busy arguing about the problem. Nobody is interested in hearing the facts. All that matters is money.

Re:Ok

[BrainInAJar]

Medieval serfs didn't work for an hourly wage, modern employees do.

The problem isn't a lack of employment, The problem is insufficient compensation to keep you in a human standard of living

Increased productivity is good for keeping business profits up, but it's no good at all for keeping people fed and clothed

Re:Ok

[Hard_Code]

To add on to that: there is a gradient between developed and developing nations. Harnessing that gradient is profitable Cheap labor produces cheap goods which further fuels the migration of labor jobs to developing nations. The laborer in the developed nation then requires more cheap goods because wages are depressed, and the cycle continues. The playing field needs to be evened, but it needs to be evened on OUR TERMS, which means fair trade not "free" trade. Laborers in developed nations will almost certainly have to take pay cuts, but not the rampant and excessive cuts, or cuts in jobs, that is resulting from wholesale migration to sweatshop and slave labor. As Alan Keyes put it in a recent Senate debate, the American worker produces a "free" good while the sweatshop worker produces a "slave" good. It is not fair to pit the Free good against the Slave good in the market.

Re:Ok

[Anonymous Coward]

Increased productivity is good for keeping business profits up, but it's no good at all for keeping people fed and clothed

That's an artefact of the system. In the end the society needs X amount of goods and X amount of goods is being done with less and less work.
The only problem is distributing them to everyone in a fair way when the system is built on the assumption that it's realistic to expect all "good people" who deserve the goods working full time.

Re:Ok

[tehdaemon]

You have never tried to take care of kids, cook, do laundry, and keep the house clean. It is full time work.

Re:Ok

[pjt33]

I don't deny it, but it's not full-term employment in the sense in which the phrase was used in the post to which I was responding, and it doesn't count towards statistics on the workforce, which is the word I used.

Postmodern(Apple) != Postmodern(Orange)

[Anonymous Coward]

The problem lies in the definitions. A bookkeeper of today is nothing like a bookkeeper of the 19th century. The "postmodern" bookkeeper is dependent on a new set of tools -- software written by those outmoded programmers, even. They no longer calculate books by hand, but use automated tools, and they're not so concerned with tracking business' money as they are with audits, GAAP, and finding interesting new ways to conjure a penny more than consensus earnings estimates from the Street.

Similarly, the role of "blacksmiths" evolved. Today, you might call them machinists, or perhaps they're the assembly-line workers. You don't see many blacksmiths making auto parts by hand, but then you don't see many bookkeepers calculating General Electric's quarterly results by hand, either.

It's merely a linguistic quirk that we retained the occupational title for "descendant job 1" and not "descendant job 2". Recall that "computers" themselves were once human beings that did arithmetic.

Similarly, programmers aren't going away, though we may stop calling them "programmers" and the day-to-day details of their methods and tools will change.

And if you are concerned simply with a narrow focus on specific methods of today, then programmers have already survived the postmodern shift. As scarce as blacksmiths are professional programmers who key in word values with switches on the front on their machine, and almost as scarce as those that deal with any significant amount of assembly language. Programmers and their tools have already evolved with time, and a modern programmer mucking about in .NET RAD would be nearly as alien to a systems programmer of forty years ago as a machinist with CAD-driven power machine tools would be to the blacksmith of four hundred years ago.

No Knuth?

[Eric Giguere]

The reader comments on these articles are themselves interesting. I especially think this one is telling: "As a Java programmer I haven't looked up searching and sorting algorithms in Knuth". Being a Java programmer myself, I can certainly appreciate having these kinds of algorithms in a standard library. I still think you should spend some time with Knuth or some other book (or online documentation) and understand how they work. That's one problem I see, not enough programmers today learning the fundamentals and just letting the libraries do all the hard lifting for them.

Eric
Thanks to All-Bran, William Shatner 'goes like no man has gone before'

Re:No Knuth?

[bob65]

So saying programmers will become extinct really depends on your definition of programmer. Some consider Java programmers that don't know how to implement basic sorting algorithms or do basic O analysis "programmers". Others outside of the tech field think "computer scientists==programmers" (don't get me started on this). *Some* group of programming-related people is going to become extinct, we just aren't sure who yet. Blacksmiths were replaced by machines, so a good place to start would be, "what type of programming can be replaced by machines?" As design patterns become more and more mature, and the software engineering industry becomes more standardized, I suspect we'll reach a point where the class diagram (or some description of the implementation) will be enough to generate working code. We've already seen a trend in programming going from lower level programming (assembly, etc) to higher level programming (Java, etc). Abstraction permeates computing in every sense, and I expect the abstraction to continue.

Dam'it Jim! I'm an Engineer.

[Anonymous Coward]

"So my work experience and part of my academic background was all for nothing this says? Great news! Now that I've studied for more than 4 years, I learn it is going to be useless."

Utter nonsense. Programming and programmers aren't dead until AI matures. In which case programming will not be the only thing to suffer.

You're an engineer, that's the difference between the architect and the guys who actually build the house.

Think big picture view, and while you're doing that, get some OSS experience running a project.

Use tools that give you that god's eye approach to problems, while others actually dig the trenches.

Re:great news!

[smallpaul]

They have been predicting the demise of programmers since the invention of COBOL in the 60s. It was supposed to turn ordinary business users into programmers thanks to its easy, English-like syntax. We're still waiting. Now this writer is talking about "codeless development environments" which are (like the 5GLs and expert systems of yesteryear) supposed to replace programmers.

The analogy is not right

[vmaxxxed]

While I think that programmers are "evolving", the
analogy with "Blacksmith and the
Bookkeeper" is not right.

Bookkeepers, blacksmiths, even secretaries are
disappearing because their jobs can be automated and computers are taking over.
Programmers are disappearing because of a different reason. It is not because computers are
taking over us; to this day no computer can write a program or fix a bug.

I believe that the problem is that the software is
becoming so complex that what could have been done by one-person years ago, takes a team
these days. I remember that in 1991 I wrote several systems like inventory managers, stuff
like that. Today you require testers, DBA expert, GUI designers...

I think it is more like what happened to the
aviation industry. In the 1900 you could build your own airplane, there were people building
airplanes all over the world. Today, only a few companies can do that,
and the demand for airplane-building skills is gone. Well, not gone,
its just that it has changed, now you have to specialize in some area, you can not know ALL what
is required to build an airplane. ( well, unless you are Burt Rutan)

I think that software is evolving the same.
Eventually we will have to specialize in GUI, or DB or whatever. The generic programmer is not
disappearing, just evolving. ( Read with Hannibal Lecters accent. )

My 2 cents...
-Ale

Again, myths and rumors

[Headius]

Posting my comment from the other "extinction" article here...

These judgement day scenarios are based on a big fallacy I haven't yet seen addressed:

The market for software developers is not standing still; it's growing tremendously. We're just not seeing it because a lot of new development is going overseas. However, there's no sign that the demand is going to slow down, and there's not an infinite number of tech workers overseas.

Already Indian workers are concerned about having their own tech bubble, as other countries start coming online with cheaper workers. China, Phillipines, and others are starting to take work away from India.

Further, despite claims to the contrary, it's not just as easy to move programming jobs overseas as it is for manufacturing jobs. Indian programmers aren't just plucked from the trees...they've gone through years of training and education just like we have. It costs a lot more time and money to train a programmer than to train an assembly-line worker. Again, there are not infinite resources available. It just seems that way because India has been building up a highly-trained workforce for a long time--without work to give them.

Our own tech boom and bust resulted in scads of untrained, unskilled workers getting paid too much to do too little. Reality check: there's no such thing as an HTML programmer. Writing VB is not going to earn you $50/hr. If you don't like what you're doing, you're not in the right line of work. The lion's share of jobs lost to offshoring are jobs that were filled by wannabes during the .com years. I personally know at least 5 administrators and programmers that refused to ever accept a lower-paying job when things went bad. They lost their cars, their houses, and their dignity as a result, and all for a job none of them liked doing in the first place.

Finally, as other posts have noted, the cost of paying a programmer is not the largest portion of developing software. Gathering requirements, testing, working with customers and clients, managing change, administering systems; all enter into it and have similar contributions to the overall cost. In the case of offshoring, almost all of these become more expensive...in some cases much more expensive.

Blacksmiths are still in demand...

[Anonymous Coward]

They're just not as prevalent today as they were yesterday.

As long as there is hardware that has to do a job, there will be a need for a program to control that hardware.

The important part of this conjecture is this:

A program is merely a series of logic routines executed for the purpose of controling and dealing with the input of data and output of results.

E.g. In an environment like a dank hole in the ground the logic could as simple as "Water present? Turn on. Water absent? Turn off." and the hardware nothing more than a simple pump with a moisture sensor, but someone had to create the "program", or input and output receptive part of the machine to control it's on/off state.

There will always be a need for someone to create a program, whether that program be written on a computer in C, Java, ASM, what-have-you, or even if they're just creating a physical object that can change it's logical state in order to "adapt" to a changing environment.

Maybe the label of "Programmer" will be one day gone*, but I don't think the essence of the programming field will disappear.

*Probably replaced corporatively by Serf.

Naah

[mav[LAG]]

There was a time when programming meant the encoding of finely-honed algorithms, parsimoniously pruned to the fewest possible instructions; algorithms which would then also have utilized the least amount of memory for execution.

That time for me was this morning. My clients still appreciate small, fast code that can be developed quickly, doesn't fall over and is agile enough to adapt to new requirements. Most of it is in Python, some of it is in Pyrex or C and yeah, some of it is still in assembly. Assembly is not a waste of time or a hangover from the Apollo days but a useful addition to any coder's arsenal especially given the insight into Leaky Abtractions that it brings. YMMV of course - asm is really not required for high-level business apps but it comes in really handy for certain kinds of video processing.

The vast majority of coding today is ignorant of such constraints. A 2K limit for even the simplest of applications -- even those written in Java, which was ostensibly designed to minimize an executable's footprint -- would today be considered absurd. And that's just for the minimalist application. Never mind the JVM, which is sort of required for anything meaningful to occur. But an entire operating system squeezed into 2K? It is obvious that the skills required for a successful programmer in 1969 are very different from the skills required for a successful programmer today.

And yet they're surprisingly similar. Given:

• this set of instructions
  
• this manual on how to use them (sometimes incomplete or even misleading)
  
• this platform
  
• your willingness to learn new techniques and technologies and
  
• a considerable degree of focus and dedication
  

can you do X for us in time T? I contend that all programming is about limitations and overcoming them: whether they be memory, time, operating system capability or human. No super-genetically evolved system is going to replace a smart human anytime soon. This guy needs to do two things: a) learn a real programming language and b) read the Mythical Man Month. There are no silver bullets.

Field far from perfected

[bigberk]

Is this guy joking?? The field of programming or software engineering is far from obsoleted/superceded/replaced.

Go and ask computer users (both novices and professionals) about what they think of the software they use day to day. Most will complain about the software they use (unreliable, unstable, missing essential capabilities...). That alone tells us there is still much more work to be done.

I don't think it matters whether you use UML diagrams, Java, C#. What matters is that everyone in business today is a computer user, and a large percent of these computer users are not happy with the software they are using.

There is plenty of work for those that can solve these problems. IMHO.

Fad is Why We Always Need Programmers

[reporter]

Programmers will always be needed -- for the same reason that mechanical engineers will always be needed in the auto industry.

The automobile is a mature technology. If everyone would settle for one styling of car, then we could drive the price of the car so low that the profit margin is barely 1 penny. There would be no need to change the metal stampers. There would be no need to modify anything in the car. All 200 million adults in the USA would drive the exact same 4-door sedan. The economies of scale for the exact same car would be enormous. The cost could probably be driven down to $5000 per vehicle.

Yet, this scenario will never materialize, for people's tastes change constantly. This constant change requires new metal stamps and the ensuing alterations to the engine and powertrain. So, mechanical engineers will always have work.

Now, consider the word processor. It is a mature technology. There is nothing fundamentally new on the horizon. Yet, we still need programmers to change the look and feel as the tastes of the customer change.

Further, one payroll processing system does not meet the styling tastes of all the corporate customers. Each company will want programmers to tailor the payroll system so that it just suits the tastes of the company's management. No two companies will have the same tastes.

People are different, and these differences change eternally. Rod Stewart sung "The First Cut Hurts the Deepest". Now, Sheryl Crow sings "The First Cut Hurts the Deepest". The songs are identical, but the delivery is different. The new style of delivery now addresses the tastes of the new generation of punks in high school.

"...do you think I'm sexy, and you want my body....come on sugar, tell me so!" -- Rod Steward, 1977.

Reality check

[wombatmobile]

Extinction of the programmer species means for me that I lost a few years of my life and now I have to change my carreer.

If it happened, but it hasn't.

To succeed as a programmer, develop better powers of discernment (by observing others you respect) and learn how to maintain confidence in your decisions.

I Read this Goofy Article About...

[Trolling4Dollars]

..."self writing programs in the armchair IT publication "PC World". I think it was the July or August edition. It was about this company that is writing the "next generation" of development tools. They suggest that in the near future everyone will be a programmer because this new approach basically lets anyone build their own applications by pointing and clicking and writing basic text descriptions of what they want the application to do. Then this development environment sets out to building the application to their specs.

To me, I thought... "Aha! He wants to make programming go the way that word processing and data entry did after the 80s". For the youngsters in the audience, believe it or not, but there used to be a very short lived period of time when people went to school to learn how to become word processors or data entry specialists. As soon as GUI based word processing and office suites were released, those people lost their jobs because "anyone could do the tasks". Is that really true though? Read on...

Yes, it is true that pretty much anyone who is plunked down in front of GUI based word processing software or an office suite can build a pretty impressive looking document. It doesn't require schooling anymore, just some introduction and you're off. But, does that mean that same person can design a glossy publication like a magazine or even a well done news letter? No. What about the person who can set up a basic spreadsheet? Would you want them being the DBA for your Oracle database? Probably not. So the titles have changed and the availability of jobs shrunk. I think you're going to see the same thing with programming.

There probably will be a point when the average office jockey will be able to build a basic application for use in their workplace. But that guy isn't going to be a programmer and the program he built isn't going to be efficient, compact or reliable. Programmers will still be needed, but just not in the way that they are now. It's likely that programmers will become more specialized. (Just like you now have graphic design people who are the extension of the more talented word processors of the late 70s) Programmers aren't going to disappear. They are going to become something else.

Re:great news!

[delco]

Now this writer is talking about "codeless development environments" which are (like the 5GLs and expert systems of yesteryear) supposed to replace programmers.

The way I see it, if the current demand for software systems doesn't change, this argument may possibly hold *SOME* water. However, users keep pushing what they want onto the cutting edge of what current systems can do.

For an analagous example, look at processing power. When companies come out with faster processors, do we just run the same applications, but now just faster? No. Not only do our standard applications do new and *cough cough* wonderful things, We find new and exciting uses for our extra processing power. which in turn pushes the need for even faster processors.

It seems that the extra time made available by the eases of case tools, intelligent IDEs and more intuitive syntax is being put to use implementing more sophisticated architectures, developing more robust business logic and implementing more useable software overall.

Without the increase in productivity associated with evolving programming tools (ie languages,environments,frameworks) would we have more programmers working on projects of the same scale as today, or the same number of programmers working on projects with less usability?

BZZT! Wrong, sorry!

[Safety Cap]

... from the article, the author states that 'coding is dead' because

The advent of the Codeless Development Environment

Ahh, the mythical Blue Unicorn; users will be able to make their own Uber Apps and developers will sink into the La Brea Tar Pits. That could be the case if development was only banging code. It is actually deconstructing a business process/issue into manageable chunks, then writing code to make those chunks work together. I have yet to meet a business person who can explain the function of their job in a linear process, so there is no way that person could write their own program, unless you had a magical CDE that could translate abstract, random process descriptions into real code (forget about efficient code/processes).

The emergence of the software factory

Ahh, the anonymous factory that churns out components that our business person can then assemble into a working application? Been there, tried that with EJB. Somehow that whole promise of component EJBs never really took off. I wonder why? Could it be that this idea is CRAP? Hmm...

The decrease in the number of programming jobs in IT

Non Sequitur. The number of quality jobs aren't decreasing. The monkeys that thought they could get a MCSE and start making $65k with no experience are (hopefully) flushed out of the pool. Soon the monkeys who are unable to do the work of Business Analyst/Junior Project Manager will also be gone/outsourced. The jobs are there; you just have to know how to look for them.

Fewer students enrolled in computer science courses

Another non-sequitur. A degree means nothing in terms of fitness for doing a job---except to PHBs who don't know how to judge a candidate's worth other than by dead trees (resume/degree: equally worthless). You can certainly be a good developer without having ever taken one college class. What matters is skill, experience, adaptability, talent, and self motivation to learn new stuff.

More:

When was the last time you cracked open one of Knuth's volumes? Do you even know who Knuth is?

Yes, last week, in fact. I've been plowing through Vol 3 looking for a better way to search a bunch of stuff across multiple iSeries/DB2 files. So what? Reading Knuth doesn't mean you're a better code monkey. It can help, but just like any other resource, it is in the application. Einstein said something to the effect that the most important fact one could know is the Library's address.

Re:BZZT! Wrong, sorry!

[Tablizer]

The number of quality jobs aren't decreasing. The monkeys that thought they could get a MCSE and start making $65k with no experience are (hopefully) flushed out of the pool.

I disagree. My observation is that managers generally LIKED these people. They had people skills that their other developers lacked. Perhaps they were not as productive, but sometimes managers don't care. They have to live with the person for 9 hours a day, so they really want a friend. I once learned that I was rejected from a position because somebody percieved as "less skilled but more personable" was given the job.

Except for Alpha Geek research posititions, being technically smart does not pay anymore. People skills are more and more the bottleneck.

AI can already pen better articles than this tripe

[guidryp]

Slashdot will become extinct if they keep posting this mindless dreck.

I posit that AI out there can already write articles better than this dreck. So perhaps Dreck writing should already be extinct, but here we yet again see more dreck on /.

About the time we have code writing AI, we will have passed Vinge Singularity and all professions will be extinct.

I have to wonder if the author ever worked on anything other than web pages. Or very simple coding issues.

I work in telecom with a university educated work force (including PHDs and Masters degrees) and these folks have a tough time building quality systems given the constantly changing requirements , competitive demands and huge complexity of the systems.

Complexity is only increasing as we go forward. In almost any field you now have much larger programming teams working on more complex problems.

This is as Naive as when the thought high level laguages (moving from assembly) Cobol, Visual Basic, OOP, 4GL, 5GL etc... would eliminate programming as a profession.

As the tools became more powerfull, the problem space became more complex. If anything there has been a push beyond the toolsets and coding today is more difficult than it has been in the past, rather than simpler.

Programming large scale complex systems is so difficult now that hardly any groups out there consider it manageable.

If there is any profession where we currently can't easily get a handle on, it is programming in the large. Once you have AI that can handle this, what couldn't it do?

very long-winded statement of commonsense ideas

[geg81]

The article is a long-winded statement of some simple and commonsense notions:

• Programming will splinter into a number of more specialized occupations (e.g., database programmers, GUI developers, etc.; that's, of course, already a big trend)
  
• Tools will let people with no programming experience create new applications (some tools like that exist, but they have rather limited applications).
  
• Programming will become "industrialized", carried out by assembly-line software factory workers (maybe, but noone has made this work really well yet).
  
• AI will obviate the need for programming altogether (I'll believe it when I see it).
  

Context matters...

[hagbard5235]

You do realize that historically it is normal for around 35-40% of the non-institutionalized working age (ie not imprisoned or commited) population not be employed?

ftp://ftp.bls.gov/pub/special.requests/lf/aat1.t xt

And if you'd like to see about part-time workers you can get recent statistics here:

http://www.bls.gov/webapps/legacy/cpsatab5.htm

Please note, there's nothing wrong with a part of the population being employed part time by choice. The BLS statistics differentiate these as people employed part time for economic/non-economic reasons... if you look at the stats you'll see around 4-5 million people are employed part-time for economic reasons. You can find those stats here:

http://www.bls.gov/webapps/legacy/cpsatab5.htm

According to the BLS the labor force is around 150 million:

http://www.bls.gov/webapps/legacy/cpsatab1.htm

Out of a labor force of around 150 million that means that we have about 3-4% of the labor force working part time because they can't get full time work. While I'd prefer this to be a lower percentage, it doesn't seem to be a great and shocking problem.

Also, would you please source your data above. Absent primary sourcing I've absolutely no reaon to believe your facts.

my take on the article

[Tablizer]

1. Article is a bit wordy. If the author's Java code is as wordy as that article, then no wonder he sees code generators replacing him. The trick is factoring repetition so that you don't need code generators. If Java is not dynamic enough to factor, then use Python or Lisp or the like.

2. Automation has NOT been a significant factor in my domain. Code generators are usually used by people who write bloated code to begin with. OOP's poorly-defined relationship with relational databases has also created a demand for tools attempting to bridge that gap by generating code. If anything, this OO/R conflict increased demand for coders. (Note I am not bashing OO here, just saying there is a philosophical conflict with relational that has yet to be resolved, resulting in tedious translation efforts.)

3. AI is irrelavant because cheaper bandwidth means that machines and people can hook up with cheap overseas human brains. Why invent cheap brains when we have billions of starving eager brains around the world?

4. I don't see bookkeeping safe from the same forces. As more and more transactions are electronic, your bookkeeper can be in Asia also, along with the coders. Programmers may be the first hit, but I doubt they will be the last.

Re:my take on the article

[Tablizer]

Cheap bandwidth is a temporary thing of the present. In the normal market, the increased demand will generate the increased prices and effectively all those offshoring companies may find themselves at some point at the mercy of telecom companies.

Maybe in the short term, but in the longer term there is a kind of "Moores law" equivalent to telecommunications that will bring the prices down to make multimedia cheap.

If the field is ever to evolve, we need to lift programming to another level of abstraction and create tools that will eliminate mundane typing of same code again and again.

They have been trying to do that since the 1950's with only incrimental success. We are already using tools that simplify a lot of stuff: GUI designers, report writers, scripting languages, databases, etc.

Re:great news!

[mitchus]

The great thing about engineering is that it's very polyvalent! It's about the greatest education you can get, because up to a point (at least not within your lifetime) the basic principles are never obsolete, and lateral mobility is relatively easy. Remember, you're an engineer before you're a computer engineer! I wouldn't say that you've lost any years of your life. Maybe you can't work as a coder anymore, but if during this time you became a proficient programmer, just realize that this is something that will be useful to you throughout your life, regardless whether it's not your job anymore. You just might find that there are actually more interesting tasks than coding in the engineering field :) Anyway, don't throw the towel is all I'm saying. Cheerio

Goff has no idea what he is talking about

[Flyboy Connor]

Goff thinks AI will be able to take over the job of a programmer. Hah! What a ridiculous notion. Evidently he fell for some vague claims of specialists in genetic programming, of which he understands nothing himself. Well, evolutionary algorithms (of which genetic programming is a subclass) are one of my specialities too. And I can tell you that, while evolutionary algorithms are fun to toy with, and can be used for some interesting real-life applications, they will never be able to design complete programs for us. And why is that? Because their results are notoriously untrustworthy. You will only use them for problems for which you can't design a solution by hand, and for which you are satisfied with an approximate solution. Most applications are simply not like that. Furthermore, evolutionary algorithms can only be applied to problems for which a fitness function can be designed. I wonder what Goff thinks a fitness function for a Word Processor looks like.

Obviously, Goff knows nothing about the practice of computing science. But he still manages to hawk out three crappy articles about it. Which get on the /. frontpage... Man, that guy is good.

The arguement is flawed

[LittLe3Lue]

First of all, your argument comparing programmers to blacksmiths is already a counter-proof for what you are saying.

Sure, we don't have many blacksmiths, but let's look at the role of a blacksmith:

All they did was make things out of metal. They were what we now replaced with machines. The few innovative blacksmiths, who created new advances, were what are now engineers. Nothing has changed - only evolved. And what was once a role for a person is no longer.

Now, you are saying that there will be no need for programmers in the same sense of how we no longer need blacksmiths. But I ask you, do we have a programmer re-type every line of code for each and every piece of software that is sold? Because that would be the role of a 'blacksmith' programmer.

A programmer (or large group of) writes the first piece of software, which must usually be innovative, or better then what exists, in order to provide a use. Sure, there are countless pieces of code being re-written, or written poorly, but that is more for learning than for business, at least if they plan to survive. And when there are copies of software in the mainstream it is merely for economical competition, which will never change.

With the open source movement, or even with API's of old, there is no need to write every line of code over again, but rather to use them to make whatever new piece of software it is that you are planning to make. This involves making software that has never existed, which cannot be automated.

So I conclude that software programmers have never from the start been blacksmiths, or at least have not been for a very long time. What we provide is a constantly changing, evolving, and adapting use of logic in the form of code. No automation will ever replace that.

The number of programmers will not be decreasing to extremely low numbers either, since the more advanced software becomes the harder it is for individuals to create it. Large groups are required, for long periods of time (How long does it take to make a new game or operating system compared to 20 years ago).

You argument is not only wrong in practice, but wrong in theory. It's ridiculous.

The Mythical Man-Month

[Toby The Economist]

Writing software is hideously complicated.

The skills required to successfully design and impliment software remain light-years beyond quantification - unlike the skills required to build a cart wheel or hammer, or, for that matter, a car body or a silicon wafer.

The concept of a "software" factory, building generic components, is more or less exactly that originally touted by OO back in the 80s.

This approach failed because large components have not been made (people tend to make "stack" and "list", not "user interface" and "VPN") and because of the huge front-loaded cost of making such components general-purpose rather than one-shot.

--
Toby

Complexity trumps components

[earlgreen]

C'mon already -- most of the software I've worked on probably has about the number of inter-related working parts as a 747 jet.

So somehow this stuff is going to start writing itself? I don't think so.

The software components idea is like building a machine out of stock parts rather than designing every little bit yourself. That's great but it still doesn't make up for the vastly increasing complexity of software. If you count parts and inter-relationships, you've still got more complexity that needs to be engineered.

OK, so yes, some narrow classes of software like certain office apps can be designed and built in more and more automated ways, but this also doesn't make up for the explosion of what computers and software can do.

Just because people don't readily see code, they seem to think it can be magically woven in ways that engineered physical objects like ships and planes and cars obviously cannot. They also think that code is code is code, which is like saying a machine that designs and builds toasters automatically can also design and build ships.

Sorry, code is like matter and components are like parts and you still need engineers to put the crap together.

Re:The Future for Nerds

[ltsmash]

I wonder if once nerds lose their jobs and therefore their money, they will return to the bottom of the social pyramid again.

I'm not sure if you're just trying to troll, but there is a lot of truth in this. When I was in my 20's, before so many people were making so much money with software, a nerd was an extremely pejorative term to call someone. Today people in our industry get so much more respect. I fear that once money starts leaving high tech, we will again be returned to this position.

Re:great news!

[bitingduck]

Now that I've studied for more than 4 years, I learn it is going to be useless.

College isn't a trade school, and you shouldn't treat it that way. The most important thing you learn in college is how to learn. In many, if not most, fields what you learn in college is outdated by the time you finish (if it wasn't when you started) but it does (or should) give you a strong background from which to learn other things. In graduate school you learn how to learn things that nobody knows yet.

As an example, an undergraduate physics degree from a pretty decent school will get you to about the mid-1950s as far as physics knowledge, with a few little tastes of stuff from the 70s(and maybe even the present, if you work in someone's lab). You can fake your way into a lot of engineering jobs from there, and if you go to graduate school, you catch up to maybe the 70s (or even 80s and 90s) in a lot of areas, and you take one small piece of physics right up to the present day and become a world expert on it, adding new knowledge at the leading edge. All the stuff you learn along the way provides important context and background knowledge, but the most important thing you learn is how to obtain new knowledge. If you need any of that stuff that you didn't have time to learn (because the field has gotten very large) you at least get the tools to go back and catch up quickly. Computer Engineering has to be much the same, if not more so, since things are changing even faster than in physics.

pompous and vapid at the same time

[uncadonna]

I don't see much of interest in the article beyond some amazingly bad writing. In fact, someone who writes this badly in their own native language is almost invariably not worth listening to.

I think the author was rewarded in school for writing long sentences with big, trendy words. This sort of ridiculous blather is the result.

Some of my favorites:

• Overuse of double quotes: It is important to note that autonomous agents compete to "make a living" in a fitscape, and as such tend to fill all possible "life sustaining" niches such as to provide for their specific "needs."
• Wordy and vapid way of saying "sometimes economic activity changes the economy": While this may seem painfully obvious, when it comes to the study of economics, that feedback mechanism (i.e., the fitscape evolves as a result of agent activities) is often dismissed; a function of exogenous variables which are outside the scope of consideration for the problem in question.
• Is this a poetic essay or a dry academic analysis? Try not to change your mind in the middle of a sentence please. Grotesque neither-fish-nor-fowl writing. At first, the smithy was not aware that his hammer would soon ring no more, that machines would take the place of hammer and anvil, more efficiently creating the tools and weapons for an increasingly sophisticated cultural milieu.
• Sixth grade level spelling error. A possesive is not a plural. Smithy's also made the tools that were necessary for the daily household chores, such as pots and pans for the fireplace.
• Uhh, "first order derivative"? If this doesn't rub you the wrong way you've never had a calculus class. There's no law against that, of course, but this comes in a paragraph about the scientific method! And the essence of measurement, in a capitalist system, will always be a first-order derivative of the much-maligned bottom line--it's not personal, it's just business.
• Let me introduce you to the idea of a dictionary. "rubric" means "name", or at best "name of category". Applying a similar rubric to the postmodern computer programmer might now yield some insight as to the coming fortunes of that trade.
• (add your own comment here, I'm ovewhelmed) The anarchistic, eclectic, and often incoherent ramblings that characterize much of the Pomo Jones mindset belie the importance and value it may conceptually provide. Computer programmers are icons of the postmodern age and sometimes idiosyncratic of its nature. As much as film, and more than television, architecture, or art, the postmodern programmer has enabled the cyber-collage, globally diverse viewless worldview that is the hallmark of Pomo entrails. Indeed, computer programmers (those of the Java, C#, and Perl caste, at least) are the Pomo elite.
  uh-huh
• and here I thought it was about algorithm and syntax... Today, programming means breaking the bounded, solitary-node paradigm, enjoining a network that is effectively limitless but unique at every node.
  Sure, they'll mechanize that one any day...
• Ooh, let's throw in a random bit of computer history to show we know something about computers. But just as the entire notion of "writing code" has changed dramatically since the earliest _stacked job batch systems_ [link to an undergrad CS lecture] of the 1960s ...
• "ignorant"? The vast majority of coding today is ignorant of such constraints.
• Do you know anyone who talks like this? The fact that the nature of programming is changing even as the demand for programming skills has abated, to serve what is otherwise an increasing global demand for IT-related products, should be a glaring indicator and something of a concern to skeptics; even the computer industry is not immure from the ephemeralizing virtuous cycles wr

Damn...

[natet]

I knew I should have studied accounting...

Seriously though. I believe that the same things he uses to point to the demise of the programmer could also be applied to his test case of the bookkeeper. I believe that it would be nearly as easy to create a "bookkeeping factory" as it would be to truly create a "Programmer factory". But, I don't believe that either of them are going to happen any time soon.

Other pieces of his logic are suspect as well. Yes there has been a decline in the number of computer science students enrolled, but the article doesn't say if that statistic takes into account the number of students enrolled in Information Technology majors, or other related majors. I know that a lot of the students that started in Computer Science with me eventually switched to Business Information Technology to avoid having to take all the math required in CS.

Re:great news!

[Madcapjack]

College isn't a trade school, and you shouldn't treat it that way. The most important thing you learn in college is how to learn.

You're right , of course, but, its no consolation to a recent graduate with 30,000 in debts and no job prospects.

Re:Codeless Development

[lachlan76]

I tried using the Lego development environment.....it made me wish I had assembly. Sure, I didn't write any complicated programs, but when you have to click the motor 1 button, click in the workspace, click motor speeds, click power level 3, click in the workspace, then click wire, and connect the two, you really appreciate how much easier

MOV dx, 1
OUT dx, 3

is than using a codeless development environment.

A shiver runs down my spine whenever I hear "without a line of code".

Some kinds of programming are already dead. But...

[argent]

30 years ago, a large percentage of the programmers out there were writing what was basically spreadsheet code. Some manager needed some information, and that meant a programmer had to write a program to read data, generate reports, and add them to the daily jobs. Then came spreadsheets, and the manager could do that work instead. The programmer was out of a job...

Well, no. Instead, he worked on more sophisticated analyses, software that didn't fit in Visicalc or Lotus 1-2-3.

Time after time over the decades there's been predictions that some new productivity tool was going to replace programmers. It hasn't happened. Because what a programmer does, basically, is to do the things that people havethought up that haven't yet become popular enough or common enough to be canned in a tool.

The advent of the Codeless Development Environment

If what you're doing doesn't involve code, it's not development, it's configuration. And that's great, because having programmers wasting time on configuration is a drag on productivity. Anything you already know how to do, that you've already written code to do, you don't do it again. If you do, well, that's a problem you need to fix, write a program to do it, go on to something else...

Re:great news!

[bitingduck]

Job prospects in tech things aren't what they were 6 years ago, where you could practically walk out of school and roll in money, but there are still tech jobs out there. And 30K in debts isn't chump change, but it's nothing like doctors and lawyers collect in debt, and loads of them don't wind up in the big $$ jobs.

The thing that most bothers me when I check out job web sites is that the only work that seems to stay in the US anymore is defense stuff that can't be offshored because of the need for cleared people. There needs to be some of that kind of work, but if the only thing produced in the US anymore is pizza delivery and weapons systems, we ought to be worried.

Evolution

[DaChesserCat]

The author posits that blacksmithing has died out, while bookkeeping has continued, because the knowledgebase that the blacksmith needed was ripe for scientific analysis and automation, while the bookkeeper has to deal with constantly changing requirements. Therefore, since bookkeepers have to always change/upgrade/modify/tweak their craft, they aren't as likely to be automated.

He asks how many people are acquainted with a bookkeeper, and how many are acquainted with a blacksmith. No, I'm not acquainted with many blacksmiths. How many mechanical engineers, or mechanics am I acquainted with? More than bookkeepers, I'm afraid.

The trade of blacksmithing, where one person forges, casts and manufactures something, cradle to grave, is largely gone. However, there are an extensive number of people in this world who know how to shape metal into useful things; I believe they're better known, these days, as engineers. They may not do the actual bending, machinging, forging, etc., but they do the design, which is something a blacksmith did. Also, mechanics may not manufacture their parts (something to which the author alludes), but blacksmiths were ALSO responsible for repairing existing items. Sometimes, that meant fabricating a part, sometimes that meant taking a pre-fabricated part and applying it (or modifying a similar part) to get it working. Mechanics do that, these days, and they have a much larger catalog of pre-fabricated parts to work with, so they don't typically NEED to fabricate a part for the job. Although, in my shade-tree mechanic experience, I've been known to take an existing, not-quite-what-I-needed part, and make it work.

So, while the trade of blacksmithing, as a cradle-to-grave manufacturer, mechanic and engineer has diminished, the different roles they played have spawned their own career fields. Engineers use machines to produce much larger quantities of parts than traditional craftsmen could manage, and their quality is more consistent. Is this the demise of a trade, or simply the application of technology to improve its productivity? I would argue that, if you took all the people who are now working in roles which blacksmiths once filled, the various sub-fields of blacksmithing have MORE people working in it than the bookkeeping field.

I develop websites, for internal use, for my employer. The language and technologies I use were pretty well evolved before I finished college and went to work in this career field. I don't have to code machine language or even assembly; some other developer developed the tools I use. Because I don't have to re-invent the wheel every time my boss wants a new web app, my productivity is much higher. The original programmers were the "blacksmiths" of this field; they created some practical, useful things for their clients, and they create tools which improved their own productivity. Their tools led to the development of other tools. Consequently, there are few if any programming "blacksmiths" in the world today; most all programmers are engineers, mechanics, or factory workers feeding the automated systems. The titles just haven't changed; that's all.

As for bookkeeping not being as ripe for automation, how many bookkeepers do you know who would be willing to go back to pen and paper? Give up their calculators? Give up their computers? I'm willing to be the answer is very few. Their field has also benefitted from automation; a bookkeeper today, with his computers and software, is at least a couple orders of magnitude more productive than an old "quill and paper" bookkeeper of a couple centuries ago. However, the title has changed little, although the career has changed significantly.

As a programmer, I see the technologies I'm working with evolving on a regular basis; I have to adapt to those changes, like the bookkeeper in his analogy. Does that mean that my career will become extinct? I sincerely doubt it; I'll probably be more productive ten years from now, but I sincerely doubt that an automated softwar

Re:great news!

[Thangodin]

I will say that I agree with him on hyper-human skills. Computers are being used in the wrong way. When I worked for a dot com, everyone was raving about alltheworld.com, a flash in the pan search engine. I pointed out that I didn't want 60,000 matches--I just wanted the one that I was really looking for. They all looked at me funny. Computers are being used to flood people with data. What they should actually be doing is giving people the few pieces of data they actually need. And nobody in the dot com world seemed to grasp this. They had no idea what people actually wanted or needed. As Winston Churchill put it, "Give me this day, on one sheet of paper, the following information..."

But the current great extinction has nothing to do with any of the things he's talking about. It has to do with the outsourcing, the bust of the dot com bubble, and a popular belief that now that the dot com bubble has burst, you don't have worry about that computer stuff anymore (I kid you not--a lot of technophobes out there actually think it is all going to go away.)

There is no question that programming will evolve, and that we will work at higher and higher levels in the future. But it is equally true that as the methods ramp up, so will the target, and competition to produce software that reaches that target quickly, efficiently, and easily will demand the best solution, which is always hand tweaked. Build by number tools are always general--2,000,000 lines of code used to do what 1000 lines would do better, because the code is literally written with no idea of what the real task is. And to use these high level tools effectively there are always "Tips, Tricks, and Traps," which require a background knowledge of the underlying architecture to grasp.

And by the way, assembler still does everything better than anything else. Check out Gibson Research (http://www.grc.com/) for proof of that.

As for genetic programming, allow me to contrast the hand coded solution to the genetically designed solution:

Hand coded: 2 + 2 = 4;
Genetic: 1 + 1 + 3 - (4 * 3) + 3 + (6/2) + 2.5 + 1.5 + ln(e) = 4

Yeah, it will give you the right result... eventually. But if you want to read your email, you better book a couple days time with WETA's server farm. Not to mention that it takes as long to train one of these as it does a human--and it takes someone who knows a lot about computers (a very good programmer) to specify the criteria. A good example: the Pentagon wanted a genetic algorithm that could recognize a tank. They got one that seemed to work, until a new set of photos were used. It turned out that all the criteria photos with tanks were taken on a sunny day, and all the others were taken on a cloudy day. So if the sun is out, it's a tank. Doh!

The more sophisticated the software is, the harder it is to fix when something goes wrong. Superficially simple applications for complex tasks are that much harder to diagnose and fix when they fail. Windows appears easy, until a bug rears its head, and then it takes thousands of man hours just to track down the cause. Linux has a steeper learning curve, but a core simplicity once you're over the hump. Simple software is simple because the brain is built in. But that doesn't mean that the brain will always work.

Eventually, someone will have to go down and sing to the metal.

The blacksmith isn't gone-- it's called machinist

[bitingduck]

I finally went and RTFA, and I disagree that the blacksmith disappeared. The blacksmith became the machinist, and the machinist population is slowly decreasing, but is unlikely to go away.

Production machinists (who were in many cases still quite a skilled population) are being replaced by CNC machines, and the "machinists" now are the people who turn art into things that the machines can cut, set up the machines, diagnose their problems, etc. Eventually they may be replaced entirely by mechanical/manufacturing engineers, but the function will be much the same-- turn hunks of metal into useful stuff for people. They just do it in a higher tech environment.

The more blacksmith-like machinist is the toolmaker. These people are ever fewer in number, but are likely to never disappear entirely. Just like there is a small but steady need for technical glassblowers (to supply chemists) there will be a need for toolmakers/precision machinists in the foreseeable future.

Programming could be seen in a similar way-- the tools are improving drastically-- my dad started with toggle switches to set the bits, then moved to punch cards. By the time I was programming, the punchcard was all but dead, the microcomputer had just come on scene, and the mainframe was about to become a brontosaurus. Things were done with a command line. Now there are fancy IDEs for all sorts of languages, graphical languages and development tools, automated code generators, etc. It's still programming, just with different tools. The need for codemonkeys may decline, but it probably won't go away. There will be a market for them just like there still is for analog electrical engineers in the digital age. There won't be a need for people to hand produce boring code, but there will be clever things that need to be coded by hand, and probably lots of them.

My "academia" meter was flying high

[JavaRob]

Maybe you guys didn't catch this sure-fire sign of overly-academic writing (sometimes also found outside academia, alas):

The pomo coder stripe has evolved into something quite different than what it once may have been.

You probably thought that word was "porno" -- but no, it's short for post-modern, and it's thrown around all the time for no good reason in academia. I stopped reading the article after seeing that, and a smattering of insofars and posits when I flipped ahead to part 2. Oh yeah, and I ran into this:

But as we will see, the impact on the exogenous environment of remittable activities of autonomous agents can be profound indeed.

Indeed, indeed. That's a great sign that you are reading a paper that will take thousands and thousands of words to argue something that could have been said in three crisp, short paragraphs. Nothing against the writer -- he's got a great vocabulary, and he's using the words correctly... but you see this all the time in academic writing, because "it just doesn't sound right" without a few latin phrases and a "dialectic" or two, because everyone around you is writing like this, too. Sadly, it only complicates or even loses the real argument, and most of your readers who have never been force-fed this kind of fare and made to like it will also be lost. So, to the author: Fight the impulse!! Turn away from the dark side! Therein lies a twisty maze of passages, all alike....

Re:My "academia" meter was flying high

[GCP]

most of your readers who have never been force-fed this kind of fare and made to like it will also be lost.

I'd give you mod points if I could.

I've been force fed this kind of academic drivel for decades and nothing will make me ever like it. I was even one of the editors of an academic, peer-reviewed technical journal years ago, but my attempts to turn the academic gibberish into clear English were ultimately fruitless. Both authors and senior editors seemed to feel that clear writing posed a threat to their reputations for academic seriousness.

This article was a particularly egregious example of academic-ese. As near as I can tell, the three part series boiled down to the simple thesis that programmers are creating automated systems that will eventually replace most programmers, effectively eliminating their own profession. Why couldn't he just say so in the first paragraph?

With a clear and concise statement of the thesis, the arguments and data needed to support it also become much clearer, as do holes in the argument and obvious counter arguments that must be dealt with. But with enough academic encruftation, you can't see an argument clearly enough to reliably judge it. Maybe you just haven't understood it. Maybe you're not capable of understanding it. Maybe he's a genius and you'd better just take his word for it.

Sorry. I call pomo BS. I'm sure many current programming jobs will disappear, but many will evolve and new ones will appear. How those factors will ultimately work out isn't clear to me yet, and I'd welcome any article that actually clarified it.

Standardized businesses

[yow2000]

Blacksmiths do custom work, to fit the need - but those needs became standardized.

Bookkeepers also do custom work - but what stops those needs from becoming standardized? Why don't businesses deliberately organize their affairs, so they fit within an exact model of a business, so that accountancy does become standardized? I think the reason is because different organizations of a business is often a competitive advantage - they don't want to standardize it; and if they did, straight off someone else would come up with a better organization and put them out of business.

Complete nonsense.

[jotaeleemeese]

In rich countries people normally don't starve to death, don't die of preventable diseases, live longer than in most other places and only have bad old years because they are too lazy to go out and do some exercise, stop smoking and eat sensibly.

Unless you are posting from Sudan or a jail cell in China I just think you have no idea what a human standard of living really is.

People in rich countries, even poor people in relative terms, are well fed (heck obesity is frequently associated with poverty in rich countries) and adequatelly clothed, so frankly cut the bullshit man.

In this case, the analogy must be almost perfect!

[lamz]

In the end, the blacksmith and the bookkeeper are only convenient metaphors, not to be confused with inherently meaningful symbols. Any number of contrasting metaphors may have served just as well.

Umm, no. This essay purports to draw an analogy between blacksmiths and programmers, with a contrast to bookkeepers. In the end, the analogy better be as close to perfect as possible. You don't get to cop out in the last paragraph of a three-part essay! If the blacksmith isn't perfectly analogous to a programmer, then keep looking -- you haven't found your analogy yet!

Industrialising book-keeping

[jkozak]

One of the article's central claims is that a "book-keeping factory" is impossible. On my home world we have entities called "accountancy firms" which are exactly that.

A more accurate analogy...

[greywire]

The problem with this analogy is that it's too simple. "Programming" spans a greater skill set than "blacksmith". A more accurate comparison would be not to "programmers" as a whole, but to "low level programmers" -- people doing tightly coded assembly, or even low level libraries and objects in C, C++ etc. Because these things are equivalent to the things smiths made -- commodity components like fasteners, gears, and even small machines like electric motors.

Like somebody else said, programmers will become more like the engineers who take common components and put them together, only sometimes making any low level changes to the common parts.

Yeah, it was great when we all did amazing things with our 2K of program space in assembly code. I really miss those days. But I dont want to do that anymore than I want to design and fabricate a gear so I can build a robot. Now certainly there are going to be a few people somewhere who need to design things like that, on a low level.

You can find all sorts of similar analogies. How about textiles? Do you know anybody who makes their own fabrics? Do you know anybody who makes their own clothing and who laments the fact that they dont get to weave the fabric first?

I'd rather stand on the shoulders of giants -- and create bigger and better things using the components people have perfected before me.

When I first started to read this article (which I found tedious to read) I thought sh!t, he's right! And then I realized I'll still be programming in the future, just using more tools than low level programming. Already I use code generators and libraries of functions and object components. Programming will be more like engineering the overall structure, or designing the data base or protocols, etc. That's more fun anyway.

Nostalgia aside, hooray for the death of the programmer! Long live the Software Engineer!