It often takes an expert to know whether a job is simple or not. And it often takes that expert some investigation to find out. DIY is fine, but specialization of labor is fine too.
Also, web scraping is a great example of something that can either be simple or complex depending on many non-obvious things.
A very good advice. One thing was left out, I think. Implementing the program one self the implementor may:
* Gain an insight into the problem he was trying to solve he would not have otherwise had
* When faced with unexpected complexity, the implementer may choose to solve a less complex version of the problem that may still provide added value, or choose to go some other route entirely.
I think this last bit cannot be overstated enough. The easy problems in programming are usually so easy that even novices can solve them in a sensible time, while the hard problems can be hard to get right no matter how many years you have under your belt. Often by changing ones approach the problem can be reformulated into an easier one.
If the work is hired, then the consultant may lack the initiative, intuition, courage or experience to go for the possible alternative approach to solve the given problem.
Of course, all this depends on the quality of a problem.
Thinking of "programming" as a completely separate field, to be "done" by specialists only, is highly unfortunate.
All "programming" really is, is formulating what you want done precisely. A compiler can take it from there. Where a specialist in "computers" comes in, is he can optimize your algorithms to run more efficiently. But in more cases than not, what "programmers" are tasked with, is to "just make it work", with little more specificity than that. And then the garbage inputer gets upset because he gets garbage back out.
SQL was originally, back when clarity of thought was still considered fashionable in executive suites, intended to be used by decision makers. "Specialist" IBM "programmers" provided the code that accessed and processed the relevant records efficiently, but the person who wanted an answer, often a C level, was the one tasked with specifying what he wanted done. With enough precision that his question made sense.
I agree with you that programming shouldn't be done by specialists only. I also think that's the way computers are supposed to be used (as Steve Jobs put it, a bicycle for the brain).
However, I also experienced the problems this will cause. I was working at a software company, where the software was started by non-specialists who then took over the development section. It was a mess.
It all depends on whether something is worth your time or not. A person whose time is worth more might find it more valuable to hire someone whose $/hr is half of his (not necessarily cheap) who can get the job done in less than twice the # of hours he would take. There are some problems / jobs that are just tedious that don't require a person with 10 years of experience to complete - those are the kind that are worth hiring for.
My takeaway from this article: if a client tells you he needs a hacker to do an "easy" job, tell him to do it himself. If he thinks it is easy he won't value your work anyway.
"I know it is easy, I mean, how hard can it be? It's just not my area, you know, so I think you should do it... but since it's easy, I won't pay you much and I won't appreciate your work. So when can you start?"
Also, web scraping is a great example of something that can either be simple or complex depending on many non-obvious things.