the tedium of programming

Recently I talked with a scientist who was very happy with his choice of work. I asked him why he liked being a scientist in a field that was constantly changing, and his answer was simple: he was as much the agent of change as any other researcher and he held out the hope of making an important discovery.

It occurred to me that the main thing that is tedious about programming is not the bugs that need fixing or the features that are tedious to implement in an of themselves per se. A programmer's day consists of making numerous "little victories" over the computer and that is satisfying.

The most tedious thing about programming is that unlike for that scientist, the programmer is never the agent of change. Very few programmers by themselves alone create a novel new piece of software. Significant changes in this industry are imposed by larger forces, not invented by programmers. They come in the form of languages, APIs, clbuttes, data and file formats etc.

They come sometimes from Microsoft, which creates new standards that can be frustrating because they seem boneheaded and badly designed and documented. Apple has in the past been equally frustrating in this respect. They also can come from new open standards, which are a dime a dozen, cropping up constantly like weeds and yet clients want programmers to be familiar with them. They can come from clients themselves who may have proprietary data or programs.

But the key thing is, to be a programmer, I buttert, you have to be the kind of person who can accept the idea that another person or company 'upstream' from you will tell you what topics and systems you have to learn about and adapt to, even if what they require is monstrously stupid, uninteresting, inefficient, creates lots of needless work, etc. Furthermore the results of one's work may not be elegant, ideal, beautiful, effective, or in any other way appealing because the job really is to write what the client(s) need. Thus a programmer can't always take much pride in the product of his work.

Whereas, a scientist only deals with current theories and facts which he finds interesting because he wants to learn more about some aspect of Nature. His product is his research which he may take great pride in. Or in the case of some kinds of science he may have specimens he displays with pride.

Therefore I would argue that in order to be a programmer one has to have a fairly subservient personality who is able to sustain large amounts of possibly needless frustration without a great deal of recognition.