There is definitely a subset of programmers for which this observation doesn't directly apply - the set of programmers that spend a good deal of time in open source communities. Why? With open source work, there is no set of "helper functions" and almost every contribution that you make is against a code base that is likely to have new technologies and languages, *and* both new and seasoned developers for it. Thus, the default is a steep learning curve, and you adapt to it.

Thus, the open source *default* is akin to what you describe in changing to a new job. So I'd modify your statement to say that *for some* changing jobs makes you (temporarily) a worse programmer, but it depends on the person, and what they are used to.

I don't know. I disagree that you become a "worse" programmer. I posit that, even if you're fighting a learning curve, you should always be getting better.

In a well-organized team, newcomers should be able to get oriented within days or weeks. Compared to the usual length one keeps a job or the scale to which one should measure professional growth, this is an insignificantly small period of time to judge a change in productivity.

I think there are trade-offs here. While it's true that a lot of your knowledge is worthless, but at the same time, you gain the superpower of seeing the future. You can tell your new team: "You're proposing X; my old team did X and here's what happened." If you never switch jobs at all, then you only know of 1 example (your current team). If you've worked at a few different places, you can see more patterns of how different decisions will work out.

That's why joining a startup, in which you build that body of knowledge, is a good way to overcome these issues.