Are there too many Linux distros?

TuxRadar asks that interesting question. Here's what I said:

A Linux distro is like having your own customized OS. So one would imagine a loooong wizard that asks you what you plan to do with your new machine, and then selects the best distro to suit your needs. Or better yet, creates a new distro out of your answers.

Of course, that's way too idealistic, but I think that people create their own distros based upon their specific itches.

Constrast this with Windows or MacOS X which come in one or a handful of flavours. What the OS designers did in those companies is to imagine or research the most common usages across their user base, which they then used to make decisions on how to package the OS. Ubuntu and OpenSUSE can be said to work similarly, although they have the advantage of a much larger inventory of software components (which is also a burden to integrate).

So is having many distros (how many are there? around 400 on DistroWatch) a good or bad thing?

As with most other things, depends how we deal with them. DistroWatch makes an honorable attempt at inventorying and indexing the available distros. On their search page, for example, you can choose many criteria to limit your search. Unfortunately, it's an old-school site that doesn't benefit from user recommendations and other Web 2.0 niceties - which could be useful in that case to let people rate, recommend distros to each other, etc.

The problem that remains is symptomatic of the whole open source movement: how to reach the mass market (Windows users) with that seemingly messy situation? That's not for Linux distros specifically to worry about, and besides, I think Ubuntu is becoming the "default" distro for Linux-curious people. We can thank them for this.