Quicksearch
Syndicate This Blog
Wednesday, July 15. 2009
Open Source vs. Free Software vs. "Open Source"
Comments
Display comments as
(Linear | Threaded)
I quite agree with this OSS vs. pseudo-OSS thing. I’d like to express a bit more subtle view though. What I mean is that there are kind of “socially ideal” OSS projects and there are less ideal ones. A project with friendly upstream which welcomes patches and opinions and makes contributing easy are kind of more OSS than projects with more closed development.
For example, Vim editor is an OSS project but it’s controlled entirely by a one man. Occasionally there are some other people who send small bug fix patches but really the developer community is almost non-existent. Discussion about features happens almost only by users. So Vim is an OSS project (and an excellent text editor!) but perhaps not quite ideal as a project. Another example are GNU projects which require copyright assignment for Free Software Foundation. There is a community but political reasons make it more limited than it would probably be otherwise.
So, one could define some kind of metrics for projects’ open-sourceness. Many metrics are social like friendliness of upstream and project’s social organization. Many metrics are technical like the use of good version control system, communication channels, availability of resources (e.g., web pages, API documentation etc.). Obviously in preudo-OSS projects most of these factors tend to be on the “closed from public” side.
...then we complain when people claim that the FLOSS is not business-friendly.
FLOSS licenses have nothing to do with having a community of developers instead of having a company driving the development.
You are implying that, for example, Zabbix and MySQL shoud have never been released as FLOSS - same thing for a lot of kernel patches from many companies including Vyatta and Red Hat.
I'm saying that users of Pseudo-OSS software aren't getting all the benefits they could and should get from using FLOSS. I'm not saying what those companies (you name MySQL and Zabbix - the latter I haven't really looked at; I also thought about Zimbra and OpenXchange and a few others) should do or shouldn't have done. I didn't name names on purpose, but since you've brough up specific cases: OX is one of the worst examples I've seen so far: the mailing lists are just silent, with very few user questions and even fewer answers, most of the answers by OX staff (state as of a few months back). So not only is there no developer community outside OX, but there doesn't even seem to be a working user community. MySQL, on the contrary, demonstrates a somewhat working compromise, but the number of derived products / forks that are floating around (including those created by former key developers) and the fact that complaints about release management and development processes come up again and again tells me that it's still not working as well as it could (I'm curious how this will continnue with the new owners.) Of course it's not only the commercial background that matters here: eglibc and wodim demonstrate that similar things can also happen for other reasons.
One good metric is whether software is available in Debian - pseudo-OSS generally isn't. This particularly filters out the badgeware licensed applications.
This is a very interesting question. I think at this point you're going to have to do some experimentation. Try a few of the proposed metrics and see if they classify projects in accordance with your expectations and beliefs. Keep us updated on the results.
Comments