netlabs.org and its role in 2012

In September this year it will be 15 years that I founded netlabs.org. I never had any plans about how long I gonna run it and I’m amazed to see that it is still running strong. So time for a little stop to think about what we did and where we go in the future.

First, as most of you noticed I am not really active anymore regarding OS/2. I stopped using it several years ago and while I still miss the power of the Workplace Shell, it was time for me to move on. I don’t necessarily like current solutions (including MacOS X, which I use) but it was a pragmatic decision I do not regret. About four years ago I got introduced to the Semantic Web and that’s what I’m interested in nowadays. Me and some other guys from netlabs.org invest a lot of time into it and there is a fair chance that you can play with the first few things we create later this year.

So the next question is what happens with netlabs.org. In short: Not much. I will continue to run it and netlabs.org will be the place where all open source code gets released we create for our semantic web work. I will continue to maintain and support services like Subversion, TRAC, the minimal webpages, ftp, mailing lists, Jabber accounts,  SIP server and so on.

In the past months I got several requests for new things, including support for hosting code with Git or Mercurial. I was thinking about providing that and had a look at both systems. They are very powerful and outrank Subversion in many ways but in the end it is also a matter of personal preferences. In our semantic web work I started to use source hosted at GitHub as well and I have to say I’m very impressed by the service. It is how sharing source code needs to be done nowadays and this is something I cannot offer at netlabs.org.

When I started netlabs.org there was no SourceForge, no Freshmeat (now Freecode), no GitHub. What I created back then was a simple version of those things and it did its job for a really long time (remember CVS :). Later I introduced TRAC and Subversion because it made sense and facilitated project handling. But systems like GitHub and bitbucket are far too advanced to copy them and it wouldn’t make any sense to waste time on that.

As a conclusion: If you want git or Mercurial for your project feel free to host the code anywhere you like. netlabs.org was and is about the people, the IRC channel, the mailing lists and so on. netlabs.org is the history we share, the events we participated, the software we released. And I’m pretty sure it will stay that way for another few years and hopefully some new people join for the upcoming semantic web projects we work at.

I’m more than happy to host new projects as long as you choose things we offer and I’m also very happy if you attribute code hosted somewhere else to netlabs.org. There is a fair chance netlabs.org will get a re-release the day our semantic web software is available and this might again change the way we collaboratively work on software. But for the moment GitHub and all the other solutions win and you should use them!

One last remark: While we build a lot of stuff around open source software I get the impression that in some cases changes in the code (aka patches) do not get distributed with released binaries. This is not really smart, open source software is about sharing code and at least for GPL based software it is mandatory to release modified code. But even if you are using BSD or Apache licensed code we should share our code. The OS/2 community is definitely too small to waste time on searching recent patches for compiling it on our platform. So if you keep some stuff for you, please release it with the binaries!

Leave a Reply