Releasing Open Source

time to read 2 min | 266 words

Chad Myers asks in the ALT.Net mailing list:

Why are OSS folks to scared of calling something 'released'? I'm guessing that MonoRail's "RC" code is 10x better than the released version of most corporate software out there.

While I am not going to argue about the quality of the Castle code base, there is quite a bit more to the producing a release than just having good code. No, being scared of bugs is decidedly not the reason.

Of the top of my head, in order to get a good release out the door you need to compile a release guide, documentation, upgrade guide, installer, wizards and much more. All of this stuff is (a) boring, (b) time consuming. 

Throwing code over the wall and slapping a x.0 on it is not good enough.Commercial software has an advantage here because they actually pay people to do this. OSS has a problem because very few people enjoy writing the documentation. I am certainly included. Certainly not on my free time.

For Castle, Hammett is the one that has to carry that load, most of the time. For NHibernate, it was Sergey that did most of the actual preparation for release.

I, personally, have had very little to with prepping the releases. Actually, I feel slightly bad about that.

Anyway, if you want to see releases from OSS projects, the best way to ensure that is to contribute. It can be with time, helping build the documentation, release guides and all the other necessary stuff, or it can be with monetary donation, going to the same purpose.