Still the best quote I have about how to move an idea forward among a group of Open Source kernel developers. Pure gold.
Back on OpenSolaris years ago we had a lot of community conversations and flame wars that started out with the “Why don’t you just …” construction and then things spun around from there. At various points (usually after hundreds of such mails) the kernel engineers would occasionally chime in with a precise gem like this one below. Get to work. Write something.
[ogb-discuss] Re: [osol-discuss] Re: Project Proposal - (what is/was Indiana)
Bryan Cantrill bmc at eng.sun.com
Thu May 31 15:41:47 PDT 2007
Previous message: [ogb-discuss] obtaining a contributor grant
> More source code and less comments/opinions!
Yes, please. I don't think one can necessarily fault the OpenSolaris community for being bureaucratic; if a project wishes to lead with the process and follow with the prototype, life will naturally feel very bureaucratic. I (rather strongly) believe that this is _not_ the way that software should be developed; in software, ideas are expressed in _code_ -- the implementation _is_ the idea. If an idea has not been implemented, it is (in my opinion) premature to call it an "idea" -- it is rather a notion, a daydream or perhaps a fancy. If one has such a thought, the first order of business is not to send out proposals or give speeches or navigate through process but rather to sit down and write some code: build a prototype, share it with some like-minded people, incorporate their feedback, expand the community (note, lowercase "c") and iterate. If one does this and one's ideas are sound, the process (in my experience) naturally follows; people naturally gravitate to good ideas.
This is not to say that the OpenSolaris processes can't be used to help this process of iteration -- just that they should be used to help the process, not initiate it.
- Bryan