On Sat, Jul 23, 2011 at 5:59 PM, Ross Gardler <rgard...@opendirective.com> wrote: > On 23 July 2011 22:50, Shane Curcuru <a...@shanecurcuru.org> wrote: >> If people find this work interesting, that's great. But in terms of rules >> and procedures, sometimes it's fine to not over-document the rules until >> there's a case where they're really needed. > > +1000 (I'm referring to the general concept not any specific > discussion about potential rules) > > I don't mean to say that the establishment of rules is a pointless > exercise. However, for the majority of software developers rules are > at best a diversion, at worst restrictive. >
For the vast majority of software developers, unit tests and coding standards are also a diversion, and at worst restrictive. So what? The PPMC has oversight responsibility for the project. This is addition to roles we might individually also have as software developers. Written rules are tools, and can be a useful one at that. ASF certainly finds them to be useful on occasion. So have other Apache projects. > In general the vast majority of situations you will come up against > (or at least can imagine coming up against) will already have been > experienced somewhere in the ASF at some point. When trying to figure > out how to handle a given situation there will be someone who can help > guide the decision making. > Great. I'd like to think that a good rule is based on experience and not just made up out of thin air. But as we've seen from my hypothetical question about when a committer becomes a committer, different people have different experiences and different opinions. So this PPMC needs to chart its own course on this, synthesizing what wisdom we can collect from others' experience. But once we've done that, I can think of no better way of ensuring that such wisdom is applied consistently going forward than writing it down as a rule. > Remember an ASF project is about consensus, not about rules and > regulations. It's about the community deciding what is best for the > community as a whole at that point in time. There is no shortage of > people to help you build consensus by providing options. Even when you > graduate you will be able to call on the experience of everyone here > and everyone in the ASF as a whole. > Yes, yes, yes. But a rule is one way in which consensus can be recorded. It is not different than consensus. It is about making that consensus be transparent, reusable and consistent, so project members (and non project members) can have reasonable expectations about future actions and decisions. This has been true since Runnymede and King John. It is not sufficient to have consensus. We also need to be seen as being fair, consistent and predictable. This is important for members as well as non-members. A capricious consensus helps no one. -Rob > Ross >
