Jeffrey 'jf' Lim wrote:
> On Wed, Sep 9, 2009 at 11:54 AM, Nick Holland
> <n...@holland-consulting.net>wrote:
>
> * Commitment to doing it right in one way, not twenty different
>> ways ("pick one, maybe you get lucky").
>>
>
> just one question - how do we determine this "one right way"? (like there
> could be a multitude of ways to do the same thing)
First of all, it isn't "we" who determine the "one right way", it is
the developers, and if there is a conflict, it's Theo. (and yes,
internal conflicts take place, and if you think the "abuse" on the
lists is vicious, you don't want to see some of the internal disputes!)
it's a lot simpler than it might seem.
The process varies, but usually starts by someone writing code (and man
pages). If the code is good, it is improved, if it sucks but solves a
problem, it may inspire someone else to write good code. "Talking about
the idea" is not part of the OpenBSD development process.
But often in life as in OpenBSD, the key to making a good decision is
simply MAKING A DECISION, then you worry about making it the RIGHT
decision. This is a problem you can see in almost any committee designed
system, compromise takes place to try to keep everyone happy-ish, but no
one is completely happy, and the result is usually far from ideal. "A"
could work, "B" could work, "A+B" and "neither" suck. Most committees end
up going for the "A+B" or "neither" option for fear of pissing off the A or
B camps.
Simply making a hard decision quickly and focusing all efforts behind that
decision produces better results than compromises that split development
efforts and drag out problems produces better results.
Nick.
