Code Ownership - Individual or Collective? (was Re: Refactoring Requirements)
- --- In firstname.lastname@example.org, Brad Appleton <brad@b...>
> One thing that always struck me as a bit "out of sorts" wasI take this as a place where XP has consciously taken a different
>an area where there didn't seem to be "alignment" between the
>org-patterns and XP, and that is on the issue of code-ownership.
>Cope's org-patterns clearly called out code-ownership as a
>pattern of hyperproductive software development organizations
path, as a sort of social experiment. In a code ownership
environment, you're building in delays and handoffs ("could you add
this for me once you finish what you're working on?"). There's a
theorem in queueing theory that if you have specialized resources
you'll be less responsive than if you have uniform resources. Shared
ownership tries to leverage that.
Of course, translating back to people, we're not uniform resources.
XP can in a sense pretend to have uniformly competent people by
pairing, standups, judicious task selection, expert-in-earshot (one
room), and other communication efforts. (But note that Cope's
patterns use some of these things too, e.g., pairing.)
>(tho I don't recall if it was exclusive code-owner access orThe version in PLOP 1 says "Each code module in the system is owned
>something more like a code "steward" for individual
by a single Developer. Except in unusual, explicit circumstances,
code may be modified only by its owner."
Bill Wake William.Wake@... www.xp123.com