On Sun, Sep 19, 1999 at 12:45:10AM -0500, Chris Lawrence wrote: > My point is simply that changing policy does not translate into making > things happen by fiat. Perhaps you missed the entire point of my > mail, because it seems like we agree with each other.
Actually... rereading your original post... You seem to be saying that issueing new policy doesn't prevent a person from issuing a package under an old policy and that therefore there's no advantage in writing policy so that the new policy still accepts the old practice. I'm suggesting that individual policy changes are not necessarily related, and that some new, relatively untested policies should have a longer release cycle than other new policies. The "you can release your package under an older policy version" mechanism should be a last resort. It's a very crude mechanism. Please tell me if I misread your post, but it looks like you were suggesting that we have a series of mutually contradictory policy versions and that the developer must pick one appropriate to his needs, and that at some point older policy versions would become obsolete with no real transition period. -- Raul
