From owner-freebsd-hackers Thu Apr 20 09:51:13 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id JAA02204 for hackers-outgoing; Thu, 20 Apr 1995 09:51:13 -0700 Received: from nomad.osmre.gov (nomad.osmre.gov [192.243.129.244]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id JAA02198 for ; Thu, 20 Apr 1995 09:51:09 -0700 Received: (from gfoster@localhost) by nomad.osmre.gov (8.6.8/8.6.6) id MAA18678; Thu, 20 Apr 1995 12:47:13 -0400 Date: Thu, 20 Apr 1995 12:47:13 -0400 From: Glen Foster Message-Id: <199504201647.MAA18678@nomad.osmre.gov> To: freebsd-hackers@FreeBSD.org Subject: Release stability Sender: hackers-owner@FreeBSD.org Precedence: bulk Features are nice to have but the measure of any operating system is how dependable it is. In my book features account for about 20% and dependability (in which I include both predictability and stability) about 80% of the "quality" of a release. Please, please, please do not compromise the quality of 2.1. If I remember correctly, the CSRG releases were "odd number == features added, even number == stable versions of the odd number" (actually, maybe it was the other way around :-). This scheme had a lot going for it from the consumer's POV. I suspect it also moderated the expectations of the developers, at least during the bug-fix release cycle. Has (re-)adopting this philosophy been considered by the core group? Glen Foster