From owner-freebsd-stable Sat Jan 8 12:17:36 2000 Delivered-To: freebsd-stable@freebsd.org Received: from mail1.gmx.net (mail1.gmx.net [194.221.183.61]) by hub.freebsd.org (Postfix) with SMTP id AD8DB157F0 for ; Sat, 8 Jan 2000 12:16:46 -0800 (PST) (envelope-from Gerhard.Sittig@gmx.net) Received: (qmail 29180 invoked by uid 0); 8 Jan 2000 20:16:35 -0000 Received: from p3e9d514e.dip.t-dialin.net (HELO speedy.gsinet) (62.157.81.78) by mail1.gmx.net with SMTP; 8 Jan 2000 20:16:35 -0000 Received: (from sittig@localhost) by speedy.gsinet (8.8.8/8.8.8) id SAA00146 for freebsd-stable@freebsd.org; Sat, 8 Jan 2000 18:04:53 +0100 Date: Sat, 8 Jan 2000 18:04:53 +0100 From: Gerhard Sittig To: freebsd-stable@freebsd.org Subject: Re: 4.4 BSD forever? Message-ID: <20000108180453.L31561@speedy.gsinet> Mail-Followup-To: freebsd-stable@freebsd.org References: <010501bf5989$2c4b7ec0$0200000a@danco.home> <3876D48E.8046C35@wcnet.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0i In-Reply-To: <3876D48E.8046C35@wcnet.net>; from jestess@wcnet.net on Sat, Jan 08, 2000 at 12:09:18AM -0600 Organization: System Defenestrators Inc. Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Sat, Jan 08, 2000 at 00:09 -0600, John Estess wrote: > > But what does the version number really mean? Very little. In a > world where 3.1 precedes to 95 (MS), where 4 precedes 7 > (Slackware), where original releases start with 5 or 6 > (Mandrake and Suse, I believe - DQM), what do you think release > numbers signify? Not much. I feel I have to disagree here (besides the fact that there *are* SuSE releases prior to 5.0, and Mandrake might use this numbering to reflect the RedHat releases they're derived from). If others are unable to count, I DON'T have to do so too just because of THEIR lack of basic knowledge. Otherwise we instantly had to jump on ANY train when "almost everyone else" is on it already (i.e. make any free software more behave like the MS idol and act on commercial and marketing rules instead of doing technologically based decisions). That's something which would be flamed down in the blink of an eye as soon as someone would suggest it here in the list, I'm sure. > I'm glad FreeBSD has the integrity to sequence their releases > in a logical manner, but this isn't required, is somewhat > unexpected, and is really boring. Regarding release numbers I WANT to have some orientation, so a "boring" scheme is helpful and exactly what I would EXPECT and admire if found! It's not just about "naming it different". Solaris marketing people did (breaking the sequence 2.5 .. 2.6 .. 7 .. 8) and confused updating and security mechanisms (the naming pattern did't apply any longer). This caused real trouble to the admins. I don't see any benefit in breaking functionality just to attract some uneducated people with shiny names since they cannot even count in little numbers' domains. Keeping this in mind I could live with leaving out lower versioning numbers (e.g. starting with major number 4 since customers cannot imagine that "foo 6" is not more developed than "bar 1") or hopping to "catch up" with the competitors you're being compared to. But this always leaves a dubious impression behind once you THINK about it. It turns out to be a question of "Whom are we targetting for?". Do we want to attract those who take a shiny name or four digit number for more than the features and development a product has undergone? Or do we focus on the ones who can handle real info? > As far as designating a release, a name is only a name, whether > it be a sequence number (4.4) or a real moniker (Reno). People > will tend to know the current (usual meaning) distribution in > either case, and that is what matters. Sorry, but there's more to it than "just a name". You have to have an identifier and when it's not the release number you suddenly have two of them ("name" and "number") which raises complexity for NO added information or functionality. You have to provide this information to the application software as well (think of uname(2)) and how many of these programs can cope with nonnumerical input? Someone remembers the OS/2 DOS emulation (labelled "DOS 20.0") which made many programs tell you "You need at least DOS version 3.0"? And once you have to follow more than just one "versioning scheme driven by marketeers" (admin a heterogenous network) or if you are new to a certain environment, how do you tell one version from the other even before you come to attack your real problem? In Linux-land there are already problems to make newbies realize that kernel 2.2.12 is more recent than 2.2.5. :) Why should anyone try and add even more fuzz to it instead of establishing a pattern to name them like "2.2.05" (or use even three digits for the -current line). And the same points hold true to BSD IMHO. This discussion reminds me in some way of the O'Reilly books and how they are referred to. When someone calls it "THIS ANIMAL's book" everyone not having it in the shelf has to ask "What is it about?" causing one more cycle and lots of redundancy in a discussion to fetch information which could have been provided in the first place. Not everyone might be as familiar with certain associations as you could be. virtually yours 82D1 9B9C 01DC 4FB4 D7B4 61BE 3F49 4F77 72DE DA76 Gerhard Sittig mail -s "get gpg key" Gerhard.Sittig@gmx.net < /dev/null -- If you don't understand or are scared by any of the above ask your parents or an adult to help you. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message