Date: Mon, 01 Mar 1999 01:15:11 -0800 From: "Jordan K. Hubbard" <jkh@zippy.cdrom.com> To: "Robert A. Bruce" <rab@pike.cdrom.com> Cc: Dave Yost <Dave@Yost.com>, freebsd-advocacy@FreeBSD.ORG Subject: Re: The Linux PR firestorm disaster (w.r.t. FreeBSD) Message-ID: <32820.920279711@zippy.cdrom.com> In-Reply-To: Your message of "Sun, 28 Feb 1999 16:58:42 PST." <199903010058.QAA24952@pike.cdrom.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> What does FreeBSD have that Linux doesn't? I don't think that such > a table would be very "forbidding". It's not so much that it's forbidding so much as highly angle-of-view dependent. Once you set out to work on enunciating what you find so attractive about FreeBSD from your particular perspective, you quickly see how much of it really is just a matter of personal preference and a "feel" to the OS that's also very difficult to describe in textual form to someone else. It's a lot like one's preference for automobiles of a certain manufacture. If you'd driven 4 different Mazdas and 4 different Toyotas in your lifetime and the various Mazdas just always impressed you in some set of ways as being "better" than the Toyotas, then one would probably understand your having some preference for Mazdas as a general rule - you subjected 8 different models to your selection criteria and the Mazdas "won." Get into a Toyota mailing list and say "The Mazda is a better car", however, and you'll very quickly find out that a lot of your selection criteria are not shared by the members of that list, or perhaps they share your selection criteria but draw completely different conclusions from it; either way it's a very subjective thing. So it is with FreeBSD and Linux. I've installed and looked at both (one would hope :) and I have a penchant for certain organizational methodologies in an OS which FreeBSD simply comes closer to providing. I can install the base bits and get a /usr/src tree containing the source for every piece of the OS I might ever want to build (or even just INSPECT), available right there in the same form that was used to build the binaries I'm running in case I want to reproduce the same results. I don't need to learn multiple arcane autoconf/configure/whatever build systems to build each separate component, it all just fits together under the Berkeley make system. Even stuff which is wildly outside "the BSD way" gets forced to toe the line through the ports collection, a mechanism which at least makes sure that things don't attempt to pollute things directly in my "system" directories like /bin, /usr/bin, etc. If I then wish to syncronize this source (or ports) tree with the official project version, there are several commands provided (CTM, anoncvs, CVSup) which will also allow me to track the changes to this whole pile of bits on an ongoing basis, public mailing lists also giving me the option of actually seeing the change log messages as things in my tree get modified. These may seem like a small things, but when you're trying to figure out just what parts of a system are supposed to be there and which were just added by the trojan horse weenies who broke in at 2am, the more organized your binaries are the better. If you've got nice, organized build systems on tap, you can also rebuild and replace all those binaries in a systematic fashion and if the problem involves a security problem which has since been fixed by FreeBSD.org, you can run your usual source-syncronization command to catch up with these latest changes. No hand-applied patches to figure out, no special build procedures to follow, just cvsup, build the world and call me in the morning. I guess if I had to encapsulate this in a sound bite, I'd say that FreeBSD was simple more developer friendly. We have a well-organized source tree, a publically exported CVS repository, lots of fancy web tools for querying or indexing our sources (http://www.freebsd.org/cgi/cvsweb.cgi and http://lxr.linux.no/freebsd/source to cite two examples), we simply have the basic MIND SET that developers should have all the best tools available for developing and managing the OS, that just happening to suit the needs of a certain kind of user rather well also. I think a lot of Linux folks would *like* that sort of approach but I don't see any of the big distributions putting as much work into making the OS developer friendly as they are into adding desktop gee-gaws and such for the traditional Windows user. Then we get into "stability", another one of those words which is very much viewpoint dependent. I'll go out on a limb here and say that any OS which makes it easy for developers to fix problems quickly and for users to track those fixes in an organized fashion (not just "here are 20 patches on my FTP site! Come and get 'em!") is going to be the "more stable" OS in the long run. I think that also applies to any OS which is run by inherently conservative people who, in many cases, are old and wizened and a little closer in temperment to Yoda than Luke Skywalker. Luke might be flashier and get all the girls, but he makes more mistakes. :-) And no, I don't want "FreeBSD is Yoda, Linux is Luke Skywalker!" on this comparison - it was merely another analogy. :-) Like most things, I think we're not going to "win" so much here by attempting to score points off Linux as we are by enumerating our strengths. We run Yahoo, Hotmail, Link exchange, etc. and are very popular among ISPs. We have a long history. We pull our socks up. We're bad-ass. And so on... - Jordan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?32820.920279711>