Subj : test To : Roy J. Tellason From : Lawrence Garvin Date : Fri Jun 06 2003 01:40 pm Roy wrote to Lawrence at 04:06 26 May: RJT> I'm wondering, when would I want to go with something like this RJT> as opposed to say, sticking with the Slackware that I'm running RJT> on the other box and am already fairly familiar with? Or maybe RJT> when would I want to go with some other flavor of *BSD, or even RJT> some other unix? LG> Roy, that's really a matter of preference. RJT> Not a matter of this or that is better at certain things? I can't say that one or the other isn't better at one thing or another. However, I also don't have any immediate knowledge that one or the other /is/ better at one thing or another. I'd be entirely surprised if each flavor didn't have something or other that it excelled in. But as to what that actually is, I'm still a few research projects away from that personal knowledge. :-) LG> My choice for going with the *BSD product is because the BSD source LG> tree has been around forever, it's a true Unix, it's derived LG> directly from the original AT&T Unix, and it's very easy to LG> install. RJT> I've not had any real hassles installing any of them, but am not RJT> sure of the reltative weight of the rest of those consideration. RJT> Why would something being "a true unix" be a plus? Again, it's mainly a matter of bigotry. The advantage to being a "true unix" is that the code used to write the kernel and utilities is mature, and well tested. While I certainly don't mean to malign any Linux code or programmers, it is new, relatively speaking, and sometimes it takes several years before some of the more obscure issues with a product come around to see the light of day. Another immediate issue, though I really don't think one of significance, is the SCO Group's fight-picking with IBM over alleged impropriety with the code in the Linux kernel. It's enough material for a whole new thread, but the short answer I can give is that the SCO Group has stepped in a big pile of it, and it'll take 'em months to extricate themselves from their little power play. Unfortunately, there's always the possibility that IBM /did/ convert some AIX code into the Linux kernel, and if so -- if that really did happen, I'd want to be as far away from a Linux installation as possible until I see exactly who the SCO Group is going to aim their poison darts at after IBM. RJT> I've not gotten the impression that there are really that much in RJT> the way of differences between them. As far as /content/ I would agree, they're all probably pretty much the same. Where they differ is in areas such as: a. Installation methods b. Customer support c. Price, packaging, sales channels, and delivery methods. Very few of the Linux distributions available today are entirely "without cost" to the end user. Many of them are only available through retail or direct sale channels. All three BSD products, though, are available absolutely 100% free-of-charge -- though a couple of them do sell media kits for a nominal fee. RJT> I remember reading somewhere that NetBSD (?) was supposed to be RJT> better for things like firewalls, though I can't recall why. RJT> Dunno what the differences are among these three, anyhow. The fundamental different between NetBSD and the others, is that NetBSD's "headquarters" is in Canada, and thus was able to avoid ALL of the U.S. imposed export restrictions on cryptography. Beyond that, a lot of the "security" code, other than cryptography, contained in NetBSD may actually have either existed in the original BSD4.3 code, or was written by the FreeBSD group. Perhaps one item to note of the three BSD groups is their familial associations. Both OpenBSD and NetBSD came about much more recently than FreeBSD, but all three of them are direct descendants of the BSD4.3 commercial product from the 1980s. RJT> I suspect that they're all good products, and that there might be RJT> choices as to which would be better in certain applications... Exactly. I agree. --- * Origin: lawrence@eforest.net | The Enchanted Forest (1:106/6018) .