Date: Tue, 24 Apr 2001 07:26:54 CDT From: Dave Leimbach <dleimbac@earthlink.net> To: "Albert D. Cahalan" <acahalan@cs.uml.edu> Cc: questions@freebsd.org Subject: Re: How Is The FeeBSD OS Like and Different Than Say Redhat or Suse LINUX Message-ID: <200104241225.FAA03888@scaup.mail.pas.earthlink.net>
next in thread | raw e-mail | index | archive | help
Well obviously we can agree to disagree but..... On Tue, 24 Apr 2001, Albert D. Cahalan wrote: > Date: Tue, 24 Apr 2001 00:49:30 -0400 (EDT) > To: dleimbac@earthlink.net > From: "Albert D. Cahalan" <acahalan@cs.uml.edu> > Subject: Re: How Is The FeeBSD OS Like and Different Than Say Redhat or > Suse LINUX > Ever heard of "designed by committee"? How about "too many cooks..."? > Nobody but Linus himself is _completely_ happy with the choices he > makes, but at least we can be sure that the kernel will be the result > of a fairly coherent vision. I don't think FBSD suffers from the too many cooks problem. The releases are stable and coherant. The kernel is fairly easy to read source-wise. (Probably as far as kernel's go). The reason committee works better sometimes is that you have a better chance NOT to make the mistake that Linus did in the 2.0.x kernels of designing a kernel that only allows one CPU to be in kernel data structures at a time which thusly thwarts SMP. 2.4.x is different but most of the changes that a pro-SMP are in the network stack. If you want to look at too many cooks you should see that there are many linux distributions all with different library sets. Ever run something that was "RedHat Linux only?". I have and I think its stupid. > > > Look at PPC linux today. Its way behind in the > > mainstream kernel cuz Linux didn't like it. > > No, this is not why PPC support is behind. It is behind for > two reasons: > > 1. The developers are lazy about getting patches in to Linus. > Linus expects patches to be delivered in logical chunks. > Linus will refuse a big blob of CVS diff output, because > this makes it difficult to examine the changes for problems. > > 2. Partly because of reason #1, the PPC developers missed the > 2.4.xx deadline. Linus then refused to destabilize the 2.4.xx > kernel with large patches right around release. Linus also > refuses to let the early 2.4.xx kernels suffer this fate. > It's a good thing that Linus doesn't want to destabilize the stable 2.4.x kernel like the 2.2.x kernel releases did. They were a nightmare at the beginning. He said so himself. > > Also GGI was a great idea that > > was all but killed due to combinations of things including being told > it > > would never make it to the kernel by Linus. > > Yeah, this sucks. Be glad, because this would have killed BSD. > Without Linux, there wouldn't be XFree86. > Linux being more popular than FreeBSD you are probably correct. You should know that I don't hate linux. I keep it on my machine because I get paid to work on it at work. I am fairly familiar with the kernel innards. (Well moreso than a complete kernel newbie). There are many people who used to like linux more than they do now. > > Also I want to eventually start using some IPV6 stuff and linux, even > > though it supports it or claims to in the kernel config, doesn't have > as > > cool a networking structure so I figured FBSD would be the way to learn > how > > to use IPV6. > > What is so cool about BSD networking structure? Well I am learning about this stuff now but since FreeBSD is a direct decendant of the original BSD. And BSD was the first Unix and the first OS anywhere to have TCP/IP who do you think most likely has the most mature and sophistocated methods of dealing with it? FreeBSD 4.2 still beats linux in every network benchmark except email handling... <don't know why there.... :)> even in the 2.4.x kernels. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200104241225.FAA03888>