Slashdot Mirror


NetBSD in 2003 - Annual NetBSD Status Report

jschauma writes "On February 7th, The NetBSD Foundation held it's annual meeting, during which the developers discussed, among other things, how NetBSD progressed over the last year and what things are planned for the comming year. The Annual NetBSD Status Report summarizes this meeting and provides an overview of past, present and future of the NetBSD Project, the NetBSD operating system, pkgsrc and the NetBSD Foundation both in general and from the perspective of each group, to give users and people interested in NetBSD insight into the project. Please join our mailing lists for participating in ongoing discussion, and see our web site for more information about the NetBSD project, http://www.NetBSD.org."

5 of 49 comments (clear)

  1. good report by H4x0r+Jim+Duggan · · Score: 5, Interesting

    It's a pity such annual reports aren't more common among free software projects.

    One funny thing is that the pkgsrc section mentions a patch in the GNATS database to allow pkgsrc to run on GNU Hurd systems! (it already runs on GNU/Linux)

    I doubt any established distros are going to ditch their existing package systems, but if a new distro was to begin (as a non-Debian testbed for GNU/Hurd) - it would be worthwhile evaluating the *BSD package systems.

    1. Re:good report by Anonymous Coward · · Score: 5, Informative

      Well, Bluewall Linux uses NetBSD pkgsrc; though I don't understand why anyone would use it when there is the real thing.

    2. Re:good report by kivaapina · · Score: 5, Funny

      The reason why such annual reports aren't more common is because they occur once a year. If they were to become more common they would cease to be annual.

  2. Linux isn't a company by boobsea · · Score: 5, Informative

    Talk about specific Linux development companies instead.

    Debian might fall into this category, but I wouldn't know about others

  3. Re:NetBSD Status Report by Anonymous Coward · · Score: 5, Interesting

    Portable code tends to be rather clean code.
    Clean code tends to run better and have fewer bugs.

    Yep. One of the more interesting ones to me was the new RC system, which when implemented on Vax was very noticably slower than the old one. I remember reading the whole thread on ways to resolve it. No flame wars, no little ranting "children", just a solid technical discussion on ways to fix it.

    It points out things when something that maybe adds .1 seconds to your startup on a fast x86 box adds *minutes* to the startup on your old Vax. Things like that might never have been noticed without it being so portable, and it leads to better code/performance for *all* platforms.

    Plus, having the same PCI card driver on your PC and your PPC Mac makes for easier fixes. Find a bug, fix it in one place, and all PCI platforms are fixed.

    I run NetBSD everywhere I can. The people on the boards are generally always very "professional", I don't see people getting flamed for asking "stupid" questions, and in general I see a lot of thought going into the *design* before things get implemented. I always find the Linux approach of "just start coding, figure it out as you go" to be a recipe for bad things... like a VM change in the middle of the 2.4 chain. Sorry, I want risk, I'll run -current. I want stability, I run a normal release. Never gotten bit on NetBSD yet.