Interesting Links – August 2012

When preparing to move my Firefox profile to a new computer I noticed that I collect a lot of bookmarks that I do nothing with. Instead, I have decided to make a monthly post of the “interesting” stuff I collect and so clean them out on a regular basis.

  • The GCC cxx-conversion branch was merged. That means that GCC is now compiled as C++.
  • The Hijinks Ensue podcast is back.
  • The Chakra Linux distribution is dropping their i686 support.
  • A forum post be Tom Gundersen explaining why the move towards systemd in Arch is a good idea.
  • Blog post about progress on the Cupt package manager. No features I immediately want to implement in pacman…
  • The Slackware documentation project is underway.
  • A systemd-to-sysvinit converter is being developed in a Summer of Code 2012 project.
  • How does Arch Linux compare to openSuse Tumbleweed? I do not know because I can not access this article (hint, hint…)
  • Found this Pixies cover via reddit. Had to watch Fight Club again…
  • The results of Gentoo offering a bounty to fix a bug. This reminded me of Arch Bounty, which failed to even get a single project up.
  • Despite this blog post, I understand even less about the version numbers for packages in Debian!
  • A project to write a tutorial on x86_64 assembly.

10 thoughts on “Interesting Links – August 2012

  1. > The Slaceware documentation project is underway.
    ‘Slackware’.

  2. “The Charka Linux distribution is dropping their i686 support.”

    Hope Arch won’t follow its example. ^_^

    • @Cippaciong: Could you tell me how can you call Arch linux a bleeding edge distro when supporting i686? The x64 instruction set has been with us from late 2003 (AMD) & 2004 (Intel). It has been 8 years since then. It will become increasingly hard to develop and properly test i686 when nobody from the developers use it anymore (at least my guess). I would think 10 years should be the maximum age of supporting one architecture – if there is replacing one, is established, and most of the people are using it anyways.

      If you want to have support for older computers you should probably move to distribution for older computers (puppy, wattOS etc. – dedicated to older computers). You can’t have support for older architectures indefinitely for rolling updates distro anyways.

      • > The x64 instruction set has been with us from late 2003 (AMD) & 2004 (Intel)

        > I would think 10 years should be the maximum age of supporting one architecture

        You mean we should drop x64 in a year or two?

        • Please do not copy out only parts that suits you. I will add the part that is missing here -> “and most of the people are using it anyways.”

          If there is a major architecture which majority (lets say more approx. 80%) of people use, why not. I don’t see it happening in year or two.

      • @Patrik : Let’s say ok for the 10 years delay before considering something is “obsolescent” : my i686 laptop was bought new on 2007, only five years ago, so I would be expecting its architecture to be supported until 2017 at least. Arch Linux is working like a charm on it, and it is able to run the latest power-hungry sofrware and desktop environment.

        The “bleeding edge” thing is for the software, not the hardware. It is a bad idea for a big distribution to say that 5 years old computers are already too old.

        • @Vincent:
          The most important part of my message was the percentage of people using it – “..majority (lets say more approx. 80%)..” && “..hard to develop and properly test i686 when nobody from the developers use it anymore ..”. a) I think it is nonsense that 80% of users are using suboptimal system because of 20% of people. b) Nobody who is actually developing && testing is actually using it – then these 20% get untested system which in the end is much worse then it could be with specialized system for older computers – users, developers && testers are using the same platform.

          quote: “The “bleeding edge” thing is for the software, not the hardware. It is a bad idea for a big distribution to say that 5 years old computers are already too old”.

          The software – hardware gap can be only so big. After some time you start to limit the newer hardware because you want to keep supporting the older one.

          I wonder what kind of cpu you have. If you had brand new cpu in 2007. When I double checked the list on year 2007 most of the cpus were already supporting 64bit to some extent.

  3. looking at the preview for that issue of Linux Format, I wouldn’t bother. Not sure if the audience is the mainstream computer user, or what, but in the first round they knock out Arch in favor of OpenSUSE Tumbleweed, and at the end hand the ‘crown’ to Mint.