Interesting Links – September 2012

What has interested me in the world of the internet in the last month?

Lets start with some release announcements:

  • Linux From Scratch 7.2 was announced.
  • Python-3.3 was released.
  • The first alpha of Akabei (Chakra’s pacman replacement) was announced.
  • Not exactly a “release” announcement, but Clang will be the default compiler on FreeBSD 10.0.

A good rant or two is always entertaining:

  • Thomas (Arch dev) on systemd.
  • Anti /tmp on tmpfs.
  • An Arch Bang moderator has discovered I can be… let say opinionated… about incompetency.
  • And this was just, well…

And some random links:

  • Some posts about dealing with misbehaviour on the Debian developmental lists.
  • The Open Build Service can now create Arch Linux packages.
  • A warning about automatic updates in Fedora.
  • An interesting approach to teaching programming (specifically Python) online.
  • The first commit I pushed to the glibc git repo.
  • Discussion about a format to replace tar. I want the ability to easily remove/replace files…

Finally, I do genetics too…

  • The ENCODE project gave us a major insight into just how complex the human genome is.

6 thoughts on “Interesting Links – September 2012

  1. The way you treats Arch Bang developers and their users are still unnecessary rude, you know Allen? Technical issues aside, do they deserve that sort of comments/sarcastic remarks just because they are not as technically competent as you or the devs?

    I kinda expected you to at least acknowledge that you went to far with your remarks in that last blog post. Guess I really didn’t know you all that well (and I thought you would be a better guy than this).

    I kinda like systemd, but this whole systemd fiasco and especially how Arch devs handle the situation have left me pretty disillusioned with the whole Arch “community”.

    • I think I have be rather polite given all the Arch Bang team seem to do is complain on their forums about the decisions the Arch Linux team make and never actually contribute anything during the decision making. In fact, the amount they contribute back towards Arch Linux itself is practically zero.

      I even went to their bug tracker to see if there were issues reported there that should be forwarded to the Arch Linux bug tracker. Their bug tracker has been filled with spam for several months, which really did not alter my opinion on the competency at all…

  2. Had a shit time on Arch forums, but cant blame Allen too much,
    First he’s got a kickass avatar,
    Second the mans got taste,
    Third he got a commit to glibc,
    Fourth i used his sequence to build up my toolchains, turned out to be better than gentoo,s or LFS sequence,
    Finally im sure he wasnt the M^%&^%F who defaced my github page.

    Just putting a final post in to say did the no-sytemd modifications as said would, all 60 odd PKGBUILDS, everyone with sysd carefully removed from the compilation, the writing of conf files, and where the original tmpfiles etc dir is created ( in glibc package really ?? )
    In the process i have fixed up 11 packages from ABS/ Arch SVN ( i always check both cos they often are different ) with gentoo and other patches ( specifically a few of the gnome packages which no longer compile against cups 1.6 , few others with crap url’s , and old md5′s ) created a bunch of new ones and generally done a good housekeeping for you. ( since for the optimization script i ended up building EVERY pkgbuild in ABS, some of the authors sites were down, busy etc so ended up using debian mirrors of the sources, wonder why you dont ? also provided proper –ALLSOURCE packages)
    I have delivered exactly as i said i would, and thats my word kept, now totally defeated with my other goals im off back to my own field where I rule the roost :)
    This is prob my last post in the “linux” world, so prob last time you’ll hear from me, so have a good one !! and cheers.

    • Hey Jubei,

      I never saw the comment on your github, but I did see your reply (which is still in your public timeline…). So I won’t comment on that beyond noting the big “fork” button creates a working copy of your repo which is how people can submit pull requests to you. So “copying” is not quite right.

      Anyway, it is good to hear my toolchain build order works. Although I am fairly sure its is almost the same as the LFS build order as that is where I did my learning.

      We do not focus on keeping packages building from source, so some of the PKGBUILDs can become outdated with time. I try and make sure all of [core] builds from trunk on our SVN. We also host sources for many packages in the source directory on the mirrors, although they are not used in the PKGBUILDs.

      Also, glibc does not create the tmpfiles directory per se… But it does put a file there for the nscd daemon.

      And finally, I am going to give some advise because I am like that! Hope you do not get offended as that is not the intention. If you are still continuing with this project (you say it is dropped here, but it seems active on the Arch Bang forums), I would suggest taking a big step back. Start with the initial goal – have a repo that allows an Arch install to be free of systemd – and complete that. Building everything for i7 immediately alienated the majority of people that were interested in this. So, rebuild glibc to get rid of the systemd files for nscd, but don’t rebuild the whole toolchain. Leave replacing gvfs with some thing from Puppy Linux until later. In fact, most rebuilds just to remove service files are largely unnecessary for the start.

      I’d focus on just getting the non-systemd boot-up working with whatever udev you decide on. Nothing more. When that is working and continually maintained, you will build a community around the project which will allow larger goals to be tackled.