Become a fan of Slashdot on Facebook


Forgot your password?
Operating Systems Software Linux

Linux Kernel 2.6.29 Released 265 writes "Linus Torvalds has released Linux 2.6.29. The new features include the inclusion of kernel graphic modesetting, WiMAX, access point Wi-Fi support, inclusion of squashfs and a preliminary version of btrfs, a more scalable version of RCU, eCryptfs filename encryption, ext4 no journal mode, OCFS2 metadata checksums, improvements to the memory controller, support for filesystem freeze, and other features. Here is the full list of changes."
This discussion has been archived. No new comments can be posted.

Linux Kernel 2.6.29 Released

Comments Filter:
  • by Anonymous Coward on Monday March 23, 2009 @10:18PM (#27307105)

    I just want the government to leave me alone. I feel like the police watch me. I pay all my taxes, I've never been violent, I try to be good, I help others, and I feel the lasers eyes. I feel like I'm being hunted. I feel watched. I feel like the government is coming for me. I think of this song:

    Johnnys in america, low-techs at the; Wheel; No-one needs anyone, they dont even; Just pretend; Johnnys in America

    Im afraid of americans; Im afraid of the world; Im afraid I cant help it; Im afraid I cant; Johnnys in America

    Johnny wants a brain, johnny wants to; Suck on a coke; Johnny wants a woman, johnny wants; To think of a joke; Johnnys in america

    Im afraid of americans; Im afraid of the world; Im afraid I cant help it; Im afraid I cant; Johnnys in america;

    Johnnys in america, johnny looks up at; The stars; Johnny combs his hair and johnny; Wants pussy and cars; Johnnys in america

    Im afraid of americans; Im afraid of the world; Im afraid I cant help it; Im afraid I cant; Johnnys in america;

      God is an american; Im afraid of americans; Im afraid of the world;

    Im afraid I cant help it;
    Im afraid I cant;
    Im afraid of americans; Im afraid of the words; Im afraid I cant help it; Im afraid I cant; Johnnys in america; Johnnys in america

  • by Brain_Recall ( 868040 ) <> on Monday March 23, 2009 @11:24PM (#27307557)
    The reason why it's "scientifically interesting" is:

    Devil Facial Tumour Disease (DFTD) is extremely unusual as it is only one of three recorded cancers that can spread like a contagious disease. The cancer is passed from devil to devil through biting. The live tumour cells aren't rejected by the animal's immune system because of a lack of genetic diversity among Tasmanian devils.

    See: []

  • by SpazmodeusG ( 1334705 ) on Monday March 23, 2009 @11:36PM (#27307633)
    It uses the hostap driver on Mac - that's HOST Access Point driver, same as what's used in Linux. It creates full access point, not an ad-hoc network. It works well, you can connect to it with other devices just like you would to any other WAP.
    It isn't built into the kernel though (nothing much is on Darwin). This Linux feature they are talking about now has been around for a while. It hasn't been built into the kernel before though.
  • by wrook ( 134116 ) on Monday March 23, 2009 @11:45PM (#27307681) Homepage

    If one wanted to help the plight of the Tasmanian devil, what would be the best thing to do (keeping in mind that I'm not in Australia)?

  • by k-macjapan ( 1271084 ) on Tuesday March 24, 2009 @12:25AM (#27307931)

    Hello. I have looked into this and found the following site. []

    The link to their donation form is []


  • by pedrop357 ( 681672 ) on Tuesday March 24, 2009 @12:59AM (#27308151)

    That's not the same thing. Bridge literally bridges the two NICs much like a switch (network people please correct me if I'm wrong). From what I've seen, ad-hoc is one client to one client, no more.

    I have a Zydas ZD1211 something or other USB 802.11b/g whose Windows driver allows to act as an access point. In my case at least, Windows doesn't provide the necessary options-I can choose ad-hoc or infrastructure client mode.

    I need the driver support for AP mode to set (useless) MAC filtering, SSID, broadcast SSID yes/no, encryption type and for it to be accessible (associative?) with more than one client at a time.

  • by slash.duncan ( 1103465 ) on Tuesday March 24, 2009 @03:49AM (#27308893) Homepage

    No kidding.

    I'm honestly not sure it still applies to KDE in the 4.x era, but at least in the 3.x era, the philosophies were quite different. Gnome's policy of (pseudoquote) "there's one best way to do it and we don't want to confuse the users with too many config options" was extremely frustrating for many KDE users and devs, particularly the power users that /like/ to configure the desktop until it uniquely fits them like a glove, while likewise, the KDE "if it can be configured, different people are going to want different things, so let's expose every single possible configuration option to the user in the GUI" was extremely frustrating to many Gnome users and devs, particularly those who just want it to work, damit, because they have work to do.

    The point is, forcing the devs and users who find the one policy most useful to follow the other one, surely is effective... at causing useless squabbles and getting nothing done! Keep the "there's only one true way" folks away from the "make it configurable for everyone" folks, and both types can continue to improve their product without getting in the way of each other.

    Similarly of course with all the other "Linux is too divided" debates, from too many distributions, to vi/emacs, to... whatever. It's a free community and part of the strength therein lies in the freedom. Even if it were possible to take away that freedom to create one's own product, there'd be little point, as were it to happen, we'd just end up back with the monopolistic monstrosity that is MS. One size does NOT fit all, and encouraging differentiation and innovation, certainly based on common standards, but /only/ /based/ on common standards, is a /good/ thing.

    That said, the one thing that does keep the Linux community from incompatibly splitting up much like the proprietary Unix community did is again, that it's all open and shared. Each distribution and individual app therefore has an interest not only in doing what it was created for really well, even if that splits from the community, but ALSO in following the common solution where it really doesn't matter for what it was created for, because every deviation from the common solution costs maintenance time and resources, time and resources that could otherwise be invested in bettering either the differentiating aspects further, or in advancing the common ones. In practice this dynamic ensures that individual solutions only diverge from the common where it really matters to them, because every divergence costs resources, and divergence just for the sake of it is thus less efficient and dies out relatively quickly, compared to those who focus resources on divergence only where it directly furthers their goals and on otherwise bettering the common solution, submitting patches upstream, etc. Thus, unlike the proprietary Unix solutions, divergence for the sake of divergence simply isn't efficient enough to survive, and ultimately dies. But where there's a good reason for divergence, that only serves to drive a sharper focus on bettering the different solutions that remain, driving the evolution of the community as a whole even faster.

    (Umm... (looking around) I guess it's pretty obvious that I'm a "True Believer" (tm), isn't it. Yes, I am, and for that I'm not going to apologize! =:^)

  • ext4 fs corrupted (Score:2, Interesting)

    by assert(0) ( 913801 ) on Tuesday March 24, 2009 @04:43AM (#27309073) Homepage

    When upgrading from .28 to one of the latest .29 rc's, one of my ext4 filesystems got corrupted. Something to do with resize inode. Had to reboot my old .28 kernel to be able to fix it with fsck. No data loss that I'm aware of (lost+found was empty).

    I think this was originally an ext4dev fs from the .26 era. I have been staying with .28 since. This is a 322 GB fs which is not critical for boot, but obviously I don't want to lose data.

    Perhaps I ought to backup and recreate the fs under .29.

  • Re:Drivers??? (Score:2, Interesting)

    by Computershack ( 1143409 ) on Tuesday March 24, 2009 @05:28AM (#27309281)
    Not in the UK either. Mine have always defaulted to A4.
  • Re:Drivers??? (Score:3, Interesting)

    by SuperAndy ( 1414157 ) on Tuesday March 24, 2009 @06:05AM (#27309457)

    A4 isn't meant to be equal to the golden ratio. It is meant to be 1/Root2, or Root2, depending on your outlook on life. Designed such that if you stick two of them together, you get the next size up. Or on cutting one in half, get the next size down; again, depending on your outlook on life. []

  • Re:Drivers??? (Score:5, Interesting)

    by xouumalperxe ( 815707 ) on Tuesday March 24, 2009 @06:10AM (#27309479)
    The golden ratio isn't really all that useful for writing paper. The A system is based on A0 being 1 square metre, and the sides being in a sqrt(2) proportion so that each time you cut an A(n) sheet in half along the long edge, you get two A(n+1) sheets. Plus, it makes calculating letter weights really easy: given the paper weight in g/m^2, you just divide that by 2^4 = 16 for A4, 2^5 = 32 for A5, etc, then multiply by the number of sheets.
  • by Vanders ( 110092 ) on Tuesday March 24, 2009 @06:12AM (#27309495) Homepage
    No. Linux only needs to understand initrd because it uses initrd. If Linux supported it, GRUB (or any other Multiboot compliant loader) can load all of the required kernel modules on behalf of the kernel, including the root filesystem module.

    This is exactly how Syllable does things. All of it's filesystems, bus drivers and disc subsystems are loadable modules and we rely on GRUB to load them all at boot time, then the kernel initialises them all before it mounts it's root filesystem.
  • by Chemisor ( 97276 ) * on Tuesday March 24, 2009 @02:52PM (#27315853)

    Yes, but it's a more general issue of changing the order of operations. If I write a file and then rename it, I expect the writing to happen before the renaming. If I write to the file in two different places, I expect the actual writes to occur in the order I made them. ext4 designers evidently decided that they knew better how to order the writes, breaking everyone's expectations in the process. Sure, it probably improves performance because you can minimize seeks by ordering writes in a particular order, but the KDE problem illustrates that if you do that, you must at the very least ensure that the dependencies between files are honored. In KDE's case the dependency is between the file and the directory in which it is contained. When such a dependency exists, the filesystem MUST NOT reorder the writes! It isn't just KDE config files. The Firebird database engine, for example, does not fsync at all, relying instead on correct write ordering to preserve data integrity. If the ordering is broken, all Firebird databases will become corrupted.

"Well, it don't make the sun shine, but at least it don't deepen the shit." -- Straiter Empy, in _Riddley_Walker_ by Russell Hoban