Slashdot Mirror


Linux 3.3 Released

diegocg writes "Linux 3.3 has been released. The changes include the merge of kernel code from the Android project. There is also support for a new architecture (TI C6X), much improved balancing and the ability to restripe between different RAID profiles in Btrfs, and several network improvements: a virtual switch implementation (Open vSwitch) designed for virtualization scenarios, a faster and more scalable alternative to the 'bonding' driver, a configurable limit to the transmission queue of the network devices to fight bufferbloat, a network priority control group and per-cgroup TCP buffer limits. There are also many small features and new drivers and fixes. Here's the full changelog."

314 comments

  1. Yea! by colinrichardday · · Score: 2, Funny

    Yea!

  2. Way to go....... by WyzrdX · · Score: 0, Offtopic

    Finally 1 less reason to use anything windows based. I have been looking forward to the code getting out there. I just hope I can continue to learn without returning to basics.

    --
    M O O N... That spells Slashdot.
    1. Re:Way to go....... by MiG82au · · Score: 3, Informative

      A good file system with restripeable RAID is pretty nifty. Admittedly you won't use the feature often, but it's a nice alternative to backing up and restoring onto a new array.

    2. Re:Way to go....... by tpstigers · · Score: 5, Insightful

      Fanboism? Don't be a douchebag. This is a post about a new Linux release. The people who comment on this post should naturally be expected to be Linux users, probably fans. Just as a post about a new OS X release would naturally be populated with Apple fans. It's the people who purposefully comment on threads about products they DON'T like that create a problem. Then there are people like you who just enjoy acting like a dick. Give it a rest.

    3. Re:Way to go....... by BitZtream · · Score: 1, Informative

      So you mean Solaris or FreeBSD then with ZFS I assume?

      --
      Persistent Volume manager for Kubernetes - https://github.com/dwimsey/openshift-pvmanager
    4. Re:Way to go....... by markdavis · · Score: 1

      You are an ass. -1 Troll

    5. Re:Way to go....... by GmExtremacy · · Score: 4, Insightful

      It's the people who purposefully comment on threads about products they DON'T like that create a problem.

      I wouldn't say it's a problem. They, like to the people who like it, are simply stating their opinions.

    6. Re:Way to go....... by Electricity+Likes+Me · · Score: 3, Informative

      ZFS has no support for resizing or restriping it's RAID pools, or shrinking the storage units.

      It's a giant missing feature on an otherwise excellent FS.

    7. Re:Way to go....... by DeathFromSomewhere · · Score: 0

      It's the people who purposefully comment on threads about products they DON'T like that create a problem.

      Agreed. Nothing like a good echo chamber.

      --
      -1 overrated isn't the same thing as "I disagree".
    8. Re:Way to go....... by unixisc · · Score: 2

      You are right, but the OP of this thread is the one who started it w/ snide references to Windows. He could have kept it to a discussion of just Linux, and nobody from the Windows camp would have bothered.

    9. Re:Way to go....... by mug+funky · · Score: 3, Interesting

      the android merge means i can play angry birds without having to use wine...

      the various fixes means my decaying old netbook will still remain usable, and even appear snappy next to one half it's age.

    10. Re:Way to go....... by Anonymous Coward · · Score: 5, Insightful

      So... which part of this release actually provides a compelling reason to use Linux over any other OS?
      You've been itching for something to run on that TI C6X system you built?

      The fanboisim here makes me gag. Apple has nothing on you guys.

      Hey Cowboy did you know that the Linux Kernel currently runs most smart tvs, bd players, and other home entertainment devices. I would be willing to bet that the number of Samsung, Sony, LG and other tvs and devices running on the Linux kernel is much greater than the number of Macs, and PC currently in use combined!

      The reason for this is that any manufacturer can use OpenSource software like the Linux kernel and modify it to their own needs without sending money to Redmond for every device they sell. This is why Microsoft and Apple have failed in the embedded market with perhaps the exception of some car companies like Ford Motors. Ballmer can rant, rave and do all the paid shill crap he wants. Fact is as the kernel becomes more open to modification from companies like Google with Android optimisations and slick coding Microsoft will become irrelevant in many markets.

      The post was about the most important core software released in history so go pound on your PC, and post how linux sucks somewhere where someone cares. The Linux kernel is one hell of allot more that just the base of an OS as you perceive it.

    11. Re:Way to go....... by Anonymous Coward · · Score: 1, Insightful

      >It's the people who purposefully comment on threads about products they DON'T like that create a problem.

      So what are your thoughts on how Slashdot has used a broken window to represent Windows stories for at leats a decade?

      And what are your thoughts about how until just a few months ago Slashdot used the borg-ified face of the world's largest philanthropist as an icon for Microsoft stories?

      Are you (and all the people who gave your post karma) implying that Slashdot itself and all its editors are douchebags who create problems?

      Because I would agree with that sentiment.

    12. Re:Way to go....... by Anonymous Coward · · Score: 0

      Shrinking or restriping are things I would never do on a live FS, so it's probably bloat anyway.

    13. Re:Way to go....... by Anonymous Coward · · Score: 2, Insightful

      Right. Because if you wouldn't do it, it's useless.

    14. Re:Way to go....... by CSMoran · · Score: 5, Funny

      the android merge means i can play angry birds without having to use wine...

      Note that playing sober raises it to another level of difficulty.

      --
      Every end has half a stick.
    15. Re:Way to go....... by MrHanky · · Score: 4, Informative

      Bullshit. Not only does a merge of Android kernel features not mean you can play angry birds under some regular Linux distro (you'll need, oh, Dalvik and Android's windowing system which is not X11), you can already play Angry Birds in Chrome, no Wine required. The kernel is entirely irrelevant. If you don't know what you're talking about, just shut up.

    16. Re:Way to go....... by unixisc · · Score: 1

      Question - what is Android's Windowing system?

    17. Re:Way to go....... by pegdhcp · · Score: 4, Insightful

      Then neither you are not working with web developers who change their mind daily about basic system design nor you are working in an environment where uptime is not a concern nor you have very expensive load balance boxen that protects you from mundane facts of life. Either way, you are a very lucky person... OTOH we, of lesser levels of humanity, are grateful for one more tool that would make our lives easier.

    18. Re:Way to go....... by chuckymonkey · · Score: 3, Insightful

      There's valid criticism and then there's just being an asshole, they're very different.

      --
      "Some books contain the machinery required to create and sustain universes."-Tycho
    19. Re:Way to go....... by GmExtremacy · · Score: 3, Insightful

      And what constitutes an "asshole" is subjective. I don't see anything wrong with stating your opinion and doing nothing besides that, even if someone thinks it's harsh.

    20. Re:Way to go....... by WyzrdX · · Score: 3, Informative

      So... which part of this release actually provides a compelling reason to use Linux over any other OS? You've been itching for something to run on that TI C6X system you built?

      The fanboisim here makes me gag. Apple has nothing on you guys.

      My post had nothing to do with fanboisim. I currently use Win, Mac, and 5 different flavors of Linux not counting my tv, car stereo, smart phone, and at least 3 other devices that run modified Linux code.

      I have been using Linux off and on for years. But it has only been recently that I have really been modifying it and making it do what I want and how I want. Currently I have my MBP that I use that I need to have Windows installed on due to either software differences in Win and Mac versions (ie. Quickbooks) or because I need access to some windows items while on my Mac.

      I use a Windows machine for some games only because the gaming industry seems to feel that there is only 1 OS that is worth the time. Even Mac is lacking on a lot of the games I play. And I am not a big fan of running some games in a X over for Linux.

      When I use Linux it is for everything else from my firewall to Development. But I am dependent on my Phone and Tablet which run Android. Now with android kernel merge I may have a greater use for using linux than before.

      And as I stated, I just hope I dont have to start at the basics. I am no guru by far but I really dont want to have to thrash what I have learned in the last year or so and start again. Will have to wait and see what the distros look like.

      --
      M O O N... That spells Slashdot.
    21. Re:Way to go....... by tpstigers · · Score: 2

      Your points are only vaguely related to the subject at hand, but I can see why these particular issues distress you so. I don't think anyone would think any less of you if you just stopped visiting Slashdot.

    22. Re:Way to go....... by Anonymous Coward · · Score: 2, Insightful

      Don't forget the 250+ million Android devices floating around out there.

      Linux is everywhere these days although mostly invisible (some would call that "seamless").

    23. Re:Way to go....... by Anonymous Coward · · Score: 0

      Don't forget the 250+ million Android devices floating around out there.

      Right, but he was talking about devices for which manufacturers don't have to send money to Redmond. If you've seen the news lately, most Android handset makers (most being defined as over 50%) are sending money to Redmond for "patent protection" over undisclosed patents that Android may possibly violate.

    24. Re:Way to go....... by RoboJ1M · · Score: 1

      Nothing. Linux is a kernel, not an operating system.
      If you don't know the difference, either STOP reading Slashdot or START reading slashdot, whichever applies to you.

    25. Re:Way to go....... by eric_herm · · Score: 0

      Sorry but asshole is well defined. It cannot be placed anywhere on the body. And if nothing go out of it, this is not a asshole.

      For example, ear hole are not ass hole.

    26. Re:Way to go....... by DeathFromSomewhere · · Score: 1

      Indeed. For example I would say that discouraging someone from posting because they hold a different opinion is being an asshole.

      --
      -1 overrated isn't the same thing as "I disagree".
    27. Re:Way to go....... by leuk_he · · Score: 1

      expand o matic raid z

      It is "only" an implementation issue.

      I would love it: just add a HD, and get extra diskspace that is raid-protected, or (schedule a ) remove a HD, and the data will restripe over the remaining HD's.

    28. Re:Way to go....... by Electricity+Likes+Me · · Score: 1

      Now that would solve the problem entirely.

      It's kind of an important issue for making it a general purpose file system, since with things like SSD's and the like you can easily end up suddenly finding yourself wanting to migrate down a size in data capacity. Also the whole "throw disks into the pool" attitude that ZFS takes means you can end up creating non-optimal situation quickly, but have a much harder time backing out of them.

    29. Re:Way to go....... by bill_mcgonigle · · Score: 2

      ZFS has no support for resizing or restriping it's RAID pools, or shrinking the storage units.

      Not entirely true - if you replace all of the devices in a pool with larger ones (one or two at a time depending on your pool) when all of the devices are of a larger size, ZFS will automatically expand the pool.

      So, if you have a pool of 7 1TB drives, say with 4.5 TB of usable storage and you replace them with 7 2TB drives, when the last one is done rebuilding you'll have 9TB of usable storage without doing anything more as an admin.

      Remember also that ZFS dynamically allocates filesystem storage out of a pool, so the need to shrink filesystems is much less relevant than with LVM and ext*/xfs. I'm sure there are cases where it could be handy, but I've personally never run into one in the 4 years I've been running ZFS systems.

      --
      My God, it's Full of Source!
      OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
    30. Re:Way to go....... by larien · · Score: 1

      I have, albeit with Veritas. Took about a month of restriping 20TB to change the stripe widths. Just because you don't do it, doesn't mean that others don't.

    31. Re:Way to go....... by unixisc · · Score: 1

      Aside from any license issues, are there any reasons that the Veritas FS can't be included along w/ the other Linux file systems, instead of inventing a new one?

    32. Re:Way to go....... by mug+funky · · Score: 1

      wow, i see Angry Birds is a touchy subject for you. be sure to take it up in your next session, for all our sakes.

      aside from that, Angry Birds doesn't exactly run at usable speeds in Chrome. i'm not wanting to upgrade my hardware to run slow software when a fast alternative exists. i don't want to roast my crutch with an alienware monstrosity just for a few minutes of gratification on the way to work. YMMV.

      no, i don't know jack about window managers. you got me. you're clearly superior to me in every way. i'll be sure to remove myself from the genepool, and while i'm at it, i'll remove my offspring as well, just in case. the world will be better off without my ignorance of window managers for smartphones.

  3. Keep it up. by Severus+Snape · · Score: 5, Insightful

    The Linux kernel guys show that constant steady frequent releases are the way forwards, note to GNOME and KDE guys, you got it wrong.

    1. Re:Keep it up. by houstonbofh · · Score: 5, Insightful

      That and the kernel guys actually put in features people want and need, not shove unwanted changes down the users thoughts...

    2. Re:Keep it up. by Anonymous Coward · · Score: 0, Troll

      Last time I checked, KDE makes releases nearly every two months. Too-rapid releases with major API changes (such as the ACPI changes in 3.2) can cause rather annoying problems with third-party drivers.

      @World, if you're going to use a three-part version number, please follow semver. Don't make API-breaking changes in a minor version.

    3. Re:Keep it up. by MrEricSir · · Score: 5, Funny

      shove unwanted changes down the users thoughts...

      You can uninstall GBrain and/or MindKontrol to prevent Gnome and KDE from controlling your thoughts.

      --
      There's no -1 for "I don't get it."
    4. Re:Keep it up. by Anonymous Coward · · Score: 0

      The Linux kernel guys show that constant steady frequent releases are the way forwards, note to GNOME and KDE guys, you got it wrong.

      Not enough information... obvious bug has no merit whatsoever and we are doing everything perfect. marked as incomplete. /sarcasm

    5. Re:Keep it up. by Tanktalus · · Score: 3, Funny

      I was actually considering that the other day. I'm not sure why, but I changed my mind. ThEy ArE wOnDeRfUl ApPs.

    6. Re:Keep it up. by MurukeshM · · Score: 2

      Let me rephrase GP for you..

      ... the kernel guys actually put in features people who care about such things want and need, while not pissing the hell out of others...

    7. Re:Keep it up. by Smauler · · Score: 4, Funny

      The Linux kernel guys show that constant steady

      I agree... 1 sec.

      frequent releases are the way forwards, note

      Argh... just got to...

      to GNOME and KDE

      update firefox...

      guys, you

      Again?

      got

      Must

      it

      finish

      wrong.

      comment.

    8. Re:Keep it up. by Barsteward · · Score: 1

      thats Fedora's fault for upgrading, not KDE's. They should have tested first, then i guess you should have too. Who puts a new version of the OS on a primary working machine without first testing it??

      --
      "The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
    9. Re:Keep it up. by Anonymous Coward · · Score: 0

      Yeah, everyone loved it when Firefox did it.

    10. Re:Keep it up. by Kjella · · Score: 1

      Oh boy, the same "it's not our fault for releasing KDE 4.0, it's the distros who ship it" just this time it's for KMail2. Distros ship what you release. Users use what you release. They only package it, if you release turds the users get gift-wrapped turds. If it's not a packaging issue - and a quick search indicates many distros share the pain so it's not - then it's a developer issue. Call it alpha, betas, previews, RCs, anything you want but when you call it a release then it's nobody's fault but your own that people think it's ready to upgrade from the old release.

      --
      Live today, because you never know what tomorrow brings
    11. Re:Keep it up. by unixisc · · Score: 1

      That and the kernel guys actually put in features people want and need, not shove unwanted changes down the users thoughts...

      Okay, so what are the kernel changes that users need? Filesystem - we currently have a choice of ext2, ext3 and ext4 - what's inadequate about any of them that couldn't be resolved in an ext5? Any reason why re-strippable RAID can't be in that? Reason I ask is that we already have a plethora of Unix file systems - UFS, XFS, ZFS, ext2/3/4, Hammer, Veritas, and who knows what else. Linux did well to support OpenRISC, but having it support VLIW DSPs is overkill. One thing they could do as far as the Linux kernel goes is work on drivers - particularly Wi-Fi drivers, and do what's possible to ensure that 3.3, or 3.4 support just about every peripheral device there is out there. Aside from that, as far as I can tell, the Linux kernel is pretty much complete.

      In the field of kernels, I think work could be done making Minix 3 complete, using it along w/ Hurd to make a complete OS (fork one of the 2 if the licenses are that incompatible), as well as porting Minix to Raspberry Pi. Oh, and maybe have a complete Java OS consisting of a Java VM that runs all Bytecode, so that it can be a native platform for Java, w/o any of the overhead of running any underlying non-Java OS.

    12. Re:Keep it up. by semi-extrinsic · · Score: 5, Informative

      Okay, so what are the kernel changes that users need? Filesystem - we currently have a choice of ext2, ext3 and ext4 - what's inadequate about any of them that couldn't be resolved in an ext5? Any reason why re-strippable RAID can't be in that?

      The general notion is that btrfs will "be" ext5 (i.e. it will be the next "updated" but still stable and mainstream FS), and that there will not be a filesystem with the actual name "ext5". For those who don't need btrfs features, ext4 will suffice. This is also the intent of Theodore Ts'o, the principal developer of ext3/4.

      I believe the reason for this is that the innovation going on in filesystems is centered around some big rethinks, e.g. btrfs uses a copy-on-write B-tree (a concept introduced in 2007). It would be a pain in the neck (or impossible) to innovate like this and remain backwards compatible with ext2/3/4, thus btrfs is not called ext5.

      One thing they could do as far as the Linux kernel goes is work on drivers - particularly Wi-Fi drivers, and do what's possible to ensure that 3.3, or 3.4 support just about every peripheral device there is out there. Aside from that, as far as I can tell, the Linux kernel is pretty much complete.

      How about you RTFS? To quote:

      There are also many small features and new drivers and fixes.

      --
      for i in `facebook friends "=bday" 2>/dev/null | cut -d " " -f 3-`; do facebook wallpost $i "Happy birthday!"; done
    13. Re:Keep it up. by Verunks · · Score: 2

      The Linux kernel guys show that constant steady frequent releases are the way forwards, note to GNOME and KDE guys, you got it wrong.

      I don't know about gnome, but kde release a new version every six months since 4.0

    14. Re:Keep it up. by TheRaven64 · · Score: 1
      I think what you mean is:

      That, and the kernel developers implement the features people are willing to pay them to implement because they have a business need for them.

      --
      I am TheRaven on Soylent News
    15. Re:Keep it up. by Anonymous Coward · · Score: 0

      shove unwanted changes down the users thoughts...

      You can uninstall GBrain and/or MindKontrol to prevent Gnome and KDE from controlling your thoughts.

      ... but GBrain and MindKontrol make you not wanting to do that.

    16. Re:Keep it up. by chrb · · Score: 1

      The firefox guys should've kept their versions at c.y instead of dropping the constant c, and everyone would've been happy. A number that changes scares people - but prefix it with a number that doesn't change and people are ok with it.

    17. Re:Keep it up. by turbidostato · · Score: 2

      "They only package it"

      Oh, no, they do much more than this: they choose what to package.

    18. Re:Keep it up. by LordLimecat · · Score: 2

      So why then all the hate when Mozilla follows the same release mentality?

    19. Re:Keep it up. by slashgrim · · Score: 1

      So why then all the hate when Mozilla follows the same release mentality?

      If Linux kernel followed the same release mentality of Mozilla, we'd have Linux 10.3 already.

    20. Re:Keep it up. by Anonymous Coward · · Score: 0

      GNOME has had regular releases every six months since 2002. Fedora and Ubuntu both based their schedules on GNOME's because it's so steady. But don't let a pesky thing like facts get in your way of people thinking you're insightful.

    21. Re:Keep it up. by Anonymous Coward · · Score: 1

      I consistently find it hard to believe how much Slashdotters get worked up over Mozilla's version numbers.

    22. Re:Keep it up. by RoLi · · Score: 1

      I guess the emphasis is on the "not shove unwanted changes down the users thoughts"-part.

      Only a miniscule minority may truely need the new features, but they don't hurt the other, existing users.

    23. Re:Keep it up. by Anonymous Coward · · Score: 0

      The funny thing about ext? version is it that there have been a plethora of different innovative file systems, while ext just evolved to incorporate some of the most needed things, and remained a defacto standard. Will we all be using btrfs or an evolution of btrfs in ten years? Time will tell.

    24. Re:Keep it up. by mister_playboy · · Score: 2

      Because Slashdot is full of OCD-ish folks who find deep meaning in numbers that most people pay no attention to?

      --
      Do what thou wilt shall be the whole of the Law ::: Love is the law, love under will
    25. Re:Keep it up. by Anonymous Coward · · Score: 0

      GNOME has had a 6 month cadence for at least ten years. This guy probably has a real problem with GNOME and/or KDE but when he can't even articulate what the problem is, it's just not a useful starting point for a discussion...

    26. Re:Keep it up. by sjames · · Score: 1

      He said frequent releases, not frequent pestering.

    27. Re:Keep it up. by GuB-42 · · Score: 1

      obligatory : http://xkcd.com/619/

    28. Re:Keep it up. by ConceptJunkie · · Score: 1

      If Linux kernel followed the same release mentality of Mozilla, we'd have Linux 10.3 already.

      So? Is that any less arbitrary than when they decided to bump the version to 3.0?

      The version number is nothing but a name. Numbers make it easier to know in what order the releases came out. Using the numbers to denote Major/Minor/Revision provides more information, but again, it's all relative. Major is bigger than minor, but how much? It's fine if you decide that the major number will be bumped only if the ABI changes or something like that, but that doesn't always apply to an end-user app. Does it really change anything Firefox is up to version 11, vs. version 4.6? In either case, the important piece of information is whether you're running the latest version.

      less is up to version 436!

      --
      You are in a maze of twisty little passages, all alike.
    29. Re:Keep it up. by ConceptJunkie · · Score: 1

      They're just jealous of my IQ.

      --
      You are in a maze of twisty little passages, all alike.
    30. Re:Keep it up. by ultranova · · Score: 1

      The firefox guys should've kept their versions at c.y instead of dropping the constant c, and everyone would've been happy. A number that changes scares people - but prefix it with a number that doesn't change and people are ok with it.

      More to the point, c+1.y is supposed to indicate that the release has major new features compared to c.*, while c.y+1 means some bug(s) in c.y have been fixed.

      Firefox changed this scheme partially for marketing reasons, but mostly to remove the ability for users to stay on a particular c series. In other words, you can't get bugfixes without new features, which inevitable end up introducing new bugs. No, you have to act as beta tester for whatever the Firefox team comes up with. At the same time, if the Firefox team decides you don't need no stinking status bar, you will either give up either it or bug fixes. To put it even blunter, Firefox versioning scheme changed so that users would have less choice, and would have harder time forming a critical mass at any particular "feature level" to make a branch.

      That's why people hate the new Firefox versioning scheme: it's a deliberate attempt to make them as powerless as possible in an open-source project. Which is understandable from a developer point of view, since backporting bugfixes into old branches is a waste of resources. And of course every designer wants their pet feature be used by everyone, so succumbing to the temptation of forcing the issue is quite understandable.

      --

      Forget magic. Any technology distinguishable from divine power is insufficiently advanced.

    31. Re:Keep it up. by tehcyder · · Score: 1

      Okay, so what are the kernel changes that users need?

      The gift of immortality to the user. Until then, colour me unimpressed.

      --
      To have a right to do a thing is not at all the same as to be right in doing it
    32. Re:Keep it up. by Anonymous Coward · · Score: 0

      That's not what GBrain told me.

  4. Which distributions? by cold+fjord · · Score: 0

    So, which will be the first commercial and non-commercial distributions to get it?

    --
    much of left-wing thought is a kind of playing with fire by people who don't even know that fire is hot - George Orwell
    1. Re:Which distributions? by flyingfsck · · Score: 2

      You don't have to wait for a distribution house to pick it up. Downloading and doing a kernel compile is easy.

      --
      Excuse me, but please get off my Pennisetum Clandestinum, eh!
    2. Re:Which distributions? by Anonymous Coward · · Score: 3, Informative

      It will be used for the Fedora 17 beta release currently scheduled for April 3rd.

    3. Re:Which distributions? by Hardolaf · · Score: 4, Informative

      Arch Linux will probably support it in a few days. The packages have been marked outdated and there is already a 3.3rc7-1 ( https://aur.archlinux.org/packages.php?ID=50893 )release in the wild that will probably be the basis for the updated to 3.3.

    4. Re:Which distributions? by Anonymous Coward · · Score: 0

      So which kernel will Occiferic Ocelot - oh oops, sorry, we're now awaiting Pretentious Panda or something - ship with in 4,3,2,1 weeks? I expect the major distros like Ubuntu or Mint will update to the new kernel as part of their usual schedule, right? Maybe I'm not adventurous enough to venture ahead of the distro providers who would ensure other software is updated to work with the new kernel....

    5. Re:Which distributions? by Anonymous Coward · · Score: 0

      The way Ubuntu is, I wouldn't expect to see this until Oneiric+3.

    6. Re:Which distributions? by Anonymous Coward · · Score: 1

      I expect the major distros like Ubuntu or Mint will update to the new kernel as part of their usual schedule, right?

      Hahaha, no. Ubuntu takes ages to update the kernel. And with good reason: it's an end-user's distro, it can't be bleeding edge. Try Arch if you want the newest kernel (keep in mind 3.3 is not yet on the official repository, but should be shortly).

    7. Re:Which distributions? by meow27 · · Score: 3, Informative

      Gentoo and/or Arch

    8. Re:Which distributions? by Anonymous Coward · · Score: 0

      I noticed the maintainer of Grailux (a python-centric distro) has already upgraded to the 5.5 kernel...

    9. Re:Which distributions? by quarkscat · · Score: 4, Interesting

      I cannot answer this question for any GNU/Linux distribution except for Slackware, which may or may not get Linux Kernel 3.3.xx as part of an official distribution for at least one Slackware release iteration ... But my personal Slackware machine will be getting 3.3 as soon as it finishes building and I reboot the machine. ;-)

      It's nice to have a GNU/Linux distribution that doesn't jerk users around with strange application locations, misaligned library versions, or an update schedule tied to commercial support contracts. I've tried the rest, and I returned to the best (imho), since GNU/Linux kernel 0.96. Don't try dropping a new kernel source tar-ball onto RH Enterprise Server, Fedora, or even Ubunto -- it will break your system, and your $$$$ support agreement.

    10. Re:Which distributions? by zill · · Score: 4, Funny

      5.5 kernel...

      This is either a typo or python is way more powerful than I thought.

    11. Re:Which distributions? by Anonymous Coward · · Score: 0

      Hmmm, -1 off-topic for asking what distributions will use the new Linux in a story about the new Linux kernel. I guess someone is still angry from another discussion. :D

    12. Re:Which distributions? by Anonymous Coward · · Score: 0

      Well, you obviously aren't familiar with Monarchic counting systems of 8th century Britain.
      In all fairness, I'm not sure if I was referring to African or European systems. Let me feast upon the indigenous animals, consult the book of shrubberies, and weigh the evidence.
      Isn't this a marvelous exchange? I wasn't sure how it would pan out, when I first started typing, but I'm really quite thrilled about it, now.

    13. Re:Which distributions? by gweihir · · Score: 1

      My Debian installation. Only lamers need distro-kernels.

      --
      Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
    14. Re:Which distributions? by Anonymous Coward · · Score: 0

      Did you ever try about Arch Linux? Howd you find it? (serious question)

    15. Re:Which distributions? by tao · · Score: 1

      from __ future __ import linux-5.5 ?!

    16. Re:Which distributions? by dargaud · · Score: 1

      Don't try dropping a new kernel source tar-ball onto RH Enterprise Server, Fedora, or even Ubunto -- it will break your system

      Where are the lists of additions/modifications that those distros make to the kernel ?

      --
      Non-Linux Penguins ?
    17. Re:Which distributions? by Anonymous Coward · · Score: 0

      For RedHat, in a single big monolithic unreadable patch against vanilla.

    18. Re:Which distributions? by sensei+moreh · · Score: 1

      Fedora 17 pre-beta already is close: ~$ rpm -q kernel kernel-3.3.0-0.rc7.git0.3.fc17.x86_64

      --
      Geology - it's not rocket science; it's rock science
    19. Re:Which distributions? by quarkscat · · Score: 1

      On a number of Linux distributions, installing a vanilla (kernel.org) kernel source tar-ball isn't even possible. Either the kernel will fail to successfully build due to mismatched versioning of libraries or sometimes gcc itself. The underlying issues are not merely a vanilla kernel. For other Linux distributions, the directory structure itself doesn't match up to generally accepted standards. Often the user of a specific distribution must wait until the vendor's own series of patches (kernel, libraries, utilities, etcetera) are made available -- on the vendor's timetable, not the users.

      I am not denigrating other GNU/Linux distributions -- there are plenty of reasons to stick with a particular distribution if you are comfortable with the timeliness of upgrades and the level of support from that distribution's user base. If the user pays for a support contract, deviation from the standard baseline can render your support contract null-and-void, or at least problematic.

      I personally like to try brand new whiz-bang applications and utilities from the original development teams rather than relying upon a specific distribution vendor to make that application or utility available via their support structure. It's all a matter of your personal (or corporate) comfort level with risk. No one in a production environment would change a kernel, an application , or utility on a production machine without a graceful way to back out of those changes. This is not just a GNU/Linux platform issue, but applies to any OS and any application, and I have been involved in supporting a number of commercial unixes and "hobbyist" unix-like OSes. What I have discovered is that If you stick with, for instance Red Hat, you will learn Red Hat, but if you learn a distribution like Slackware you will learn unix -- that's what kept me gainfully employed.

    20. Re:Which distributions? by quarkscat · · Score: 1

      ES, TDIAF, AC! ---------- BTW, AC, I've been immersed in unix & unix-like OSes for 15 years and am no "slackware" snob. I was merely pointing out some issues that other slashdot readers might need to be aware of before trashing their OS.

      Ubuntu has a ways to go before it is as polished & stable as HPUX, even and especially on a VME backplane, but it's a damn sight cheaper. Get over your hyper-sensitivity to fact-based criticism of your particular favorite brand of GNU/Linux -- none are perfect, even Slackware. Even IRIX 6.5.xx wasn't perfect, although I have never seen another commercial unix variant as pretty, stable, and logically laid out -- even Mac OS X could stand some serious improvements.

    21. Re:Which distributions? by Nimey · · Score: 1

      It's too bad Slack has such a crappy package-management system or I might still be using it.

      --
      Hail Eris, full of mischief...

      E pluribus sanguinem
    22. Re:Which distributions? by quarkscat · · Score: 1

      I haven't tried Arch Linux, nor seriously tried a number of other linux variants in the past 18 months, so I cannot say whether it will break with the installation of a vanilla (kernel.org) kernel. Configuring and building a new kernel from source is my preferred method of getting a new kernel, rather than merely installing a replacement distribution kernel binary. But then, I like NetBSD, FreeBSD, Dragonfly BSD, and (especially) OpenBSD as well.

      As an aside, I tend to concur with those who would say that a camel is a horse designed by committee.

    23. Re:Which distributions? by Anonymous Coward · · Score: 0

      Read the documentation about: from __future__ import *

      http://docs.python.org/library/__future__.html

    24. Re:Which distributions? by macemoneta · · Score: 1

      Already available in Fedora, of course.

      --

      Can You Say Linux? I Knew That You Could.

    25. Re:Which distributions? by turbidostato · · Score: 1

      "You don't have to wait for a distribution house to pick it up. Downloading and doing a kernel compile is easy."

      No, you don't have to wait. But since Torvalds himself stated that stabilizing and bugfixing kernel isn't its kernel tree's job no more but a distribution issue, I wouldn't count on that being such a wise movement unless discovering problems and rebuilding boxes is either your job or your hobby.

    26. Re:Which distributions? by Barefoot+Monkey · · Score: 4, Funny

      Your problems with Slackware appear to stem from your use of the wrong mock-religion meme in your signature. Eris, in particular, is known to cause trouble. Try changing your signature to The Subgenius must have Slack and the package management should improve.

    27. Re:Which distributions? by Anonymous Coward · · Score: 0

      No, building a new kernel on RH, or Ubuntu will not break the system. What kind of BS is that? The process for compiling and installing a new kernel works exactly the same on Ubuntu and Red Hat/Fedora as it does on Slackware.

    28. Re:Which distributions? by menkhaura · · Score: 2

      It already is. I've just updated my Arch and it came with the new kernel.

      --
      Stupidity is an equal opportunity striker.
      Fellow slashdotter Bill Dog
    29. Re:Which distributions? by Nimey · · Score: 1

      Well played, sir, well played.

      --
      Hail Eris, full of mischief...

      E pluribus sanguinem
    30. Re:Which distributions? by TeknoHog · · Score: 1

      My Gentoo installation. Only lamers need binary distros.

      --
      Escher was the first MC and Giger invented the HR department.
    31. Re:Which distributions? by armanox · · Score: 1

      I've been using a custom kernel on Fedora (and the older Red Hat) for years (Currently running 3.2.x on F14 at home) without issue.

      --
      I'm starting to think GNU is the problem with "GNU/Linux" these days.
    32. Re:Which distributions? by pseudofrog · · Score: 1

      One, two, five (three sir!), three.

      Also, Google tells me that Grailux doesn't exist. Yet, anyway.

    33. Re:Which distributions? by Anonymous Coward · · Score: 0

      its only in testing repository yet, but it should advance to core in about 1-2 weeks

    34. Re:Which distributions? by somenickname · · Score: 1

      Don't try dropping a new kernel source tar-ball onto RH Enterprise Server, Fedora, or even Ubunto -- it will break your system, and your $$$$ support agreement.

      The support agreement part is possible but the actual building and installation of a new kernel on Debian/Ubuntu couldn't be easier. With a single command you can build proper .deb packages from the kernel source and with a second command install them complete with grub updates, DKMS updates, etc. I've been doing this for years without problems. The days of breaking your machine with a custom kernel are long gone for modern distros.

    35. Re:Which distributions? by bill_mcgonigle · · Score: 1

      Don't try dropping a new kernel source tar-ball onto RH Enterprise Server, Fedora, or even Ubunto -- it will break your system, and your $$$$ support agreement.

      Are you aware of why people buy Redhat support agreements? If so, do you think Redhat could offer such agreements for the same kind of money if they supported every possible software version combination? Do you think buying into Redhat's system in involuntary?

      As far as Fedora, the 3.3 packages are already built for 17 and 18; 16 should be along in a day or two. But if you're impatient odds are very good that you can rebuild the 17 package on 16 with a simple 'rpmbuild --rebuild' command, and install it with 'rpm -i kernel-foo....'. Fedora folks care about engineering, so if something doesn't work as it should, they want to know about it - you won't hear them using "it's not supported" as an excuse.

      But, hey, I used to run Slackware too - it was my favorite distro on the InfoMagic set.

      --
      My God, it's Full of Source!
      OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
    36. Re:Which distributions? by neuro88 · · Score: 1

      For ubuntu/debian, try make-kpkg (apt-get install kernel-package).

      Download the tarball, untar, make menuconfig (or whatever you prefer), then do make-kpkg kernel_image (add --initrd if you need an initial ramdisk). There's various other useful options as well. The result is a nice debian kernel package that can be managed by the system. Works much better than the old make bzImage, etc.

      I used to do things by hand in debian (and I never really had trouble in debian with this approach, but I haven't tried it in debian since 2000-2001 or so which is when I first learned about make-kpkg). make-kpkg works great in ubuntu as well, obviously. Gentoo's equivalent seems to be genkernel. Otherwise I agree about the rest of those distributions.

    37. Re:Which distributions? by bmorency · · Score: 1

      If you are running Debian Sid it will ask you to upgrade to the new version.

  5. Recursion by philip.paradis · · Score: 3, Funny

    If I deploy a 3.3 guest on a host running 3.3, does it automatically become 3.3 repeating and go on forever?

    --
    Write failed: Broken pipe
    1. Re:Recursion by Vectronic · · Score: 1

      3.3 all the way down

    2. Re:Recursion by Anonymous Coward · · Score: 0

      Ask the peeps at stack overflow site.

    3. Re:Recursion by mister_playboy · · Score: 1

      Maybe... try playing this album while you install. \m/

      --
      Do what thou wilt shall be the whole of the Law ::: Love is the law, love under will
  6. *Btrfs* by Anonymous Coward · · Score: 0, Offtopic

    Sorry, I farted.

  7. C6X support is surprising by Anonymous Coward · · Score: 5, Interesting

    Wow, I had no idea there was work in porting Linux to DSP architectures. That's quite an interesting development. I wonder what the use case is, since DSPs are typically used for very specific, real-time work, not for hosting general-purpose operating systems.

    Also, it's quite surprising to me since as far as I know it's necessary to use TI's compiler to generate C6X code. I found one initiative to port GCC to it, but afaik it didn't get finished. My understanding is that it is no small job to get Linux to compile on non-supported compilers, so I'm interested in the toolchain they are using. For my own work on a C6711, I've been using the TI compiler under Wine. (Which works fine actually, although I had to generate an initial project in CodeComposer to get some of the board-specific support files.)

    1. Re:C6X support is surprising by unixisc · · Score: 3, Interesting

      Also, isn't TI C6X a VLIW - in which case, it would need some very elaborate state of the art compilers? Anybody writing a compiler for this thing would have to write one that does, in addition to the usual activities, VLIW stuff like register renaming and allocation, branch prediction and speculative execution, and so on. Would GCC (or LLVM/Clang) put that sort of effort into a compiler?

    2. Re:C6X support is surprising by Anonymous Coward · · Score: 1

      Transmeta (remember them?) had a gcc-code generator for their VLIW which was, naturally, GPL. Red Hat was maintaining it after transmeta went tits up, working on generalizing it for more architectures. I don't know if it was merged into mainline gcc (the steering committee had a stick up their ass about it at the time since it was so specific to transmeta).

    3. Re:C6X support is surprising by macshit · · Score: 4, Informative

      Also, it's quite surprising to me since as far as I know it's necessary to use TI's compiler to generate C6X code. I found one initiative to port GCC to it, but afaik it didn't get finished. My understanding is that it is no small job to get Linux to compile on non-supported compilers, so I'm interested in the toolchain they are using.

      GCC 4.7 (which will be released soonish; it's basically already done) supports the C6X architecture.

      From the GCC 4.7 release notes:

      New Targets and Target Specific Improvements:
      ...
      C6X

      • Support has been added for the Texas Instruments C6X family of processors.
      --
      We live, as we dream -- alone....
    4. Re:C6X support is surprising by unixisc · · Score: 1

      That was my point. Every VLIW implementation would have its own code generator for things that in the RISC and CISC worlds, the CPUs usually do. Since a branch prediction on a Crusoe would differ from a C6X which in turn would differ from an Itanium, there are no VLIW specific changes that can be merged into a mainline one-size-fits-all gcc. That's the difference b/w CISC or RISC implementations, vs a VLIW one.

    5. Re:C6X support is surprising by Anonymous Coward · · Score: 0

      Great! I hope they fixed the currently broken AVR support.

    6. Re:C6X support is surprising by macshit · · Score: 1

      Great! I hope they fixed the currently broken AVR support.

      The page I linked to above also shows many changes to AVR support; whether that makes it non-"broken" or not, I don't know.

      --
      We live, as we dream -- alone....
    7. Re:C6X support is surprising by aiht · · Score: 1

      Although, the Linux-c6x release notes page says "this release uses the gcc tool chain (4.5-124) from Code Sourcery."
      The Code Sourcery link says "Mentor Graphics makes extensive enhancements to the base versions, adding support for more CPUs, improving code-generation, and addressing defects found through its validation process."
      Maybe their patches got rolled into mainline in version 4.7?

    8. Re:C6X support is surprising by steveha · · Score: 4, Informative

      The TI C6X line of chips are not only VLIW, they are "DSP" chips, optimized for signal processing operations. Also, this chip has no MMU. Nobody is going to build a tablet computer or any other general-purpose device based on one of these.

      I think for the near term at least, anyone using a TI C6X will be using the TI C compiler. TI has a whole IDE, called Code Composer Studio.

      But now we have the possibility of running Linux on the chip.

      The one time I worked with a TI DSP chip, I didn't really have an operating system. Just a bootstrap loader, and then my code ran on the bare metal, along with some TI-supplied library code. Now I'm working with an Analog Devices DSP chip and it's the same situation. For my current purposes I'm not using any OS at all. But Linux support could potentially be great; for example, if you were using a platform with an Ethernet interface, you could use the Linux networking code; if you were using a platform with USB, you could use Linux USB code and file system code and so on.

      steveha

      --
      lf(1): it's like ls(1) but sorts filenames by extension, tersely
    9. Re:C6X support is surprising by Anonymous Coward · · Score: 0

      In which way is the current GCC broken for AVR?

    10. Re:C6X support is surprising by TheRaven64 · · Score: 1, Informative

      I suspect that this badly-worded Slashdot summary really means that the kernel can now expose the C6x as a device and manage userspace access to it when running on something like an OMAP SoC. Running Linux on the C6x itself would be pointless in the extreme.

      --
      I am TheRaven on Soylent News
    11. Re:C6X support is surprising by Nimey · · Score: 1

      I expect you're right.

      --
      Hail Eris, full of mischief...

      E pluribus sanguinem
    12. Re:C6X support is surprising by TeknoHog · · Score: 1
      No, the C6x can actually run Linux.

      Another intriguing scenario is running Linux on both architectures (ARM and DSP) simultaneously, using shared memory to communicate between them.

      So far, its kernel supports neither SMP nor an MMU, which means it is restricted to running Clibc instead of glibc, and it has a very limited set of applications that can be supported as long as it is missing the MMU.

      --
      Escher was the first MC and Giger invented the HR department.
    13. Re:C6X support is surprising by phozz+bare · · Score: 1

      For my own work on a C6711, I've been using the TI compiler under Wine.

      Why? The Code Generation Tools (free registration required) all have Linux versions.

  8. Great! by ArchieBunker · · Score: 1, Insightful

    Now how many of these features are out of beta and actually work?

    --
    Only the State obtains its revenue by coercion. - Murray Rothbard
    1. Re:Great! by Anonymous Coward · · Score: 0

      Three. But only in two use-case scenarios.

    2. Re:Great! by Daniel+Phillips · · Score: 4, Informative

      Features not marked "experimental" in the kernel config database are out of beta.

      --
      Have you got your LWN subscription yet?
  9. Great timing by quantumphaze · · Score: 4, Funny

    I just rebooted to apply 3.2.11 :(

    1. Re:Great timing by Anonymous Coward · · Score: 0

      I don't use linux, but I remember reading a while ago that linux introduced a feature to update the kernel w/o a reboot. Does this not apply?

    2. Re:Great timing by MiG82au · · Score: 4, Informative

      Only thing I know of is Ksplice which is a private company that offered special run-time kernel patches. Oracle bought them out and no longer releases the software and the patches for free.

    3. Re:Great timing by gQuigs · · Score: 4, Informative

      I don't use linux, but I remember reading a while ago that linux introduced a feature to update the kernel w/o a reboot. Does this not apply?

      It's not built-in for any major distros yet. It's called ksplice, which is owned by Oracle now. (It is GPLv2)

      AFAIK it has not been mainstreamed.

    4. Re:Great timing by petteyg359 · · Score: 4, Informative

      Ksplice is just a commercial tool that makes use of kexec which has been in the kennel for years. There is absolutely no need for Ksplice yo use kexec.

    5. Re:Great timing by Anonymous Coward · · Score: 0

      Based on your comment, I'm willing to bet that you're not able to reboot often because this machine is in production? If so, perhaps you should be loading a brand new version of Linux right out the gate. Mature is better than bold. Let the dust settle first before attempting something this risky.

    6. Re:Great timing by Anonymous Coward · · Score: 0

      You can also try using kexec (Debian kexec-tools) to speed up your reboots. It's definitely still a reboot though.

    7. Re:Great timing by BitZtream · · Score: 4, Insightful

      Probably never will be either. Its usefulness was always questionable at best anyway, it is a GREAT academic exercise that I'm interested in just cause I've been developing my own 'x64 os' as a learning experience so the tactics they use I like to learn about, from a practical perspective as a system admin, its silly.

      Mission critical infrastructure where you would want continuous availability is running on a cluster which can stand to have a host rebooted for upgrades so live splicing kernels is pointless in those situations.

      ksplice is for people in moms basement who want an uptime long penis, not for anyone who actually needs service availability.

      ksplice is a treatment for a symptom, which has a long list of side effects that are non-obvious to your non-developer sysadmins, which means most.

      Clusters are the vaccination/condom that prevents you from developing the problem in the first place

      --
      Persistent Volume manager for Kubernetes - https://github.com/dwimsey/openshift-pvmanager
    8. Re:Great timing by larry+bagina · · Score: 1

      And apparently, they no longer spam slashdot with k-splice stories, so something good came out of it.

      --
      Do you even lift?

      These aren't the 'roids you're looking for.

    9. Re:Great timing by Anonymous Coward · · Score: 0

      Clusters are the vaccination/condom that prevents you from developing the problem in the first place

      So you're saying ksplice is a cluster-f*ck?

    10. Re:Great timing by GNUALMAFUERTE · · Score: 1

      Really? I feel really stupid then doing it by hand years before ksplice even existed ...

      --
      WTF am I doing replying to an AC at 5 A.M on a Friday night?
    11. Re:Great timing by GNUALMAFUERTE · · Score: 0

      Really?

      Well, explain this to me:

      When you have 8 servers each with 2 PCI-E Quad E1 Digium Cards, handling a total of 248 inbound calls on toll free numbers, with an average of 200 simultaneous channels per server 24/7, how do you cluster that? When you have analog CCTV cameras running into 4 servers each with 16 channels of video, well, how do you cluster that?

      Not everything is HTTP over TCP/IP. Not everything is easily solved with a load-balancing reverse proxy or DNS balancing/failover. Not everything can be clustered. In those situations, and I speak from experience (those two are real-life situations I deal with), not rebooting is real fucking important. And you have two options: either you leave systems unpatched and wait for the next 5 minutes downtime window that might be a year from now, hoping nothing bad happens, or you live patch those motherfuckers.

      If that wasn't enough, I can mention at least an extra 20 cases where clustering isn't an option, and neither is rebooting.

      --
      WTF am I doing replying to an AC at 5 A.M on a Friday night?
    12. Re:Great timing by phantomfive · · Score: 3, Insightful

      What? If you have an 8 server system, and can't handle one of them going down, you're basically screwed, because sooner or later one of them WILL go down. If you want real uptime, you need redundancy.

      And it's doable, you didn't give good requirements, but in your CCTV example, all you need is to store the data on a SAN.

      Basically if your design depends on the fact that none of your servers will go down, then you need a new designer.

      --
      "First they came for the slanderers and i said nothing."
    13. Re:Great timing by quantumphaze · · Score: 1

      It's Arch Linux on a personal Laptop. Arch doesn't retain the previous kernel after an update like other distros causing interesting effects with fuse and adding USB devices. But at least there isn't a backlog of unused kernels eating up /boot like I have with Ubuntu.

      Anyway, I get in the office and run the update, reboot, then refresh Slashdot to see this. It's more of a matter of reloading my apps and Firefox with my typical 50 tabs (I may have a problem).

    14. Re:Great timing by Anonymous Coward · · Score: 5, Insightful

      Really?

      Well, explain this to me:

      When you have 8 servers each with 2 PCI-E Quad E1 Digium Cards, handling a total of 248 inbound calls on toll free numbers, with an average of 200 simultaneous channels per server 24/7, how do you cluster that? When you have analog CCTV cameras running into 4 servers each with 16 channels of video, well, how do you cluster that?

      Not everything is HTTP over TCP/IP. Not everything is easily solved with a load-balancing reverse proxy or DNS balancing/failover. Not everything can be clustered. In those situations, and I speak from experience (those two are real-life situations I deal with), not rebooting is real fucking important. And you have two options: either you leave systems unpatched and wait for the next 5 minutes downtime window that might be a year from now, hoping nothing bad happens, or you live patch those motherfuckers.

      If that wasn't enough, I can mention at least an extra 20 cases where clustering isn't an option, and neither is rebooting.

      The explanation is fairly simple. Your hardware does not meet the redundancy and failover requirements for the uptime expectations you've set. Equipment designed for extensive uptimes and critical services have built-in redundancy. Take your phone "server" for example. We run several types of phone service, and the one most similar to what you describe runs as a pair of servers, each with a primary and redundant connection. If either server fails, or if the connected switches/routers fail, they can failover to the backup hardware without even interrupting a call in progress. When we need to upgrade the servers, we do them one at a time.

      Rebooting always has to be an option. Always. Why? Because sooner or later you're going to have hardware failure, and you'll be rebooting whether or not you want to. Going with your 'head in the sand' approach only means your customers will feel a much greater impact from the inevitable downtime than they would if you'd properly designed your systems in the first place.

    15. Re:Great timing by Lost+Found · · Score: 4, Informative

      Not true. Kexec replaces the whole kernel, which means the system is reset. Ksplice applies and removes patches (security updates mainly) while the kernel is running, which means all the processes keep running as if nothing happened.

    16. Re:Great timing by blind+biker · · Score: 1

      Is there a distro that uses kexec? Is Oracle the only one to have a distro that does (albeit through a commercial implementation)?

      --
      "The agriculture ministry is not in charge of Gundam" - Japanese ministry official.
    17. Re:Great timing by Nursie · · Score: 1

      kexec - great for fast reboots, really annoying to find when you want to switch OS on your multi-boot machine.

      WTF? Wheres grub? How am I back at the gdm login screen so damn fast?

    18. Re:Great timing by batkiwi · · Score: 4, Informative

      "When you have 8 servers each with 2 PCI-E Quad E1 Digium Cards, handling a total of 248 inbound calls on toll free numbers,"

      This one is tricky if you're trunking with the local telco correctly. Your telco should offer a redunancy and rerouting service if you actually have 64 E1s with them.

      "When you have analog CCTV cameras running into 4 servers each with 16 channels of video, well, how do you cluster that?"

      That one's easy. Splitter before the capture card.

      If you care about it, it's capable of being made redundant.

    19. Re:Great timing by KazW · · Score: 1

      The two replies above mine make a good point of explaining this, but your system design is extremely flawed. Also, it indicates to me that you've probably been setup for a failure since day 1 due to your project having an insufficient budget.

      --
      Geeks don't grock information, they grep it.
    20. Re:Great timing by KazW · · Score: 2

      I have at least this many tabs open in Firefox across my tab groups, the "Don't load tabs until selected" option in the general tab is really quite awesome.

      --
      Geeks don't grock information, they grep it.
    21. Re:Great timing by clarkn0va · · Score: 4, Funny

      Ksplice is just a commercial tool that makes use of kexec which has been in the kennel for years.

      HOO LET THE KERNELS AOT! WHOOT! WHOOT, WHOOT, WHOOT!

      --
      I am literally 3000 tokens away from the chaotic crossbow --Stephen
    22. Re:Great timing by Anonymous Coward · · Score: 0

      If you're going to pull that one off you at least have to learn how to spell w00t.

    23. Re:Great timing by Anonymous Coward · · Score: 0

      I feel really stupid then doing it by hand

      You'll feel much smarter when you get a GF...

    24. Re:Great timing by rdebath · · Score: 1

      What!

      Those 315 seconds per year (Five minutes) that you get from the calculation for ''five nines' reliability is NOT a maintenance window! It's the time you have to fix things WHEN something breaks unexpectedly. What you're looking at is 24x7 availability this is a different problem and in this sort of situation preventative maintenance must be designed for with zero downtime even if that PM is physically replacing a machine.

      The five or ten seconds that some clusters take to switch nodes during a crash (or even sometimes a planned node migration) is what gets charged to the 315 seconds of the 99.999% reliability BUT only if it's within the availability times (if that's 24x7 well, that's what you have to design for).

      And BTW, your two examples are pretty easy to arrange a cluster, the phone calls you just bring a new node online and change your front end routers to stop sending new calls to the machine that's going out and start sending them to the new one. A few minutes later the old machine is idle and can be rebooted/removed etc. The CCTV cameras are even easier; they can be switched between frames or by allowing a couple of seconds of overlap. If you haven't got the hardware for that, well, your system isn't designed for 24x7 operation even if it's got 'five nines'.

    25. Re:Great timing by Anonymous Coward · · Score: 0

      So? reboot again, if it takes over a minute, you should replace your OS... or your computer. Unless that's a server and you expect it to have near-100% uptime, in which case you should have scheduled maintenance once a week at least, to apply security patches.

    26. Re:Great timing by rdebath · · Score: 1

      Yes, I look at it in the same way as a libc upgrade. It's all very nice that the system can continue and if you're lucky and the job the system is doing doesn't need the new version of libc but at some point the services and system needs to be restarted because you MUST be able to do a cold start.

      People make mistakes and a cold start during an emergency is NOT the time to discover that someone made a mistake during an upgrade 18 months ago.

    27. Re:Great timing by BertieBaggio · · Score: 2

      I have at least this many tabs open in Firefox across my tab groups, the "Don't load tabs until selected" option in the general tab is really quite awesome.

      Thanks for this!

      --
      If all you have is a grenade, pretty soon every problem looks like a foxhole -- MightyYar
    28. Re:Great timing by Anonymous Coward · · Score: 0

      Services can be restarted while the system is running. The only thing that can't (usually, it's not technically impossible to implement) is init, which is probably static linked anyway.

      I agree with the part about finding out that someone made a mistake, though.

    29. Re:Great timing by jimicus · · Score: 4, Funny

      Ah, slashdot. The place where you can indignantly call someone wrong because they've told you to do something that's impossible, only to have a whole bunch of people who've already done it explain that you, in fact, are the person in the wrong.

    30. Re:Great timing by pegdhcp · · Score: 1

      Even our crappy (pseudo-)government run telco offers 1-N real-time backup for E1 trunks, they even provide geographical redundancy (with no surcharge), if you are willing to distribute your ingress... All CLECs on the other hand are offering SIP trunks, with all the bells and balls attached... Albeit it is a crappy service in general, we prefer incumbent operator's E1 based service, because they know what "toll quality" really means... Moral of the story, if a telco cannot offer redundant E1 or SIP trunks you need to consider changing operator, before starting to think about servers...

    31. Re:Great timing by Anonymous Coward · · Score: 1

      Debian Squeeze (current stable) uses kexec.
      Slightly strange sensation not seeing the BIOS screen scroll by when rebooting after a kernel update.

    32. Re:Great timing by pegdhcp · · Score: 1
      This reminded me a story from ten years ago:

      A world famous telecommunications equipment provider (a company from North America with a rhyming name :) ) claimed to our management suits that their products are five nines, including maintenance window (!). Naturally after a short while in our (non North American) local conditions they started to pop like corn in a hot oven. Their explanation was (and this was somehow implied in colored pieces of papers they provided to upper echelons of our management) that their power bus (provided that supplies connected to different fuses, preferably of different phases of line) is five nines, thus the box is always online, regardless of the status of CPU :) We started to change CPUs, then supplies of boxes started to pop as well. The new explanation was although it was the recommended practice to connect PSs to different phases, the actual frequency shift in our local conditions (50 Hz 220 V) was far above the conditions in their R&D environment (60 Hz 110 V), thus our local level of variations has a greater absolute value than expected. Shocking, mentally and physically as 90 V respective voltage to the real ground manifested on chassis :)

      What I am trying to tell, as an old fart, is that redundancy starts at the bare metal level. Any design on an upper level must be supported by the hardware. Also it is important to know what is being made redundant. You can have multiple Internet connections but if your UPS is not redundant and is located near brown water facilities, you can find yourself in deep shit...

    33. Re:Great timing by Anonymous Coward · · Score: 0

      why does this always happen??? i just bought windows xp and apparently that's already out of date too!

    34. Re:Great timing by Anonymous Coward · · Score: 0

      I congratulate you, sir, for your fantasic neck beard.

    35. Re:Great timing by turbidostato · · Score: 1

      "from a practical perspective as a system admin, its silly."

      That you didn't find a corporate use case for something like that doesn't mean it doesn't exist.

      A modern, server-grade, computer can take up to five minutes to boot up, loading firmwares, waiting for user prompts, etc. This is PITA when you are developing something that requires system reboots, in example, continuous integration of a service that deploys a complete server-something.

      A practical instance is the OpenStack integration environment: booting up "real iron" servers, installing OS by means of PXE and, on top of that, installing and configuring OpenStack itself could take about half an hour... and you want a cycle to be run at each commit.

      A judicious use of kexec (a more or less ksplice equivalent) and LVM snapshots takes this down to about five minutes.

    36. Re:Great timing by swillden · · Score: 1

      What is toll quality?

      --
      Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
    37. Re:Great timing by Anonymous Coward · · Score: 0

      Cisco also uses kexec for updates of its NX-OS powered devices (Nexus switches).

    38. Re:Great timing by Anonymous Coward · · Score: 0

      ^^THIS

      This frustrated the hell out of me for at least 30 minutes the first time I rebooted with kexec enabled.

    39. Re:Great timing by pegdhcp · · Score: 3, Informative
      "toll quality" basically means that whatever path your voice takes thru operator's network, QoS, CoS or whatever the service management method applied to your frames, package, signal etc. the end result would be identical to the performance of a legacy twisted pair cable that phone systems used to be built on. A toll quality connection as a general rule:

      1) Would not require any specific equipment other than a native phone device (assuming that PRI ports are native to the phone system, which is a subject of never ending discussions amongst some old farts like me around here...)

      2) Would not cause any digital disturbance to voice quality like packet loss, jitter etc.

    40. Re:Great timing by armanox · · Score: 1

      I remember seeing it used in OpenSUSE (post-install).

      --
      I'm starting to think GNU is the problem with "GNU/Linux" these days.
    41. Re:Great timing by ScRoNdO · · Score: 1

      Hey this guy is a genius, he'll get free advice on how to properly redesign his infrastructure all while saying everybody else is wrong

    42. Re:Great timing by CAIMLAS · · Score: 1

      There are other reasons to use something like ksplice/kexec.

      You inherit an environment which has been poorly/insufficiently maintained. You're running a 3+ year old distro on 6-8 year old hardware. The machine(s) have not been rebooted in over 200 days, many over 500+ days. You have several options:

      * virtualize them all immediately and deal with upgrading the VMs (timeframe: several weeks to months)
      * replace them slowly with other infrastructure as you figure out what they do and deal with not fully knowing what they do (timeframe: multiple months)
      * reboot them repeatedly to try to get them up to date (assuming the distro supports it), risking POST death of the drives and/or other components

      Now imagine you've got 5, 10, 20+ of these machines. What do you do? Ksplice is an obvious godsend in the last scenario and will greatly speed the 1st. It's not a silver bullet, but by all means, there are times where the option would certainly be appealing!

      --
      ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
    43. Re:Great timing by sjames · · Score: 1

      If the systems are that important, you keep them isolated and running only the critical application. Then you update them during those brief downtimes.

      Similar deal for the CCTV system, or use IP cameras.

      What do you do when hardware fails (not if, when)?

    44. Re:Great timing by Anonymous Coward · · Score: 0

      Maybe he was aiming for WOOF?

    45. Re:Great timing by Hognoxious · · Score: 1

      kexec which has been in the kennel for years

      I hear it's a bitch to set up and it runs like a dog.

      --
      Confucius say, "Find worm in apple - bad. Find half a worm - worse."
    46. Re:Great timing by swillden · · Score: 1

      Thanks.

      --
      Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
    47. Re:Great timing by ChrisMaple · · Score: 1

      In 1985 I was running 78 simultaneous telco channels 24/7, bare metal, on a 6 MHz Z80. Why in the world would you be running an operating system to do that? There are too many things running an OS for which there is no excuse to have an OS.

      (Billing and some overhead was handled by another 6 MHz Z80 running TurboDOS, which could be shut down for hours at a time.)

      --
      Contribute to civilization: ari.aynrand.org/donate
    48. Re:Great timing by Anonymous Coward · · Score: 0

      He is using a bunch of E1 Digium Cards, and 8 servers. He is clearly a cheap ass, that's crap hardware for Asterisk-style cheap setups.

      The proper way of doing that is to have two SBCs and four SIP trunks to connect each of his SBCs to two SBCs on the provider side, over a NGN network. You can do it right even if you're using Asterisk.

      Transparent fail-over on the voice application side is a *lot* harder, but possible (well, maybe it is not possible using Asterisk).

    49. Re:Great timing by GNUALMAFUERTE · · Score: 1

      Sure, if you are just handling the traffic. You where just routing calls, analogue to analogue. Try routing analogue => IP in various codecs, and also running complex IVRs do that do complex DB queries, voice recognition, speech synthesis, etc.

      Also, imagine those applications changing on a daily basis. If you want to keep up with that, you want everything in a flexible Asterisk box. And you'll be doing RAD in PHP, not optimizing things in assembly.

      --
      WTF am I doing replying to an AC at 5 A.M on a Friday night?
    50. Re:Great timing by GNUALMAFUERTE · · Score: 1

      Isolating systems isn't always the answer, when you have people accessing them 24/7 from iPads, phones and PCs.
      IP Cameras suck, and aren't flexible (try changing the lens on an IP camera). Also, good IP Cameras (I'm thinking Axis) cost 10x as much as a similar analogue camera.

      --
      WTF am I doing replying to an AC at 5 A.M on a Friday night?
    51. Re:Great timing by GNUALMAFUERTE · · Score: 1

      My infrastructure is just fine as it is. Thank you very much.

      --
      WTF am I doing replying to an AC at 5 A.M on a Friday night?
    52. Re:Great timing by GNUALMAFUERTE · · Score: 1

      The data IS on a SAN. What I can't cluster is the CAPTURE. Sure, I could double the amount of servers I need and use BNC T connectors to send each camera to 2 servers, that would cause a) a decrease in image quality and probable interference (CCTV cameras usually don't have a strong enough signal to replicate on two systems) and b) duplicate the cost, try convincing my customers about that ...

      --
      WTF am I doing replying to an AC at 5 A.M on a Friday night?
    53. Re:Great timing by GNUALMAFUERTE · · Score: 1

      This one is tricky if you're trunking with the local telco correctly. Your telco should offer a redunancy and rerouting service if you actually have 64 E1s with them. >> Learn a couple of things about Telefonica and Telecom and then you let me know. Those are my only choices as inbound telco. FYI: I'm getting R2 signalling, that'll tell you how obsolete is their infrastructure.

      CCTV cameras don't have most of the time a strong enough signal to support a splitter after an average of 400m run on passive baluns.

      --
      WTF am I doing replying to an AC at 5 A.M on a Friday night?
    54. Re:Great timing by GNUALMAFUERTE · · Score: 1

      There is a limitation you are not taking into account:

      I'm doing ALL OF THIS on R2. That's how limited Telefonica's infrastructure is in this city. Do you think I wouldn't ask for SIP trunks if I could get them?

      All of our international calls are coming in through SIP, and those are properly balanced across all boxes with 2 SIP routers configured with DNS failover.
      You just can't get that for many national calls here.

      You don't know everything about my setup, so ask before you criticize.

      --
      WTF am I doing replying to an AC at 5 A.M on a Friday night?
    55. Re:Great timing by phantomfive · · Score: 1

      If you don't have a redundant server, you need to be prepared for the event that your server goes down. It's going to happen.

      --
      "First they came for the slanderers and i said nothing."
    56. Re:Great timing by sjames · · Score: 1

      If he has people actually accessing the switch system from iPads, PCs, and Phones other than through the VoIP protocols, he's already lost. The attack surface in that case is the switch software (Asterisk usually), so patching the kernel isn't likely a critical update. Ksplice won't do a thing for upgrading Asterisk.

      If the CCTV app is so critical it can't stand the few seconds downtime to move the analog cable from one server to another to do upgrades, that 10x pricetag shouldn't be a problem. OR, as I said, he could keep them isolated and not update the kernel.

    57. Re:Great timing by bingoUV · · Score: 1

      Even if you update a kernel at runtime, how do you test it? Suppose in 5 years lifetime of a system, you patched 20 kernel bugs saving you 20 restarts.

      This creates kernel version 1, 2, 3 .... 20.

      A process was created in version $i , got a signal when kernel version was $j1, another signal when kernel version was $j2 ... nth signal when kernel version was $jn, where n >> 20.

      Just to test signal related issues, this will need an enormous number of test iterations, impractical to do in a lifetime. Include process context switching related tests, file descriptor related etc. and it will never get done within the lifetime of the universe. Cheaper to add 10 completely redundant systems.

      --
      Bingo Dictionary - Pragmatist, n. A myopic idealist.
    58. Re:Great timing by GNUALMAFUERTE · · Score: 1

      It's not that this machines don't ever get rebooted, or that it's common practice to do live updates.

      I usually do update and reboot those servers when possible. But when there are critical updates that need to be applied NOW and I can't reboot, that's a nice solution until I get authorization to properly update the server.

      --
      WTF am I doing replying to an AC at 5 A.M on a Friday night?
    59. Re:Great timing by bingoUV · · Score: 1

      That makes it simpler than my example, 20 becomes 1. Still, testing complexity has increased enough that you wouldn't be able to do the update NOW. Though I agree the live update can be applied once you finish testing which can be before the authorization to reboot comes.

      --
      Bingo Dictionary - Pragmatist, n. A myopic idealist.
    60. Re:Great timing by Anonymous Coward · · Score: 0

      ksplice is for people in moms basement who want an uptime long penis, not for anyone who actually needs service availability.

      That comment suggests that you have little or no real experience in the industry. Yes, in an ideal world important setups would have redundancy but we don't live in an ideal world. A lot of companies don't have the finances to shell out for redundancy. Many others don't understand the need. And there are still more where the IT department doesn't have full control over their budgets, so you have internal politics and interference from cretins from other departments.

      For companies such as the above, Ksplice is an ideal solution to avoid 'unnecessary' reboots.

      For ISPs, Ksplice is also wonderful selling point to attract customers. Most customers start off small with only a server or two. Ksplice helps keep them happy which encourages more business from them.

      From a sysadmin's POV, Ksplice is a wonderful way to save time. From personal experience managing 200+ customer servers and VMs (including Xen servers), Ksplice removed the major headache of having to arrange and manage reboot schedules with customers (and trust me, a lot customers shitty about such things).

    61. Re:Great timing by rdebath · · Score: 1

      SOME services can be restarted, For example 'REST'full things like web servers. But for many services state is held in memory so 'just' restarting the service is almost as bad as restarting the whole box. Even with web servers it's likely that users will get logged out. If you're lucky the in memory stuff is 'only a cache' and the service "only runs slow for a little while". This is why 'live migration' is popular in some circles, where a service or virtual machine moves between physical machines without interrupting any service state, network connections or running tasks.

      On, and BTW; most Linux init like programs can be restart easily, just send them the right signal and they 'exec' themselves. Of course they do have state, which they have to save for the new instance of the process.

    62. Re:Great timing by snadrus · · Score: 1

      I'm interested in those 20 other cases. I work with mainframe hardware which can have virtually every piece replaced while the OS is still running (piece by piece). OS uptime is measured in years. Kernel patches aren't extremely important as AIX started with BSD which has a great security record & most of the apps are very safely written by the company that uses the mainframe. Systems like this still exist, but clustering has taken the vast majority of their need.
      Heartbeat pings for takeover actions work well. So does doing everything 2x-3x and later cleaning up, which is how mainframe accuracy works.
      For multi-stream traffic like http & voice, you stop routing to a maintenance target & wait for the sessions to end worst-case. You must have extra capacity.

      --
      Science & open-source build trust from peer review. Learn systems you can trust.
  10. Btrfs by Anonymous Coward · · Score: 0

    Bit rot filesystem?

    1. Re:Btrfs by unixisc · · Score: 1

      They should start by calling it btrfs1, since everybody knows that there will be btrfs2, btrfs3 and so on

  11. Re:Why Should I Care? by Anonymous Coward · · Score: 0

    Windows 8, I have seen the demo's and I just could not stop laughing, "charms" "tiles" what will they think up next. I am happy for you, enjoy it if it is your thing, but I hope you do not mind me paying scant attention to your opinions on my operating system.

  12. Re:Why Should I Care? by andrew3 · · Score: 2, Informative

    Linux is a kernel, not an operating system.

  13. Android Window by markdavis · · Score: 2

    It does appear this means the possibility of running of an entire Android "system" and "apps" under a normal Linux desktop/laptop/tablet, but without emulation. Correct? If so, I can see that being a great thing.

    1. Re:Android Window by phantomfive · · Score: 2

      Yes, you are right. You can now run Android on a stock kernel.

      --
      "First they came for the slanderers and i said nothing."
    2. Re:Android Window by Anonymous Coward · · Score: 1

      This is not what it means. It just means that kernel developers have accepted android-specific changed in the main source tree, thus sharing development with the world instead of being a google-closet product that needs to merge back changes made by kernel dev at every release (aka maintenance hell).

      The android kernel was always compilable for x86. The dalvikvm too. There's a couple of full android releases for x86/x64, the reason they aren't widely used is simply that android on a real computer is pointless and frustrating. Maybe things will change with android 5.

    3. Re:Android Window by unixisc · · Score: 1

      Android on a real computer would be something like a Windows 8. Instead of mucking w/ that, if Google is interested in desktop OSs at all, they should just keep working on Chromium OS. And unlike w/ WOA PCs, in this case, regardless of whether a PC has an x64, an ARM, a MIPS or a PPC, having Chromium OS on all those won't be confusing, since they have the option of either having FAT binaries, or using Java bytecode or its Dalvik equivalent.

      Also, the Android stuff is mainly going to be ARM binaries, so at source level, that's what Linux would be getting. Android on x86 is pointless - Android should remain there for ARM and MIPS. Windows 8 otoh would be perfect for x64 tablets, given that there is the perception issue that it needs to run Wintel apps.

      As an aside, anyone knows which windowing system Android uses?

  14. Re:Is it sporting the new Metro UI? by unixisc · · Score: 1

    Hey, you can get Unity or Gnome 3!

  15. I suppose I have to start building... by FlyingGuy · · Score: 1

    my own kernel, again. --sigh-- Or at least no more kernel patches until I get a change to review just how much cruft got shoved for Android Support. Fucking Google.

    --
    Hey KID! Yeah you, get the fuck off my lawn!
    1. Re:I suppose I have to start building... by Randle_Revar · · Score: 4, Informative

      It seems pretty clear stuff is not just being shoved in willy-nilly for android. There have been many debates about including this piece or that piece, and if the implementation should be identical to the android version. Many parts are not in yet, and some may not go in at all. The android suspending solution may not ever go in, mainline may eventually get a system that serves the same purpose in a different way, android may eventually support that. LWN and the LKML posts they link to give a pretty good overview short of reading all the code commits.

    2. Re:I suppose I have to start building... by grouchomarxist · · Score: 3, Informative

      The lwn post is here: https://lwn.net/Articles/472984/

      There is a lot of things they're leaving out for the time being.

    3. Re:I suppose I have to start building... by FlyingGuy · · Score: 2

      Well... Having taken a brief glance through the 3.3 patch file and the LWN posts I am really disappointed, yet again, that google thinks their code is special. The ashmem code is pretty much a duplicate of existing aync shared memory calls that can associate handles to memory which ashmem cannot. Wavelocks are just god awful but the "possible" upside is that perhaps they can be transmuted into something that makes power management a little better.

      The whole damn thing just makes the hair on the back of my neck stand up. If google wants to fork the kernel to suit their own Android agenda that is one thing, but this looks like a back end run to try and supplant the entire desktop ecosystem with the Android user interface which IMNSHO is a bit of a hack. I mean don't we have enough of a problem with the variants in Gnome and KDE already?

      Binders also give me pause as it is again google thinking they have a better solution to IPC then D-Buss or any of the others, when it has the possibility to introduce security holes since it can pass credentials around.

      You gotta wonder what the hell Linus is thinking on this.

      --
      Hey KID! Yeah you, get the fuck off my lawn!
    4. Re:I suppose I have to start building... by Kjella · · Score: 3, Insightful

      You gotta wonder what the hell Linus is thinking on this.

      Well, while he's a hard nail on code quality he's always been a pragmatic man. When it's an interface used on hundreds of millions of Android devices it's something worth supporting if he can do it as long as it doesn't interact badly with the mainline code. And that's exactly why something like wakelocks are still out while others are in. I don't think Linus believes in the one perfect system, if he has to support different IPCs then fine but maybe the implementation can share code and work towards supporting several approaches.

      Remember it's not in anybody's interest to diverge just to diverge, it's just that sometimes it's better to do your own thing and show that it works rather than trying to get permission to change an old recipe. A lot of branches have lived in parallel to mainline and eventually gotten merged in as the real needs and differences - not just the NIH and semantics - have emerged. Getting over these hurdles and keeping the kernel from fracturing into smaller branches that each go their separate ways has always one of the true strengths of the project.

      --
      Live today, because you never know what tomorrow brings
    5. Re:I suppose I have to start building... by Anonymous Coward · · Score: 0

      I was very surprised to see that the arrogant Android assholes actually managed to work with the kernel community this time. They still keep their stupid wakelocks hack even though Nokia has shown several times with the N900 and the N9 that it is completely unnecessary.

      --
      mchurch

  16. Bufferbloat by Wonko+the+Sane · · Score: 2

    I've been reading for a year about bufferbloat and all these tools designed to mitigate it but none of the explainations make sense to someone who isn't already a traffic control guru.

    Can someone explain how, if I'm using a typical Linux system as a firewall between my LAN and a cable modem, I should reconfigure that system if I want to not experience bufferbloat?

    1. Re:Bufferbloat by Maow · · Score: 2

      I've been reading for a year about bufferbloat and all these tools designed to mitigate it but none of the explainations make sense to someone who isn't already a traffic control guru.

      Can someone explain how, if I'm using a typical Linux system as a firewall between my LAN and a cable modem, I should reconfigure that system if I want to not experience bufferbloat?

      Note that I am in no way a network guru / expert, etc. so take my comment with a large dose of salt.

      That said, I don't think there's much you can do in a home environment to mitigate buffer bloat, it's when large ISPs, or other large networks, and backbones interconnect, for the most part.

      I'm not going to say much more at risk of being egregiously wrong. I'll just await someone more knowledgeable to jump in and enlighten us both...

      For anyone reading and is interested in the issue:

      Bufferbloat:

      This problem is caused mainly by router and switch manufacturers making incorrect assumptions about whether to buffer packets or drop them. As a general rule,[which?]} packets should not be buffered for more than a few milliseconds. Any more than this can lead to TCP's congestion-avoidance algorithms breaking, causing problems such as high and variable latency, and choking network bottlenecks for all other flows as the buffer becomes full of the packets of one TCP stream and other packets are then dropped.[4] The buffers then take some time to drain, before the TCP connection ramps back up to speed and then floods the buffers again.

      And a link that may show everything I said to be wrong:

      CeroWrt is a project built on the OpenWrt firmware to resolve the endemic problems of bufferbloat in home networking today, and to push forward the state of the art of edge networks and routers. Projects include proper IPv6 support, tighter integration with DNSSEC, and most importantly, reducing bufferbloat in both the wired and wireless components of the stack....

    2. Re:Bufferbloat by Anonymous Coward · · Score: 2, Interesting

      You want to limit your outgoing transmission speed using QoS to be just under your outgoing bandwidth limit. This prevents your ISP from buffering traffic and reduces latency, increasing responsiveness to things like incoming SSH connections.

    3. Re:Bufferbloat by bytestorm · · Score: 4, Informative

      There isn't an easy answer to your question. In general, bufferbloat is when you get latency or jitter issues because some network device upstream of you has a large buffer, which it fills before it starts dropping your packets. The dropped packets is how software relying on TCP is notified of network congestion so it knows to throttle back. Other protocols may be affected differently (you might notice VoIP delay or bad lag on your xbox).

      To combat this, the idea is to limit your traffic in buffers you control which are (typically) smaller than your ISP and modem's buffers so the ISP ones stay empty and highly interactive. In general, this means limiting your data rates to lower than your bandwidth and prioritizing packets by interactivity requirements. The linux kernel additions in 3.3 allow you to set your buffer size smaller for the entire interface with the goal being to reduce the delay induced by the linux router/bridge. It also adds the ability to prioritize traffic and limit buffers by cgroup (which is like a process categorization or pool which has certain resource limits), but this isn't particularly helpful in your forwarding situation.

      For my own QoS setup, I usually use a script similar to this HTB one. It requires some tuning and getting your queue priorities right requires some understanding of the traffic going through your network. A lot of high level netfilter tools (smoothwall, dd-wrt, etc) have easier to use tools QoS tools which may better suit your purposes. Having not used one, I'm not in a position to recommend them.

    4. Re:Bufferbloat by Wonko+the+Sane · · Score: 1

      I use Shorewall to configure packet filtering for me which does some QoS support. It seems simple enough but I'm not sure how to know if it uses or is affected by the new kernel options. I understand packet filtering a lot better than I understand traffic control.

    5. Re:Bufferbloat by complete+loony · · Score: 1

      In a nut shell, TCP was designed to share a limited network pipe by slowing down the transmission rate of any user when packets are unable to be delivered, assuming that each router would drop packets when they ran out of memory and/or available bandwidth. This way people who need a fast response from a remote server aren't slowed down by somebody else's file transfer. Another design idea in TCP allows for a number of packets to be in flight over very long network paths. If you only allowed for one packet at a time file transfers would be very slow.

      So if a network router just queue's all packets indefinitely and never drops them, no application using the network can tell the difference between a full network queue, or a link that is longer or just using slower technology. Of course these days memory is cheap, and manufacturers allocate too much memory to network queue's. Heck even the linux kernel (I believe) defaults to a packet queue length of 1,000.

      For a linux router, the best thing to do is throttle your upstream connection so it is slightly slower than the smallest bottleneck between you and the internet (eg your modem). That way the modem's buffer can't fill up at all. And then make sure your transmit queue is tiny, dropping any packets that don't fit. And then of course you can apply QoS rules to make sure packets that need an immediate response are sent first.

      --
      09F91102 no, 455FE104 nope, F190A1E8 uh-uh, 7A5F8A09 that's not it, C87294CE no. Ah! 452F6E403CDF10714E41DFAA257D313F.
    6. Re:Bufferbloat by Anonymous Coward · · Score: 0

      So if a network router just queue's all packets indefinitely and never drops them, no application using the network can tell the difference between a full network queue, or a link that is longer or just using slower technology.

      It would never be infinite, as TCP also has timeouts. For this reason, the effects of buffer bloat are exaggerated: it should not matter whether a packet is stuck in a buffer somewhere, or if it was just silently dropped. If you have a time sensitive application that needs to be run over a packet switched network, other transport protocols like UDP make a lot more sense.

    7. Re:Bufferbloat by Anonymous Coward · · Score: 0

      You are right that TCP has timeouts, but they tend to be VERY high. So its very safe to say the issues associated with buffer bloat have not been exaggerated at all.

    8. Re:Bufferbloat by bill_mcgonigle · · Score: 1

      Just a note on running the cerowrt distribution (where the bufferbloat research is implemented) - you need to run a wndr-3700v2 (I've been consistently successful getting v2 refurbished units) but now that the wndr-3700v3 is out, that doesn't work with cerowrt - you'll need the wndr-3800 if you're buying new (I haven't seen 3800's available as refurb yet).

      It closely parallels the WRT-54G and then WRT-54GL situation (without Netgear having learned from the headaches Linksys caused us with that one). The wndr-3[7,8]00 does appear to be the heir apparent to the 54G series, though. I'm happy I bought mine.

      --
      My God, it's Full of Source!
      OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
    9. Re:Bufferbloat by petermgreen · · Score: 1

      It would never be infinite, as TCP also has timeouts.

      Your right it won't be infinite but TCP's timeouts are vey long (to cope with networks that have intrinsically high latencies such as more than one sattelite hop) and are adjusted dynamically to try and maximise throughput in the face of high network latency.

      So as the buffer fills the latency will increase and the timeouts will increase with it .Eventually a window size limit or sanity check limit on the timeouts will be hit but not before the buffer contains a heck of a lot of data.

      If you have a time sensitive application that needs to be run over a packet switched network, other transport protocols like UDP make a lot more sense.

      The reality is different applications have to share the same network. If TCP is keeping a several seconds long FIFO full then any other traffic flowing through that FIFO will also be delayed.

      Of course the real soloution is QOS. Let the latency sensitive traffic go first regardless of how long a queue the bulk traffic has built up. The trouble is it's difficult to implement in a fair manner (if the user or application developer does the marking what stops a user or application developer marking everything as high priority? If the ISP does the marking what stops them favoring their own services?) on an open network like the internet.

      --
      note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
  17. Re:Why Should I Care? by Anonymous Coward · · Score: 5, Funny

    You're just jealous that it'll take at least a month after Windows 8 comes out before somebody creates a Metro-style UI for X.

    Two to match the colors.

  18. Power Management by Phantasmagoria · · Score: 3, Interesting

    Any improvements to power management? It pains me that my laptop gets 4 hours battery life when in Windows 7 but only 2 hours when in Linux. In both cases it's just idle with nothing special running in the background. Or is this a problem with the distribution?

    --
    Loban Amaan Rahman ==> Anagram of ==> Aha! An Abnormal Man!
    1. Re:Power Management by TheLordPhantom · · Score: 4, Interesting

      Mostly it is a problem with the video drivers. Especially AMD. The AMD open source drivers are horrendously inefficient. And, in my experience, the proprietary drivers aren't a whole lot better, but even worse, break everything. So I would say power issues are at the distribution level, not the kernel level.

    2. Re:Power Management by Randle_Revar · · Score: 1

      The big regression was fixed this release, I do believe.

      ah, yes: http://www.phoronix.com/scan.php?page=news_item&px=MTA2OTY

    3. Re:Power Management by MurukeshM · · Score: 2

      I am not sure, but I think the kernel power regression bug was patched months ago: http://linux.slashdot.org/story/11/11/11/2036245/linux-kernel-power-bug-is-fixed.. Perhaps the fix hasn't hit your distro yet. Or is this something new? :(
      It used to be (around Ubuntu 9.10/10.04) that Ubuntu got more life than Windows. I'm hoping to see those days come again.

    4. Re:Power Management by Clarious · · Score: 1

      Mostly it is related to bad behaved hardware or incompleted driver. In my case enable rc6 power saving on sandy bridge boost my battery capacity from 3,5 to 6 hours on linux (it is around 4,5 to 5 hours on windows), turn off or reduce wifi tx power helps too. So check your laptop hardware, there might be some that haven't had power management yet.

  19. Re:Is it sporting the new Metro UI? by Anonymous Coward · · Score: 0

    Hey, you can get Unity or Gnome 3!

    Actually, I was just being a smart ass troll insinuating that Linux is becoming a bug-ridden pile of spaghetti code albeit more-so of a few of the more popular distros rather than the kernel itself. Case in point: Unity

  20. Nice to see AOSP code in the mainline kernal. by Irick · · Score: 3, Interesting

    Seriously, they do some good work. I'm excited to see if this fixes sleep on some of the more obscure devices and gives us better power management.

    1. Re:Nice to see AOSP code in the mainline kernal. by multipartmixed · · Score: 1

      Yes.... The AOSP code is simply fableous!

      --

      Do daemons dream of electric sleep()?
  21. Anyone rebuilding their kernel still? by devphaeton · · Score: 4, Interesting

    Back in the Middle Ages (late 1990s through about 2004) I remember us all getting excited for new kernel releases, and then all rushing to download the source and build it. (By 'us' i mean myself and local geek friends, as well as our cohorts on various IRC channels).

    Nowadays with auto-configuring, rolling release desktop distributions being the norm, is kernel building now only done in server room environments and for non-PC hardware?

    This doesn't matter much, I'm just curious.

    --


    do() || do_not(); // try();
    1. Re:Anyone rebuilding their kernel still? by JeremyMorgan · · Score: 2

      Yeah in the middle ages I was one of those rushing to the source and building it, but not as much anymore. I still rebuild it on my personal machine if I know I'll be using it a while, just to squeeze every last bit I can, but I'll readily admit I don't notice the difference in performance at all. I doubt I'll rebuild for this one as I don't see any features that really apply to me.

      As a personal user, I see fewer reasons to spend a lot of time on kernel tweaking and building, not like it was 10 years ago.

    2. Re:Anyone rebuilding their kernel still? by Wonko+the+Sane · · Score: 1

      I still compile mine. I use git to download the sources so that's a lot easier now than the tarball and patch method. Compiling and installing a new kernel only requires a few minutes and then a reboot at a convienient time.

    3. Re:Anyone rebuilding their kernel still? by devphaeton · · Score: 3, Informative

      I was trying to remember the last time I built a linux kernel. It's going to be somewhere in the early 2.6.x series, on Debian Sid. Even in those early days I didn't really notice a difference in performance (unless I was compiling in drivers for specific hardware). The kernel image was smaller, and I knew that that was better, but other than that it all ran about the same. I almost wonder if the performance "increase" I saw back in the 2.2 days was all in my head now. I used to see some performance differences in compiled FreeBSD kernels on my really old boxes (300mhz K6-II with 128MB), but I think the differences have gotten smaller and smaller since 4.x days.

      Like Wonko says, it's not a huge bit of effort to build a kernel. But I don't really see a reason to do it. I should give it a shot just for old time's sake, heh.

      --


      do() || do_not(); // try();
    4. Re:Anyone rebuilding their kernel still? by Clarious · · Score: 1

      Gentoo, duh. I could still roll my own kernel in Fedora to apply some patches that haven't been accepted in the main tree yet, just fix the rpm sources, add new patch, rpmbuild and yum localinstall :)

    5. Re:Anyone rebuilding their kernel still? by drinkypoo · · Score: 1

      I used to be waiting for Linus or someone to fix something that was broken for me. Now I'm just waiting for more performance. I can wait a bit given how astoundingly fast computers are now (to me.)

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    6. Re:Anyone rebuilding their kernel still? by Anonymous Coward · · Score: 0

      I patched and compiled kernels a couple of years ago when the realtime patch set was still mostly not merged into the mainline kernels. Back then it was absolutely necessary in order to get sub-millisecond latency audio.

      Now most of those features and fixes are in the mainline kernel, but most distributions still don't configure their kernels for minimum latency so I still reconfigure and compile on occassion.

    7. Re:Anyone rebuilding their kernel still? by Anonymous Coward · · Score: 0

      After years i just did it a couple weeks ago. I replaced my mobo and Ubuntu 11.04's ancient kernel didn't have a module for the sensors.

    8. Re:Anyone rebuilding their kernel still? by MarkRose · · Score: 1

      It probably wasn't in your head. With the smaller L1 and L2 caches and memory bandwidth in the day, having a leaner kernel image helped.

      --
      Be relentless!
    9. Re:Anyone rebuilding their kernel still? by evilviper · · Score: 1

      Nowadays with auto-configuring, rolling release desktop distributions being the norm, is kernel building now only done in server room environments and for non-PC hardware?

      In the old days, people were compiling their own kernels because:
          A) there weren't bleeding-edge distros &
          B) There were major new features coming out, like USB SUPPORT, latest network cards, sound cards, modems, and more. The change has been in the PC hardware industry, more than Linux... Although it must be said that Linux is now the cutting-edge platform, rather than second-class citizen (where the community had to reverse engineering everything).

      The world of TV-tuner cards is still just like the old days. I had to upgrade my CentOS6 system to a Linux-3.x kernel to get support for my Hauppauge HVR-1250 card (used a Fedora SRPM, though), and still had to patch it myself, just to get the attached remote control working. I'm sure others can chime-in with other spots where kernel upgrades are necessary for hardware they use and need...

      In the changelog, at least those BTRFS changes seem like killer features someone would upgrade for, if they are crazy enough to be using an unstable filesystem.

      Personally, though, I very rarely see the need to recompile kernels for our servers, either...

      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    10. Re:Anyone rebuilding their kernel still? by JeremyMorgan · · Score: 1

      Yup, back then it was faster. Now, it might be in our heads?

  22. 3.3, when did they get to 3? by Anonymous Coward · · Score: 0

    user@debian:~$ uname -r
    2.6.32-5-amd64

    Doesn't apt-get upgrade the kernel or do I have to do something else? I can't be running old Kernels, I'm sure my system will be so much better if I have the newest. Sure, it's an Intel Atom D525 CPU based system, that does nothing but serve pr0n, but that is important!

    1. Re:3.3, when did they get to 3? by Clarious · · Score: 1

      You can use the Backport repo of Debian, they have a fairly new version of kernel and some other software there. Though the last time I tried it fried my system (I was using nvidia binary blob back then)

  23. New "team" network driver by Elrond,+Duke+of+URL · · Score: 4, Interesting

    I am a bit confused with regards to the new team network driver which is going to eventually replace the current bonding net driver. The kernel newbies page says that it is user-space and uses libteam to do its work, but it also says that this new implementation will be more efficient.

    How is this so? As network throughput keeps increasing, it is important to process each packet as quickly as possible. That's why network drivers and the packet filter are in the kernel. Wouldn't moving the new team/bonding work to user-space mean a lot more data for the kernel to copy back and forth between kernel and user spaces? And wouldn't this hurt efficiency? I'm sure the computer can keep up in most cases, but it seems this will require more CPU time to handle the work.

    Just curious...

    --
    Elrond, Duke of URL
    "This is the most fun I've had without being drenched in the blood of my enemies!"-Sam&Max
    1. Re:New "team" network driver by LiENUS · · Score: 1

      Something about a fight with Tanenbaum over microkernels vs monolithic kernels.

    2. Re:New "team" network driver by Technonotice_Dom · · Score: 4, Informative

      The idea I believe is more that userspace is responsible for handling which device(s) are used for transmission and notifying the kernel, rather than being responsible for the sending of packets themselves. If you've got an active/backup bonding setup, it makes sense to perform connectivity checks from userspace which can be flexible and complex, then notify the kernel to switch or remove devices that have lost connectivity.

      The libteam daemon that's in development seems to have a round robin mode planned and I'd hope 802.3ad, but I guess we'll have to wait and see how that works. I'm sure it'll still need kernel support for the bonding implementations, it's just the monitoring and management functions that are being extracted.

    3. Re:New "team" network driver by marcosdumay · · Score: 1

      Linux is pushing every complex code into the userland. I guess it doesn't make much difference anymore, now that our machines have several processors, 2+GHz and context switching optimizations. Ok, you lose all of your pipeline when switching contexts, that's what? 14 instructions? Nobody is counting instructions with that precision anymore, nowadays we optimize algorithms, not representations.

    4. Re:New "team" network driver by Elrond,+Duke+of+URL · · Score: 1

      Ah, okay. That makes a lot more sense. Thanks!

      --
      Elrond, Duke of URL
      "This is the most fun I've had without being drenched in the blood of my enemies!"-Sam&Max
  24. Re:Why Should I Care? by Anonymous Coward · · Score: 0, Interesting

    At least a month before somebody creates a Metro-style UI for X.... and 10 years before it becomes stable.
    KDE still lacks the simplest functions, like a taskbar that is always 100% of the screen width. Annoys me every day.

  25. Re:Why Should I Care? by Anonymous Coward · · Score: 1

    We already have dozens of tiling window managers, try again.

  26. Do you need the new features? by SmallFurryCreature · · Score: 2

    PHP 5.4 recently was released and it has a really cool new feature. So I did all the hard work of finding a ppa (ubuntu user thingy, stop me if I get to technical) and added it and upgraded. That was pretty hard core! Uber nerd!

    Once, kernel features were desperately needed. Now? Meh, they are probably very nice but I can wait for others to test and add them. Everything just works so why risk breaking it?

    MS has the same problem. XP and even more so Windows 7, just works. So how to sell Windows 8? And Linux ain't selling anything so why upgrade on my own when in a few months I can just run upgrade and have it all done for me?

    Maybe I just gotten lazy. I would type more, but need a nap after so much hard key pressing. *Fluffs up cowboyneal for a pillow and cuddles up with his Linus blanky*

    --

    MMO Quests are like orgasms:

    You may solo them, I prefer them in a group.

  27. Re:Why Should I Care? by Barsteward · · Score: 2

    What version if KDE are you running? I've never had that problem.

    --
    "The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
  28. Re:Is it sporting the new Metro UI? by gweihir · · Score: 1

    That is the "Windows world" you are referring to. It is several decades behind Linux, as it incorporates the GUI into the kernel. Something nobody ever though was a good idea form the technology side, but needed by Microsoft so they can claim it is an integrated part in order to push the Internet Explorer.

    --
    Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
  29. Re:Why Should I Care? by lattyware · · Score: 4, Informative

    That's rubbish. I have a triple monitor setup and KDE will happily let me make a panel 100% of any one screen, or 100% of all three (if you wanted to do that for some insane reason) at any orientation.

    --
    -- Lattyware (www.lattyware.co.uk)
  30. Re:Is it sporting the new Metro UI? by gweihir · · Score: 1

    Why would anybody sane use Gnome or unity?

    Hint to the Windows guys (they have never seen this after all): They are Window Managers, not in any way needed Linux components. In any sane OS design, you van use the GUI you like on top of the OS. Windows is one of those historical designs that force you to use a specific one, with all the massive drawbacks that architecture has.

    --
    Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
  31. Only 3.3? by Anonymous Coward · · Score: 0

    Firefox is v 11 already! Get with the times Linux!

    1. Re:Only 3.3? by galanom · · Score: 1

      You do know that Chrome is (unstable) v19 don't you? :D

  32. Re:Why Should I Care? by lattyware · · Score: 4, Interesting

    Well, it just seems highly unlikely that it wouldn't work for anyone else. I mean, it's a pretty basic feature and if I can do it with my pretty unusual and normally troublesome setup (triple monitors are not that well supported, although KDE does a good job), then I'd expect it to work with most people's. My point is, if it doesn't work for you, then it's a bug, so submit a bug report.

    --
    -- Lattyware (www.lattyware.co.uk)
  33. UEFI boot by Unsichtbarer_Mensch · · Score: 1

    Can we AT LAST boot linux without using any boot loaders? I have a 2 year old UEFI laptop and I can't even figure out how to boot Ubuntu on it :(

    --
    Du kan glomma dina ensama stunder, du kan lita paa teknikens under - Wilmer X
    1. Re:UEFI boot by ledow · · Score: 2

      How would you boot the boot-loader?

      That's the point, the boot-loader is custom-made to the computer in question. It's not even as simple as a configuration option enabled by selecting, say, an ARM-build or something - and you have an UEFI BIOS and I don't, so that'd be another config option to select during kernel compilation (and a recompile if you moved machines).

      Literally, the boot-loader is THE lowest denominator when it comes to interfacing with the hardware. It has to find and supply disk access to any and all disks you wish to boot a kernel from using only BIOS calls, load things from disk into memory (which often requires paging tricks), and all sorts. That's inherently machine-specific, which is why all PC BIOS's handle most of that stuff for you and let you load the bootloader of choice for the rest. After the bootloader, that's why you can provide a more standard interface to boot anything from Linux to Windows without them having to know about the machine.

      Go find elilo for your personal need. Or read your manufacturer's instructions on how to manage your UEFI boot sequence.

      But an in-kernel bootloader is like an in-car spare car key. Pointless if you can't get that far without it anyway.

    2. Re:UEFI boot by Anonymous Coward · · Score: 0

      On UEFI systems I've worked on, GRUB was configured for UEFI without any intervention on my part. That's with Ubuntu 11.04 and 11.10, and Fedora 16.

  34. *Ext4* by jones_supa · · Score: 3, Funny

    Ahh, that was a cleansing sneeze.

  35. Freedom of Choice by Anonymous Coward · · Score: 0

    That is the "Windows world" you are referring to. It is several decades behind Linux, as it incorporates the GUI into the kernel.

    Back in the day....most people on the NT team were furious about the integration of graphics drivers into the kernel. Marketing/management wanted to phase out the non-NT-based systems, and their plan was to make a universal NT-based one that still ran games on cheap hardware. It was determined an acceptable tradeoff to do this integration, due to the lack of headless Windows systems of the day which could practically survive a GUI crash and still be useful. [shrug] People in the OS team were mad enough to quit over this, just like when TerminateThread was mandated as a "customer need":

    http://msdn.microsoft.com/en-us/library/windows/desktop/ms686717(v=vs.85).aspx

    If Linux (kernel) is "several decades ahead" of Windows (kernel), I'd frame that by saying it's driven by a relatively open process and not serving a single agenda. Companies like Google *don't* get to push everything they want into the kernel that everyone uses, just because they feel like it that day. Linus may dismiss the "freedom" aspect of it and consider it just expedience...but I think it's more than that...most days. :)

  36. Re:Android? As in Google? As in NSA spyware? by semi-extrinsic · · Score: 3, Insightful

    I'm guessing you don't know about SELinux? As in "written by the actual NSA"? Oh shit, it's been in the kernel for almost ten years! Go troll somewhere else.

    --
    for i in `facebook friends "=bday" 2>/dev/null | cut -d " " -f 3-`; do facebook wallpost $i "Happy birthday!"; done
  37. Re:Why Should I Care? by Anonymous Coward · · Score: 1

    My point is, if it doesn't work for you, then it's a bug, so submit a bug report.

    I see. It just sounded weird when you phrased it as "That's rubbish".

  38. You can do it with Suse by Viol8 · · Score: 1

    Or you could with 12 and 13 , not sure about current. A few irrelevant features stopped working but still ran happily.

  39. Re:Why Should I Care? by lattyware · · Score: 2

    OK, I admit to careless reactionary phrasing, but still, the point stands. The phrasing of the original post implied that KDE 'lacks the simplest functions' - which is untrue, hence the rubbish comment. The feature is there, and if it doesn't work for them, that's a bug, not lacking the feature itself.

    --
    -- Lattyware (www.lattyware.co.uk)
  40. Re:Is it sporting the new Metro UI? by Anonymous Coward · · Score: 0

    What a bunch of Sheldors. Every machine in our company that is not OSX is Linux or at least some flavor of *nix, thanks for 3.3 Linus et al. I was being facetious folks.

    Actually Unix was alive and kicking when Bill was just an itch in his daddy's pants and also when he & Company purchased the software that was to be bastardized into.....I mean to become 'Windows New Technology Technology' (not a typo remember that NT stands for New Technology), which is why it bothers me when I see crap like Unity sharing the same hard drive with my Nixes. Windows will only become a leader when they actually lead rather than copy someone else's ideas, and copied quite poorly I might add. I still think the best platform for MS Windows was, is and always will be 35mm Slide.

    Any real nerd worth his salt can operate an OS without a stinking GUI, the OS should be able to operate without GUI also. I found much humor in M$' announcement that the new Windows Server will be a command-line only OS with an optional GUI for the poser mouse jockeys.

  41. Re:Why Should I Care? by Sique · · Score: 4, Informative

    According to my university lessons, the kernel and the drivers are the operating system, and everything else is shell and applications.

    MS Windows should thus be considered a distribution (combining OS, shell and applications and an install mechanism).

    --
    .sig: Sique *sigh*
  42. Re:Why Should I Care? by Anonymous Coward · · Score: 0

    Fair enough.
    Sorry for my initial offensive comment -- for me too the taskbar fails to resize properly when changing screen size (in Virtualbox, so this happens pretty often, annoying), that's why I was felt the urge to be a bit hostile towards your "that's rubbish".

  43. Linux still wont do RevoDrive3 without hacks by elucido · · Score: 1

    I've managed to get it up and running (am using it now), but I cannot upgrade the kernel because it only works on one version of the 3x kernel. When are they going to include these drivers by default?

    Then there's the video drivers which don't work well either.

  44. Re:Why Should I Care? by Lumpy · · Score: 1

    "useful idiots"

    There is no such thing as a useful idiot unless that idiot is a fountain of money.

    --
    Do not look at laser with remaining good eye.
  45. Re:Really an alternative? by mcneely.mike · · Score: 0

    --Stares at screen--

    --Jaw drops open--

    --Says, 'WTF?'--

    --Moves on--

    --
    soylentnews.org Go there to enjoy the people!
  46. Re:Why Should I Care? by turbidostato · · Score: 1

    "There is no such thing as a useful idiot unless that idiot is a fountain of money."

    AKA "average consumer".

  47. Re:Why Should I Care? by Tsingi · · Score: 2

    NO, Linux is a kernel.

  48. Re:Why Should I Care? by Anonymous Coward · · Score: 0

    That's a lot more reasonable.

  49. Is this the power regression fix? by mark_reh · · Score: 1

    Matthew Garrett (1):
                PCI: ignore pre-1.1 ASPM quirking when ASPM is disabled

  50. Re:Why Should I Care? by red+crab · · Score: 1, Informative

    OK, I admit to careless reactionary phrasing, but still, the point stands. The phrasing of the original post implied that KDE 'lacks the simplest functions' - which is untrue, hence the rubbish comment. The feature is there, and if it doesn't work for them, that's a bug, not lacking the feature itself.

    I do think its a problem if one tries to add a new empty panel after deleting the default. KDE pulled out the resize button since the 3.x release; the panel won't occupy the full width until you add enough widgets on it; which i suppose indeed is very annoying. This isn't a bug, it appears they want it to work that way. There's more to the list of KDE stupidities, you cannot drag a widget in the panel to change its position; cannot add a desktop icon for your custom binary or script etc. Compared to GNOME 3 insanity though, KDE is still a very usable desktop.

  51. Re:Why Should I Care? by Anonymous Coward · · Score: 0

    Is it the third monitor that isn't well supported, or just multiple monitors in general?

  52. Side effect by shiftless · · Score: 1

    I tried that, but it broke PulseAudio.

  53. Re:Really an alternative? by marcosdumay · · Score: 1

    My guess is the GP was trying to write a joke based on the old SCO citation that Linux developers couldn't have made everything on their own (implying they stole part from SCO's code, and part from Windows). But whatever the intent was, the GP failed to communicate it.

    Anyway, SCO jokes are as aged as compiling one's kernel.

  54. Re:Why Should I Care? by Pf0tzenpfritz · · Score: 1

    Well, for me KDE (or the way how KDE interfaces with xrandr) mostly fails even with two displays. Might be a problem with ATI's fglrx but it used to work with Gnome 2.x and it worls with with IceWM.

    --
    Oh, the beautiful gloss of greality!
  55. Re:Why Should I Care? by CAIMLAS · · Score: 1

    Really? You've been able to run Android under an emulator on Linux for some time now. Except it actually works, whereas Metro is somewhat a non-starter.

    --
    ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
  56. Re:Why Should I Care? by tehcyder · · Score: 3, Funny

    What's with all this being nice and apologising for perhaps sounding a tad rude? Don't you know where you are?

    --
    To have a right to do a thing is not at all the same as to be right in doing it
  57. Re:Is it sporting the new Metro UI? by Anonymous Coward · · Score: 0

    Do you have anything true to say about Windows in the last five years? Because otherwise you're spitting out lies and bullshit.

  58. Re:Why Should I Care? by lattyware · · Score: 1

    That's weird - I've found KDE to have the best support. You can actually manually assign the size of the desktops (mainly designed for stuff like Eyefinity where the displays are presented to the system as a single display) which may solve the issue for you, albeit at the cost of doing it manually.

    --
    -- Lattyware (www.lattyware.co.uk)
  59. Re:Why Should I Care? by lattyware · · Score: 1

    Nope, duals are really well supported these days, it's the triple that's the problem. XRandR (the good way to do it) only supports dual monitors - so you have to revert to Xinerama (which means no 3D accelleration). It's partly also down to the fact I have to have two graphics cards. I hear good things about Eyefinity - as it presents the three (or more) monitors to the OS as a single large screen - but this isn't availible on low end card (which I use both for cost and noise/heat reasons - my gaming PC is separate). I hear wayland also has much better support for multiple monitors, but that's a long way off. In short: Dual monitors are really well supported. XRandR is great, but it doesn't generalise to three monitors. Also there is the proprietary nVidia Twinview, which again is great for dual setups with nVidia cards. KDE's 4.x support is great, as was Gnome 2.x (3.x has no support for triples and poor support for duals).

    --
    -- Lattyware (www.lattyware.co.uk)
  60. Re:Why Should I Care? by lattyware · · Score: 3, Informative

    You just need to click on the cashew (or right click, panel settings - if there is no cashew, unlock the panel first) then drag the stoppers to change the size. As to dragging a widget, you can do that from the same view by dragging, and you can add an application launcher widget and point it to your custom binary or script. All of this stuff has been in there from the first KDE 4.x builds I used, even the really buggy first ones.

    --
    -- Lattyware (www.lattyware.co.uk)
  61. Re:Why Should I Care? by Anonymous Coward · · Score: 0

    More accurately: 95, Me, 7, 2003 etc should be considered (separate) distributions.

  62. Re:Why Should I Care? by Sique · · Score: 1

    Windows 95 and Windows ME aren't operating systems according to Computer Science. An operating system has to control all resources of a computer, and neither Windows 95 nor Windows ME control the CPU, both not being time sharing systems. An application can get an exclusive lock on the CPU without Windows 95 nor Windows ME being able to get the control back from the application.

    --
    .sig: Sique *sigh*
  63. User space by DrYak · · Score: 2

    you'll need, oh, Dalvik and Android's windowing system which is not X11

    In short: To run Android application, you'll need to run the Android userland.

    The kernel is entirely irrelevant. If you don't know what you're talking about, just shut up.

    Except that, in Android's case, the kernel *is* relevant.
    The Android userland relies on quite a few modification of the kernel (mostyl to handle passing signals around).

    Previously, the only way to run the android user-land, was on a special android linux kernel.
    There was one special attempt to have Android run attop a stock distribution, done by Cannocical, and this didn't went much beyond experimental, because of the massive amount of patching involved. And thus the difficulty to maintain it, each time a new Android version emerges.

    Now the necessary changes (or at least part of them, those needed for the user-land. Those needed for the power-saving are expected to arrive by Linux 3.4).
    So you can either drop any of the latest stock vanilla linux kernel underneath your android system (as long as the vanilla kernel has driver for all needed hardware). And thus it will be much more easy for project such as CyanogenMod to feature the latest possible kernel (instead of an older version, because this was the latest available with the necessary changes.).
    Or it's more easy for attemps like cannonical, to bring support for android user-land attop of a normal distribution (because now the default kernel can also contain the necessary plumbing, without needing as much patching as before).

    --
    "Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
  64. Re:Why Should I Care? by toddestan · · Score: 1

    Windows 95 (and the later Windows ME) were preemptive multi-tasking operating systems. Unless you're considering the DOS mode that you could start them in. You must be thinking of Windows 3.1 and older, or the Mac OSes of the time which were cooperative multi-tasking systems.

  65. Re:Why Should I Care? by cynyr · · Score: 1

    the panel in XFCE will be 100% if i tell it to. not annoying at all.

    --
    All of the above was encrypted with a Quad ROT-13 method. Unauthorized decryption is in violation of the DMCA.
  66. Re:Why Should I Care? by diego.viola · · Score: 1

    We need to stop developing for X and start doing things on Wayland.

  67. Re:Why Should I Care? by diego.viola · · Score: 1

    Patches welcome.

    Also, forget about X, we have Wayland now.

  68. Re:Why Should I Care? by Anonymous Coward · · Score: 0

    DOS wasn't a "mode" for Windows 95 it was the operating system that Windows 95 ran on top of. That is why you can boot into DOS and type "win" to run the Windows 95 DOS program. It is also why it's possible to run Windows 95 on top of DR-DOS and even DOSBox.

    Windows 95 was no more an operating system than DOS Shell or GEOS was.

  69. Re:Really an alternative? by Anonymous Coward · · Score: 0
    I've used linux DOS and it's a lot like MS-DOS, but not as friendly. If you want to run linux, you should probably just use windows and run cmd.exe -- that does everything Linux DOS does but Linux DOS renamed all the command for some reason. Does Microsoft have patents on the command names or something?

    Linux is ok for cell phones and toy computers, but if you want to do any work, yu should probably stick to Windows.

  70. Re:Why Should I Care? by unixisc · · Score: 1

    How mature is Wayland - does it come standard w/ any distros? I know that KDE5 is supposed to have improved support for it, but how about the others - gnome, gnustep and so on? I'd really like to see KDE and GNUSTEP on Wayland, and Wayland itself not only on Linux but on the BSDs, Minix and Hurd as well.

  71. Re:Why Should I Care? by diego.viola · · Score: 1

    Wayland is in active development, 0.85.0 was just released recently and Wayland already supports Qt, GTK, EFL, it also has a compositor called Weston and a few applications. I believe 1.0 is expected to be released this year.

    I'd also love to see KDE and Gnome working with Wayland.

    Some interesting videos of Wayland:

    http://mirror.be.gbxs.net/video.fosdem.org//2012/maintracks/k.1.105/Wayland.webm
    http://www.youtube.com/watch?v=WNXWT3ine7E
    http://www.youtube.com/watch?v=u6Jvdo55RUU

  72. Re:Why Should I Care? by Sique · · Score: 1

    Pre-emtive multi-tasking is not sufficient to fully control the CPU resource. It still requires support from the applications. Basicly pre-emptive multi-tasking created some kind of walled garden, where wellbehaving application were able to share the CPU, but each application was able to break out of the garden and lock the CPU. It is more or less comparable with today's sandbox in contemporary browsers, with the difference that the modern browsers have much thicker walls and security patroulling along the fences.

    --
    .sig: Sique *sigh*
  73. Re:Why Should I Care? by jseale · · Score: 1

    That Metro UI is probably protected by copyright laws and the closest anyone is gonna' come to that is Ubuntu with their Unity UI. Would be cool to see something like that in Fedora or one of the other big distros though.

  74. Nice by luk3Z · · Score: 0

    Nice

    --
    Recipes for USA bankrupt - http://tinypaste.com/0d66f dd = dollar deluge (printed in the infinity)