Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×
Linux Software

Taking a look forward: Linux 2.4 89

Posted by Hemos from the check-out-what's-coming dept.
A reader sent us the latest story from Joe Pranevich regarding the Linux 2.4 Kernel. Much like his original article on 2.2, he takes a look at what's changing, and what's coming. (Hopefully by this fall. Hopefully).
This discussion has been archived. No new comments can be posted.

Taking a look forward: Linux 2.4 More

Taking a look forward: Linux 2.4

Comments Filter:
  • In that case you have used up pretty much all of the resource. Your complaint is like demanding that the glass stop overflowing as liquid is added beyond capacity.
    The solution to your problem is to; reduce amount of applications running; increase amount of physical ram on the system; or increase the size of your swap space.
    You can even add swap on the fly if necessary as Linux supports swap files. It is done as follows;
    'dd if=/dev/zero of=foo-swap-file bs=1024 n=ramsize'
    Then as root
    'mkswap foo-swap-file'
    'swapon foo-swap-file'
    and bingo! you have more "ram" to use. It has been done on a system with only 8M ram and more was needed when building some libraries.
  • Im hope that the kernel guys put a little option in the configuration called "KGI drivers"

    thats will rules.
  • Why don't you log in AC? Perhaps because you know you are trolling? Why else do you hide?

    I have no problem with discussions about MS but not on slashdot. Go to microsoft.com or MSDN, I am sure they have press releases. Over here, you are just very delibarately TROLLING.
  • I assume you meant 2.3.12 is working, but it's not. It doesn't even compile on an Alpha (chokes in some structure definitions for the Alpha board architecture definitions).

    I'll look around for information on this... Alan Cox's diary specified that 2.2.10ac12 didn't have a fix for the NFS lockd problems, but I'm having more than that.
  • #1 You KNOW we don't take trolling ACs seriously. If you have something to say at least identify yourself.

    #2 "Microsoft is not just out to make money. They have done a lot of good things for the PC."
    Yeah, we're really thankful for our daily reboots.

    #3 "arcane command line strings are NOT the most intuitive way to use a computer"
    Ever heard of KDE or Gnome?

    As for your other points ... yeah yeah whatever. Thats your opinion and I can respect that, but WHAT exactly are you trying to prove? You know most people here don't agree with you( if you didn't realize that then you have a serious problem) so what are trying to achieve? This is very childish behavior.

    So why don't you run along to www.microsoft.com and have a good time, OK?


  • He certainly has a "right" to do so, but that doesn't mean it makes sense. Try going to the microsoft campus and telling them all how evil they are. You may have a "right" to do so, but a little thought will reveal that this is not the most efficient use of time. The ACs tone was inflammatory and provocative, a well reasoned argument showing the good and bad side of Linux would not be flamed. They asked for it.
  • I'm not sure if you were around when there were very long development cycles, but in case you weren't, the transition from 2.1 to 2.2 took 2 years. That's far too long; people who wanted better SMP support, more hardware support, etc, either had to deal with the fact they were using a development kernel, or wait it out. With a shorter development cycle, this is pretty much alleviated; big changes are less likely to happen (less time), so drivers can probably be pretty easily backported.

    As for major problems creeping in - uh, that's the point of patchlevel releases (as in major.minor.patchlevel). The filesystem corruption - well, as far as I know it's been mainly due to hardware problems, overclocking and faulty ram and the like. In any case, I think I remember Alan saying something about it being fixed.

    Companies want to support Linux (with hardware drivers I suppose)? Great! So just submit the GPL'd driver code to Linus or Alan, get it included, and it'll probably be maintained by some people to make sure that it doesn't break with little tiny changes.

    Oh, you mean non-Free, proprietary, binary module type support?

    In that case, they can say "This module has been tested to work on 2.2.11. If it works on anything else, that's a complete fluke." as Linus has stated, over and over, that he'll not bend over for the companies who are too anal to release their specs or source code. As far as I personally am concerned, they can go straight to hell, because I'm not supporting them by buying their hardware (Creative Labs, are you listening?) unless they release specs (or the driver's source code).

  • that happens but not that often. It is far more frequent on Linux when you have to recompile all your stuff cause some idiot didn't think that keeping binary compability between libraries is an importand thing.
  • Linux is very nice as a server but it's years behind Winas a workstation ...
  • Yes, they're difficult to write. The difference is, they do not have to be written (at least not the journalling filesystem). Stephen Tweedie has been working on ext3fs, the journalling version of ext2, for some time, even in 2.1-days. When it's finished, it'll probably still be 2.3, and it will be rolled in. Just because it hasn't made its mainstream appearance in the kernel until a certain point, doesn't mean that the work isn't largely finished.
  • 1. No more than wake-all, I would wager. Note his comment that "only one process wins", the rest go back to sleep under wake-all. With wake-one, just the one that's going to win wakes up. The rest stay dormant. I don't see how that aids "controlling all network processes".

    2. Where exactly did you see this? I didn't see a note saying "Oh, Linus thought that it would be cool to reroute the network system directly into hardware access."

    Giving up privacy and security for entertainment? All that was listed was more soundcard support, better memory allocation for the soundcards, and support for the "DoubleTalk". I didn't see DVD mentioned at all, and I'm curious how improved soundcard drivers contribute to less security.

    As far as this "If Pranevich is right", and "Mr. Pravenich never once used the words 'privacy,' or 'security,' or 'stability' " angle, may I refer you to the standard disclaimer he posted: " this is a rough draft document, it may be wrong. In fact, it may be very mistaken. It may be choppy, it may have misspelling, it might even break all the syntax rules of the English language. Most certainly it will omit your favorite "pet" change to Linux 2.3 and you may be inclined to send me nasty emails."
    I realize you're a newbie, but reading a preliminary list of changes and then extrapolating that everyone working on Linux is omitting things that are in the OS by sheer design doesn't really make sense.

  • Linux already HAS streams (Score:1)

    by Anonymous Coward
    Linux has had streams for several years now, but it's an add-on because Linus doesn't want to clutter up the kernel. Of course, Linux streams works exceptionally well. If you are insterested you must view the Linux Streams Home Page [gcom.com]
  • Can we have process/thread forking/cloning as fast as with FreeBSD. My P100 with FreeBSD makes an embarassing mess of my K6-233 running Redhat 6.0. p.s. Does anyone know why FreeBSD (and Solaris) can fork so quickly, and Linux not... (p.s. I am not a kernel hacker :-)
    John

  • Re:How about swap/shared memory? (Score:1)

    by Anonymous Coward
    Not the whole story.

    Many Linux applications (Netscape, Kde Window Manager, Gnome, Etc.) allocate memory per application, instead of in blocks for specific purposes within the application. So long as the
    application remains open, memory is not freed. This is not a "leak" because memory is released from swap and elsewhere when the entire *application* is closed. With Gnome and Kde it involves a whole subsystem of applications that share memory.

    For example, large apps like Netscape and Kde Filemanager (when used for web browsing) do not free memory allocated for web pages which have long since been swapped out, until you close Netscape or Kde. Only a small percentage is released in closing a web page. Most remains in the swap, which grows steadily in time even though you close apps and essentially have less memory in use than when you started X with perhaps an xterm and one filemanager window open. The swap continues to grow as one uses such applications, and eventually no matter how much physical ram and hard disk one has he will run out of memory, even if one is very careful in having only a little memory *IN USE* at a time. The stupid swap continues to hoard memory that was allocated long ago but was never freed by the app in question, which should have done so immediately after that memory was no longer needed instead of waiting for the entire app to close (or in many cases, for X to close). I would call that a severe design which Linux fanatics don't want to talk about or acknowledge.

    Instead, it is easier to give lectures to newbies like they are too stupid to realize that there is only a certain amount of memory (physical plus virtual - swap). They know that and do try to keep memory in use to minimum but that does no good given enough time. Even a few hours of heavy
    web browsing, even if only one page is open at a time, is enough to bring Linux to a grinding halt.

    The workaround is to close the application and restart it (in the case of Netscape) but one can't do that with Kde if one is using Kde as the desktop shell and even with apps that aren't the desktop shell one must exit X (shut it down) because X itself hoards much of that memory, since it is the parent of these apps. Why that should be I don't understand but I can assure you that it is the case.

    This means that running Kde for more than a few days if one is using the desktop heavily is iffy. When the swap gets huge, it even often becomes impossible to shut down. The application tries to fee memory buried deep in the swap and can't. Gnome is even worse because it allocates 32 megs of ram right away with *no* apps running (except a Window manager), when using Gnome-Session.

    These are design flaws, partly in the apps and partly in the way Linux allocates and frees memory. X may be to blame, also. I suspect that people who run server functions from Linux boxes start the server processes from a regular terminal and not from an X session.


  • I think we all agree that there's a very good chance that it will blow.


  • I heard a rumor the other day that Dave Miller has some pretty cool ideas about how to multithread the IP stack so that using multiple NIC's doesn't hurt as much as it does now. The MindCraft benchmarks, whilst basically being a MicroShit marketing ploy, did help our kernel guys to find bottlenecks in the kernel when dealing with MP machines with multiple NIC's.


    But it can still be shown that on a uni-processor machine with only one NIC Linux's IP stack is faster than BSD.


    Iggy
  • With the change from 2.0 to 2.2, there were no big changes from the end users perspective (even if everything was re-writen from the ground up), all you needed was a relatively current set of libs, and the transition was relitevly painless. where if 2.4 uses a new file system (ext3, xfs, etc.) this would require a complete re-install to take advantage of the file system. This also may be true with devfs if it is implemented. I realy dont want to see to many tools to convert ext2->ext3 or xfs (remember that fat16 -> fat32 utility for win98 upgrade, I lost many a file system to that tool).
  • Any code where zero equals x equals ten sounds like it already has all the bugs it needs.
    Who said I was advocating Linux, responsibly or otherwise?
    That wasn't senseless Microsoft bashing, that was humourous Microsoft bashing, wherein I turned a troll's words against him/her.
    I assume you aren't the same AC as the troll,but you really should have indicated so.



  • What I personally would like to see is a release cycle with at least 1 year between releases.
    Well, if you're having a problem with the way kernels are being released now, just wait a while until you "know" that the wrinkles you're concerned wth are ironed out, and then get the new release.. it's that easy.
  • I am running the win2k RC1. And I have been running it for around 2 weeks solid. No reboots, no crashes, no nothing. This is while running programs almost whenevr I'm at home and awake. Winamp constantly, IE5 running constantly, eudora, writing to cd's, making mp3's, etc. Not to mention running RC5. This is alot of stress on a system. My linux box is almost comparable. But guess what. I love them both! Yes! As strange as that is, I run both linux AND windows and love them both. Linux is an awesome server (altho I prefer BSD, I'm one of those people without alot of time and rpms save me that time, call me a weenie.)

    I'm actually saddened by the linux community. THe whole spirit of "Free software! Free speech!"is darkened by the "You use microsoft? hahaha, loser!" mentality. Last I checked, Slashdot wasnt a linux only website. Its dominated by linux users, but most of you probably havent even see win2k, and you're already judging. As a matter of fact, I've ran win9x boxes for over a week without BSOD or crashes. THe only problem was programs that like to not give back memory.

    I dont consider myself a member of this linux community because I'd be ashamed to be. The constant microsoft bashing is what's childish, not posting about win2k on a message board.

    Oh, and my wishlist for kernel 2.4?
    USB networking, better Video Capture support, and a better TCP/IP stack.

    /me hugs his Linux box and win2k box
  • 2.x where x is odd are the version numbers used for development versions.

  • A quick reply.. (Score:1)

    by Anonymous Coward
    'grep "^N" CREDITS | wc -l' gives me 281 names. That's for people who are named in the Linux kernel credits and therefore have had code included. I have no clue how many people also work on it, rewrite code more efficiently, fix bugs, post patches, suggest things on l-k...

  • 2.3 is alive and kicking.
  • I don't follow Linux development too closely becuase I don't know enough (yet...) to understand many things, much less contribute. But, does anyone know about how many people actually work on Linux? That is, how many are active and submit code that actually gets added to the kernel? Is there anyone out there who has a fairly accurate count or even a guess?

  • I knewthat if I waited long enough, I'd be able to use my iMac's scanner, printer, Zip, et al. on one of my boxes.

    Only one question remains: Can I use the strawberry mouse and EmulateTwoButtons?

  • Re:DVD Support? (Score:3)

    by pb ( 1020 ) on Friday July 30, 1999 @02:46AM (#1775111)
    There *is* some work on that, there is at least support for a filesystem, but I don't know how they're going to handle decoding the encrypted stuff.

    Take a look at Linux and DVDs [rpi.edu] for more information...

  • What happened to 2.3?

    2.3 is the development version that will become 2.4; all versions where the middle number is odd are development versions (like 2.1). If you're feeling very brave and have backups, you can download 2.3 from the usual places [kernel.org] and play with it. Don't be surprised if it eats your filesystems, pets, grandparents, etc., though.
  • God only knows.. You could try having a look at the CREDITS file in your kernel source, but of course that is everyone who has ever contributed (assuming their patch included an update to the CREDITS file ). I don't know how you'd go about producing stats for active contributors..

  • DVD Support? (Score:3)

    by Fict ( 475 ) on Friday July 30, 1999 @02:40AM (#1775115)
    I didn't see anything in there about dvd support. Any active kernel developers have a word on that?

    ------------------
  • *BSD has the same problems linux does. I don't know if they are working to fix them, i imagine they are, but it's still just as broken.
  • Looking at Linux HQ's Kernel Patch Summary [linuxhq.com] I see that many of the recent patches tp 2.3 have had changes in the net code.

    Check out the Lance Armstrong Foundation [laf.org]

  • I haven't heard this idea voiced yet in all the posts so I thought I'd mention it.

    Let's say we get clever and start releasing new 2.x versions every 6 months or so. As we do this and as more and more features get added the chance of some serious problem creaping in becomes more and more likely.

    Case in point the file system corruption problem that occures in 2.2.10. Maybe the latest prepatch has fixed this.

    The other problem I forsee is that having too many major releases of stable kernels, 2.0,2.2,2.4 etc will make it harder for outside companies to support. That is when they're released within a short period of time.

    What I personally would like to see is a release cycle with atleast 1 year between releases. This would give more time to iron out wrinkles.

    Shawn
  • What was the point of that? Whats the point of coming to a Linux/UNIX centered website to promote windows? People that use Slashdot frequently don't care about windows, nor do they want to hear about windows..So whats the real point in talking about it? You won't change any ones mind or their thoughts on which operating environment they are going to use. So you more or less just wasted everyones time typing that out.
  • Thats is why I get the src.rpm and do a rebuild.
    This always seem to work. Once in awhile I will have to install some dev libs or such, but that is no big deal.
  • How about fixing whatever aspect of shared memory it is that makes my loadavg skyrocket when Netscape or StarOffice slurps up the last of my swap space? It's the only way I've been able to bring linux to its knees lately...
  • ehmmm.. you're being a bit clueless here i think.

    how in heavens do you expect an OS to prevent memory leaks in apps?

    You're right that linux get's in knots when ram+swap is exhausted, but you can prevent that from ever happening by setting limits, check out /etc/security/limits.conf on RH5/6, and set some limits like RSS problem solved.

    good admin is the key...
  • Well, Linus predicted 2.2.0 would come out around October 98, then "by Christmas," and it finally came out in January 99. Based on that track record, I'd predict 2.4.0 will come out around February 2000 or so.

    Linux seems to follow the Microsoft release pattern: add six months to the predicted release date. Then, after a "final, stable" release, release 50 or so patches to fix bugs you didn't find before the release.
  • Okay, Okay, I didn't mean to start a big load of stuff bringing out points of which operating envionment is better..I didn't come on here and say "LINUX IS GOD WINDOWS MUST DIE WOOO!" I simply stated that from what I've seen the general attitude of Slashdot users is mostly pointed twords linux/UNIX type operating envionments, I'm not saying which one is better. I think that Windows and Linux/UNIX are simply tools to get a job done. I don't think that bashing Windows or bashing UNIX is right..They both have features which make them useful to one user or another. I think really choosing an operating enviornment is a matter of personal preference. There is no "right" operating envionment, and a end user isn't wrong because he wants to use winodws. So my first post was simply saying that I thought that telling the slashdot polls that win2k was the far better operating envionment was a waste of time on a Linux/UNIX centered website. I'm not stating that I have a right to tell anyone anything. But what I posted was what I thought, and maybe I was wrong, maybe I wasn't.
  • I don't know how many of you are using Linux in an NFS environment, but for anything more than casual file sharing, the 2.2 kernels and many of the 2.3 kernels can't even keep NFS going between themselves. They seem to make decent clients, but lockd is still broken as of 2.2.10ac12. knfsd still crashes servers sometimes. The user space daemon is fast enough for me (I don't have 40 clients banging on a server), but sometimes clients lose track of handles to the server. On my diskless clients, I seem to do a "mount -o remount,rw /"
    more often than I'm doing any work.

    I know NFS can work... Linux had excellent NFS support for like three kernel versions somewhere in the 2.1 series. I think we should all make Linus use _only_ NFS for all his work for three months; that would get things fixed in a hurry. :)

  • Don't be surprised if it eats your filesystems, pets, grandparents, etc., though.

    This was my main reason for downloading. The perfect crime....

    dylan_-


    --

  • Re:I'd like to see NFS fixed. (Score:1)

    by Anonymous Coward
    Well, ac12 is broken - that's a well known fact. 2.0.12 is working, I believe. There was a small thread, 2 messages IIRC, about which kernels work best for NFS on linux-kernel. IIRC this was during the last 24hrs.

  • What would be nice in 2.6 (or 3.0) (Score:3)

    by nstrug ( 1741 ) on Friday July 30, 1999 @04:28AM (#1775139) Homepage
    Kernel
    • Faster TCP/IP
    • binding NICs to processors
    • Useable NFS - one that doesn't crash IRIX 6.5 servers and implements NFS v3
    • Guaranteed rate I/O
    • Finer kernel locks (this is being done already)
    Filesystems/Drivers
    • Finer locks - threading device drivers that might benefit
    • Journalling file system
    • logical volume management (maybe in 2.4)
    • Hardware DVD decoders (please!!!!!!!!!!!!!!)
    • Better support for PCI sound cards
  • Well, if you want to go to linuxtv.org [linuxtv.org] and they havbe a proposed hardware decoder with drivers for linux.

  • Re:DVD Support? (Score:1)

    by Anonymous Coward
    I looked into adding DVD to linux but the encryption turned out to be a real problem.

    DVD's are encrypted, and the key is on the disk, but the encryption algorithm is proprietary and can be had for a fee of a mere $10,000. (i've also heard something about a clause in the contract that costs you $1M if your software is reverse engineered... obviously a linux DVD module couldn't be source distributed if this is true)


    Hardware decoding is equally difficult: the interfaces to the cards are proprietary.

  • 2.2 took so long to stabilize because Linux allowed things in after the freeze and it takes time to stabilize two years worth of changes.

    If Linus freezes 2.3 in (say) early October and sticks to it, I think 2.4 should be out by the end of the year.

  • Oh jeez, this is the stuff that I fear will hurt Linux. I love Linux as much as anyone, but if you have ever USED it (extensively) you have to know that ./configure, make, make install doesn't always work. It doesn't work A LOT of the time. All the conflicts between distros and libraries, etc etc. Face it, it's not as easy as double clicking setup.exe.

    Also, the reason I use RPMs when I can is not to make installing them easier...its so I know what is on my system and so I can easily uninstall it.

    I am not a Win user (except at work) but I am not blind enough to think that everything is easy as pie. Linux is fun but it takes some fscking work.

    And to the guy who said a redhat install takes 2 hours. I can easily do it in 20 mins (on a P150).

    Brian
  • At this point the new program over wrote mcf40.dll
    and now all your other apps stop working.
    easy HuH ?
  • Well, all that is true. I've been there, really. But.... With Linux which has a bundled free compiler you still have a choice to fix the problem yourself or find someone who is able to do this for you.
    In most cases all problems compiling Linux sources are pretty minor, unless you get a 2.2.x thingy and try to run it on 2.0.x or do alike nuts.
    And it's much better to have a free program that you can put your hands on instead of spending a heap of money on a commercial to realize later that this freaky beast won't work with this and talk to that and sends a bunch of HTML crap attached to your e-mail etc etc.
    Again, in Linux you still have a chance to bring it up, otherwise you have to decide whether to call a support or to bite own a*s.
    Just IMHO
  • Wrongo. You do not have to be root to install software. The only time it is essential to be root is if you are installing system libraries or packages. In fact you can add your own libraries and applications as a user to define your environment.
    When installing locally:-
    ./configure --prefix=$HOME (or whatever)
    make
    make install
    When done make sure that your path definitions incorporate the new libraries and executables, and all is well. That way you can do your own development and/or run your own applications without interfering with the integrity of the system.
    For example you could within user space define a libc5 development/execution environment without ever installing the libraries and include files as root. Of course you need to have the space available.
  • I don't know why people keep thinking *BSD has a significantly better IP stack. It doesn't. However, the streams based implementation that is used by Solaris is generally a good deal better.

  • Don't forget Firewire (Score:1)

    by Anonymous Coward
    There are some Linux firewire drivers in development. I haven't gotten them to work yet however.

    Linux 2.4 ought to fix scsi as well. Naming scsi disks according to load order is insane.

Slashdot Top Deals

"Why should we subsidize intellectual curiosity?" -Ronald Reagan

Close