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

 


Forgot your password?
Close
typodupeerror
×
Linux Software

Linux 2.2.7 Released 146

Posted by CmdrTaco from the get-it-while-its-ripe dept.
MazMart was the first to report that a new stable kernel. Linus decided to name it 2.2.7- a surprise move that enraged and shocked, but since it was the next available digit we shouldn't be all that surprised. Now if I only I had a T1- hemos would kill me if I seized control of the ISDN for something so selfish as as kernel.
This discussion has been archived. No new comments can be posted.

Linux 2.2.7 Released More

Linux 2.2.7 Released

Comments Filter:

  • Blimey (Score:1)

    by Anonymous Coward
    Blimey, that was quick. Can't keep up with all these patches. What's the average time between official kernel patches?

  • Re:2.2.7? (Score:1)

    by Anonymous Coward
    2.2.7
    ^ ^ ^
    | | |----> Minor revision 7
    | |
    | |---> Even for stable, hence 2.0.37 is stable, | 2.1.129 is unstable, 2.2.7 is stable
    |
    |
    |---> Major Version 2

  • NT is only on sp4 (Score:1)

    by Anonymous Coward
    How come NT has had 4 patches in 3 years and we're on 2.2.7 already?

  • Re:Blimey (Score:1)

    by Anonymous Coward
    Let's just say it varies greatly. Anywhere from a few days (i.e. the past few patches) to months on end. (Time from 1.2.13 to 2.0.0). If you think the stable patches come quick, wait till you see what pours in once 2.3 gets started (just think, it's been nearly 5 months since the code freeze on 2.1, how many kernel developers are sitting out there just itchin to incorporate their new code into the kernel). Open Source Rocks!

  • Archives, where? (Score:1)

    by Anonymous Coward
    Where are the linux-usb archives?
    There's nothing on linuxhq, and I didn't find anything on dejanews or altavista...

  • ISDN? No wonder... (Score:1)

    by Anonymous Coward
    Rob,

    Get some funding! doesn't anyone want to invest a few bucks in /.? i mean, this site is getting stinking slow and even my poor clients ante up for 1.2 megabits. c'mon, join the amazons and yahoos and get this thing moving!

  • Re:Automatic kernel patch distribution (Score:1)

    by Anonymous Coward
    Wow, that's really neat!

    Oh yeah, FreeBSD has had such a setup for a really long time, but I suppose most of you don't know that. FreeBSD uses a program called CTM ("CVS Through Mail") to send automatic updates for ports and the entire source tree for the branch of choice (src-2_2, src-3, ports-cur, src-cur). majordomo automatically sends tree updates to those who want them, and ctm_rmail searches the appropriate mail file for the "patches," and can apply the patches or just store them. The patches can also be send in pieces and reconstructed when all of them are sent.
    A really wicked setup, that has been implemented under FreeBSD for a long time.

    I just wanted to point a feature that probably isn't very well known implemented in FreeBSD.

  • Re:Linux uptime? (Score:3)

    by Anonymous Coward on Wednesday April 28, 1999 @04:50PM (#1911525)
    d00d, you overflowed your jiffies. Jiffies is the count the kernel uses to keep track of time since boot. On intel, with the default of 100Hz, it overflows in about 1.3 years ~ 500 days. You can change the default value, but its probably a bad idea because your changeing how often you processor is scheduled; your system will be considerably less responsive, etc. IIRC, Alan Cox wanted to fix this problem before 2.2 shipped, but Linus decided it wasn't that big of a deal.
  • Umm, NT is a complete OS, not just a kernel.

    NOw, if a *DISTRIBUTION* had 7 releases in a few months then it would be time to complain.

    you cant compare kernel releases to NT service packs
  • Well, this url will work fine once it is updated, and it will show the updates in 2.2.7
    http://edge.linuxhq.com/changelis t.cgi?show=2.2.6 [linuxhq.com]
  • > If you look at NT5, Microsoft is doing daily builds, and with every build, new things break and new things are fixed.

    Eh, that's just a matter of their software development model, not the actual complexity of the OS.

    I had to point that out as I'm applying some knowledge from my Software Engineering class to real life (posting on Slashdot woo!)...

    .Laz
  • This happened to me too. But i upgraded to a new version of sysklogd at about the same time. I think it was the sysklogd that did it.

    -Todd
  • Alan's patches tend to be more experimental and cutting edge in nature than Linus's patches. Often they are patches that need more testing before being submitted to Linus. He also collects the little bugfixes from linux-kernel and so they are all in one place. Linus then takes what he wants from the ac patches and rolls them into the main kernel.
  • no problem here.... mostly debian ~2.1 system(s), netscape 4.51, libc6 version, icewm... no kde, alas, i like lean wm's these days...

    David
  • Isn't it:

    2.2.7
    ^ ^ ^
    | | -> patch level
    | |
    | ---> minor (even == stable, odd = unstable)
    |
    |----> major

    I've never heard of anyone calling the last digit a minor, major.minor.patch is how I've always seen it.

    --

  • USB Support (Score:5)

    by Jordy ( 440 ) <jordan&snocap,com> on Wednesday April 28, 1999 @02:22PM (#1911533) Homepage
    Hmm, the USB support in 2.2.7 appears to be an alternate development by Linus & friends rather than from the Linux USB project.

    Anyone have any ideas why they wrote their own instead of working with Linux USB project's version?

    The Linux USB projects sources seem farther along than what is included in the kernel.

    --
  • Hmm, I haven't worked with procmail lately; I'm assuming it wouldn't be too hard to set it up to automatically compile everything (I have a shell script at /usr/src/linux/makeall that does that) and, depending on exactly how evil you were feeling, rename the old kernel, copy the new kernel, and re-run lilo. And patch /etc/issue and /etc/issue.net to reflect the new version.
  • These people already `have a life'. Linux and computers is their life (and mine). Your life of bars, movie theaters and tv just isn't our life (though there is some overlap for many of us). You obviously use (or at least are reasonably familiar with) Linux to know what 2.0.x has to offer, but it seems you don't get any real excitement from purly intellectual pursuits. Downloading, installing and running latest releases gives us a thrill because we get to see if things work for us, and if not why not.

    As to movies, we'll go to the ones the interest us (eg TPM, Matrix, ST), but I doubt most geeks go to the movies just to get out and `have a life'. I certainly don't. I prefer reading a good SF book (The Star Fraction by Ken McCloud comes to mind), playing doom (a 486 won't cut it for quake) programming, or playing monopoly or scrabble with my wife (a semi repressed geek).

    You sir, have to learn that what thrills you does not necessarily interest anyone else.

    Leave other's their otherness. -- Marion Zimmer Bradly, The Survivours

  • We came

    We saw

    We overflowed their jiffies

  • ftp6.us.kernel.org has it.
  • Posted by chiem:

    Microsoft uses security via obscurity.
    Their bugs become "features."
  • Posted by FascDot Killed My Previous Use:

    'nuff said.
  • Ya know, I think the MBone and all the other multicasting type experiments out there are perfect for distributing kernels/kernel patches.
    Too bad AFAIK no projects are in widespread use, otherwise we could have everyone download their kernels at once with no more load to the net in general than if just a couple accessed.


    Just a thought with disjointed output.

  • But it was actually fixed anyway.

    OG.
  • As the title says, Jiffies go at 100Hz. This is the kernel scheduler frequency. (i.e the bit to the right of the #define HZ in linux/include/asm/param.h) For a 1Ghz machine, jiffies will run out at the same rate (so far as our time is concerned), but the tasking will seem ~10 times coarser to the applications. That said, if the machine is 1Ghz, the overhead of doing 64bit jiffies is minor (or zero if you have a 64bit processor)
  • For the sake of a quick speed hack (i.e. make the overhead a little lighter compared to what was there previously), set HZ to some power of 2 (i.e. 128 or 1024) and use >>'s, 's and &'s rather than %'s. p.s. does anybody know how quickly the PII does mod's compared to adds, shifts and &'s?
  • I haven't gotten hold of the patch yet (and tried four mirrors already) but did Linus include the USB OHCI driver too? No use trying it because it is very incomplete. The version in 2.2.7-pre4 hanged the machine completely.
  • man syslogd

    -m interval
    The syslogd logs a mark timestamp regularly. The
    default interval between two -- MARK -- lines is 20
    minutes. This can be changed with this option.
    Setting the interval to zero turns it off entirely.

  • Re:USB Support (Score:3)

    by Aki Laukkanen ( 1072 ) on Wednesday April 28, 1999 @02:48PM (#1911546) Homepage
    Yes, this has been a prime topic on the linux-usb list for a couple of weeks. You might want to search the archives for the "Alternate USB driver" thread.

    Linus posted the original announcement for the usb-0.01 code with a note that he couldn't understand the UUSBD code and needed something simpler. Well, I couldn't understand it either (which of course doesn't mean much :). Anyway the way I see it, things could go to two opposite directions from here; The two drivers keep on going on their separate paths or that they somehow merge in the future. I couldn't really see any future for UUSBD with the former scenario since Linus controls what goes into the kernel.

    The most likely scenario I'll see is that the internal structure is kept from Linus-USB (enhanced of course as time goes by) and that the good ideas and some of the higher level device class specific (etc.) code from UUSBD will be ported and merged to Linus-USB.
  • Easy enough to do. There's Unix-based Multicast FTP software around.

    In fact, never mind getting it to the users, use this for mirroring! Have all the mirror sites run the appropriate s/w, and gateway onto the multibone if they don't already have a feed, then have the main site push the new s/w over to the new sites as and when it's available.

    That would speed up the availability of the distributions, no end!

  • At the time of writing, it's at

    - And probably at many other mirrors.

  • Microsoft Windows NT 4.0 has had 26 "HotFixes" for critical bugs since October 21, 1996. Combine this with the 4 service packs and multiple versions of vendor released drivers targeted at different subtle revision levels of NT for a confusing mess. In the same time span, Linux has been through 20 stable kernel releases. (2.0.24->36, 2.2.0->2.2.7) with almost every important device drives included in the standard kernel distribution.

  • Patching.. (Score:3)

    by suprax ( 2463 ) on Wednesday April 28, 1999 @01:47PM (#1911550)
    Rob and others, why don't you just patch the source? The last full kernel source I downloaded was like 2.2.1 I think. Since then, I have been downloading the patches, and after a while, I tarball and gzip the full source all up and I have an exact copy as the 13+ meg file. Try it. :)
    --
    Scott Miga
  • both login and telnet can insert the kernel version automatically. (as well as time/date, nr. of users, etc..)

    read the issue, login and telnetd man pages.

  • Complain? (Score:1)

    by gas ( 2801 )
    Debian 2.2 (potato) is updated at least once a day. It's wonderful - you automatically get the latest of everything and needed security fixes.

    Of course, the latest is never thoroughly tested but thats what development versions are *for*.
  • Wouldn't something like NT 2.4.345 be more correct? Second release of NT (or was it just the GUI that changed?), fourth service pack and some extra patches.
  • Because Microsoft has the funding to do major in-house testing, while the testing for linux has to occur through the community.

    Which begs the question why they don't ;)
    Lee.
  • Even is correct...as in 2.*2*.xx

    The second digit is always even in the release versions. That's how we got from 2.0.36 to 2.2.0.

  • If there were a symlink at ftp://ftp.us.kernel.org called 'latest' linked to (duh) the latest stable version of the kernel? Then you could just:

    $ ftp ftp.us.kernel.org

    login

    ftp> get latest

    Any problems with this idea? Sure would be neato, but is it unnecessary? Whatchall thank?

    P.S. Gotta love DSL:
    226 Transfer complete.
    11074355 bytes received in 189.01 secs (57.2 kB/s)

    :)
  • Yep, I'm capped, although not as much as I should be. :) I'm supposed to have a 256kb line, which would translate to approximately 32Kb, yet I regularly get 55-60Kb, and I've seen large downloads as high as 67Kb.

    Still pretty darn neat.....:)
  • > Same happens here, someone has an idea, actually
    > re-inventing the wheel, so-to-say, and "sells"
    > it as "brand new" ...

    > I said it before - Linux sucks

    So do you, Doug Michels ;-)
  • Well, can you get it to happen again? If you just run back to 2.2.6, how is the problem going to get fixed?
  • You can think of SP4 as NT4.0.1341

    ( Or maybe NT4.1.1341 more accurate if you get my drift. )


    Ken Broadfoot
  • Sorry for the mix up.

    ( Like we really care that much )

    Ken Broadfoot
  • No, the /second/ digit designates stability versus development.

    So, 1.0.x, 1.2.x, 2.0.x, and 2.2.x are all stable.
    1.1.x, 1.3.x, 2.1.x, and 2.3.x [the last doesn't exist, yet] are all development.

    It's too bad stupidity isn't painful"

  • pardon my greenery, but what (tf) is a jiffy? i take it there were no phb's around when this word was made?
  • (Presuming source is /usr/src/linux)
    cd /usr/src
    download patches (I like wget ftp://ftp.us.kernel.org/pub/linux/v2.2/patch-2.etc )

    linux/scripts/patch-kernel

    Ta da! All patched.
    /*He who controls Purple controls the Universe. *
  • That 2 always annoyed me. =)

    --
    Kyle R. Rose, MIT LCS
  • I always do stupid things when I'm drunk. "Yes, Billy, don't put parameters in ANGLE BRACKETS when the forum is HTML!!!!"

    Line 3 should be

    "In increasing order of version number, perform the following patch:

    zcat [patch file] | patch -p2"

    Useful additional content: The -p2 option tells the patch program how many /'s to discard in the filenames. In this case, it should discard the first two (and any preceeding non-/'s.) Man patch for more info.
    --
    Kyle R. Rose, MIT LCS
  • I guess I should get with the program. I don't think that script existed when I started compiling kernels.

    It took me almost a year after the appearance of "menuconfig" to start using it instead of "config". =)
    --
    Kyle R. Rose, MIT LCS
  • Yeah, so, for the newbies who've never done a patch, here's the quick patch instruction:

    1. Download the patch(es) -- starting at the version following the source tree you already have -- from the appropriate kernel mirror. Download the .gz version if you don't know whether you have bzip or not.
    2. cd /usr/src/linux
    3. In increasing order of version number, perform the following patch:
      zcat | patch -p2

      where you should replace zcat with bzcat if you downloaded the .bz2 versions.
    4. Then, make (|menu|x)config and you're off!

    --
    Kyle R. Rose, MIT LCS
  • "d00d, you overflowed your jiffies."

    What a great phrase! Let's use it as a slang term.

    "All right already, don't overflow your jiffies!"

    "When I saw httpd was using 300 megs of memory, I nearly overflowed my jiffies."

  • no no no ... it's the second digit in hte version number that indicates whether or not the kernel version is "stable". (i.e. all 2.2.x kernels are rated as "stable" while the next developement series will be the 2.3.x series)

    -Rob
  • I dunno ... I'm still on 2.2.4 ... if the patch rate keeps going like this I'm just gonna wait til 2.2.10 comes out.

    -Rob
  • The minor version number (the middle one) is even, indicating a stable version (ie, all 2.2.x versions are officially "stable"). The 7 is the revision of the minor version.
    Christopher A. Bohn
  • have the same procmail rule call a script that will do checksums and/or pgp signatures .... and rejecting all patches that failed that test then the security problem will only defer to trusting the people who put out the official patches ... and their machines ... and their pgp keys ... hmmm ....... i'd like to see a windows user make service packs auto-install -- no, wait. downloading service packs takes all day anyway ...
  • Ha - (run-on time) If I had a life would I have just finished putting in the new 2.2.7 kernel after finding that the box I was working on didn't have as86 (part of bin86), so I had to go snag that too, at nearly 3am when I have to get up at 7am? I think not.. d'oh! I'm sure there is some really good reason I chose now to upgrade from 2.0.36... I'm just not sure what it was. I'll remember when I reboot that box. so there! If it makes you feel better - I'm not touching my firewall box's kernel. It stays at 2.0.36... I disagree with your comment that any version over 2.0 is adequate - If I remember correctly every version prior to 2.0.36 had some network vulnerabilities. (?) Are you sure you're not just mad that the last time you updated your kernel you forgot to run lilo before rebooting? ;)

  • 2.2.x releases are stable releases. And, assuming that you don't need PCI audio support, framebuffer support, or any of the other nifty multimedia features, or things like OOB protection, or memory-leak fixes, or hosts of other fixes, 2.0.0 should be fine for you.

    And, I don't know how you build kernels, but I, for one, always leave the previous version in /vmlinuz.old with a corresponding entry in lilo.conf, so there's never a need to wipe anything.

    And, just out of curiosity.... When, except maybe when EXT2 was first being toyed with, did upgrading the kernel leave anything in a terrible mess?

    Sir, you need to read the documentation, so that you know the proper procedures for upgrading the kernel,if you haven't learned to keep backups of a component before you upgrade it.

    Let's face it - 640K is enough for anybody. ;)


    The following sentence is true.
    The previous sentence is false.
  • I think he has ISDN at his house. The server is on a much faster connection.
  • ftp.be.kernel.org has it already as well... full tree and patch... but only in bz2
  • > The truly wicked could setup a procmail rule to
    > feed mail from linux-kernel-patch directly to
    > "patch -p1 -d /usr/src/linux" ...

    that would imply a rather large amount of trust, tho... course most people prolly a lot of people just patch everything blind any without making sure it is trustworthy or at least comparing pgp signatures or checksums...

    but if you want to go all the way, you'll need to have that procmail rule also call a script that recompiles the kernel, and upon success of that copies it to the proper place, runs lilo, and reboots the machine.

    quite a nice maintenance free approach to sys-admining... quite a nice maintenance free approach to security breaches too, imho... but i'm paranoid... =)

  • If what I understand about Jiffies is true, then a 100MHz machine is going to run out of Jiffies in 497 days, right? So, if we ever get to gigahertz processing, then Linux will actually have the same maxiumum uptime (49.7 days) that Windows NT has (49.7 days) on ANY machine. I wonder if this is also true for dual or quad processor machines? I haven't had time to read through the source yet, does anyone know?
  • Is this is a bad thing? No way! Linux has upgrades at least once a week, meaning that there is always some new cutting edge feature being worked on. Not all the upgrades are critical however - you don't HAVE to go to 2.2.7 if 2.2.6 or even 2.0.36 is working fine for you. With NT however there is a large amount of time between each update, meaning that problems often go unadressed for an extended period of time. Also when Microsoft releases a service pack it usually makes a huge change to some aspect of the system, and most everyone MUST upgrade to it sooner or later. I would much rather use Linux's frequent updates than be stuck with an inadequate product.

    Not too long ago, there was an ac patch almost daily. Now THAT is bleeding edge!

  • For major linux industry news such as new kernel releases and new distro releases, I'd personally like to suggest that no outside contributor's name be attached to it. I'm referring to such things as "Bonzo was the first to mention that 2.2.2-ac4 has been released", not the "Posted by CmdrTaco" lines. Maybe 150 people post such a news item, or maybe only 15 - I really don't know, but awarding people as the first to report a story that will shortly be all over the major Linux sites anyway is just like people posting "first!" comments - it isn't *that* much of a contribution. News about kernel releases is available through so many mediums - usenet, mailing lists, other websites such as freshmeat.net, etc.

    The types of contributions which should be encouraged are the obscure story of someone in Australia getting a Windows refund, or a Usenet post about writing drivers for some new hardware. Post-worthy articles that receive less than 5 unique submissions by external Slashdot readers are the ones that really deserve honorable mention.
  • yep, with marla gibbs... who has a cameo in "Lost & Found" which I recommend you all avoid.

    W
    -------------------
  • One's jiffy counters use a 32 bit register@100Hz, thus overflowing in 497 days.
    There already exists a patch to make uptimes of >100 years possible.
    It resides here [lwn.net]

    --
  • never have I felt so out of date, until I've realised that my linux box is already way behind (still have 2.2.2), because it doesn't have a net connection.

    On a slightly off topic, can anyone point me to that comparison of linux v nt with the .1 and .01 database comparisons? I can't find that article on /. anywhere...

  • Slashdot is a public forum, not a money-making machine. It does make a small profit, keeping our favourite administrators in leisure time, but at the moment it is not an in-your-face commercial venture.

    If Rob needed more cash, he'd charge more for banner ads, put more of them in, or even go to the bank. Going to a VC is asking for millions, promising big returns. If slashdot was trying to make money off us (in a big way) it would turn off the traditional readership.

    Don't you think so?

    Jonathan
  • Hmph! None of the mirrors I've checked have it.

    Don't they push these things?
  • I think that the most obvious point to consider is that the Linux community is not worried about losing face. If there is a problem, it needs only fixing, not fixing + PR Damage Control.
  • > If slashdot was trying to make money off us
    > (in a big way) it would turn off the
    > traditional readership.
    > Don't you think so?

    No. I read /. for the quality of it. I don't see a relationship between quality and making money. It doesn't have to mean both can't live happily along. I'd still read /. if Rob put a few more banners on his site. But maybe I'm not part of this so called 'traditional readership'.
  • It seems to me that it really doesn't hurt anyone when people get credited... and it just might make someone's day.

    It just might encourage that person to submit in the future; he might submit something obscure yet interesting the second time.
  • It's not -in- the config.in yet for a reason, it is "pre-EXPERIMENTAL." OHCI support is not complete. UHCI support is partially working, but major development is still going on on this driver. Join the mailing list for more info.

    http://electricrain.com/linux/uusbd-www/
  • See http://electricrain.com/linux/uusbd-www/ for the project.

    The list archive is at
    http://electricrain.com/lists/archive/linux-usb/

    Yes, this is not Inaky's driver. There are two drivers under development. The smaller one was sparked off by Linus because we needed something simple & functional in the kernel today for many reasons you can figure out yourself...

    The projects are working together and will likely merge bunches of code.

    BTW, working OHCI support is coming. The stuff in the 2.2.7 kernel is -not- even complete yet (Linus put it in despite a couple hopeful plees from me and others to wait until that worked).

    No big deal though. It's not even enabled in config.in as "EXPERIMENTAL" because we all agreed that it hasn't reached that status yet.

    If you want web pages for code history and latest snapshots on the small driver you'll have to find them yourself from the list archive; its on my DSL line which would quickly /.
  • I think that he was just joking. There may be some special significance, though, tha tisn't terribly obvious.

  • To post a reply to the main topic, you have to press the Reply button at the top of the page. This is in the same row as the Threshold, and comment-ordering drop-down menus. It wasn't immediately obvious to me either :)

  • RedHat 6.0's been out for days now. Hell, I'm already on my second installation of it, and I'm the freakin' King of procrastination! :-P

    Cheers,
    ZicoKnows@hotmail.com

  • Huh?? (Score:1)

    by Zico ( 14255 )

    You do know that the whole internet thing is international, right? In fact, I couldn't even find a U.S. site that had RedHat 6.0 for quite a while -- I ended up downloading it from a site in the U.K. -- ftp://unix.hensa.ac.uk [hensa.ac.uk].

    Cheers,
    ZicoKnows@hotmail.com

  • Actually they release hotfixes [microsoft.com] between service packs. Then they roll all the fixes into a new service pack. I am not too sure on how frequently they post these though.
  • Well, 2.2 is still new as hell... so the masses who now use it find the bugs (all bugs are shallow... blah blah) so for the first little bit after a release as big and nifty as this they tend to climb the versions like crazy, then level out to .. i think its aroudn 4 stable releases a year once all its good.
  • Some people a pretty thick... :)

    It was a JOKE!


  • The Linux kernel is developed by multiple people, however Linus is the only one to add to the "official" kernel. While I'm not sure why he doesn't use RCS (maybe he does), CVS would be pointless, since there is no "concurrent" development done on the official kernel. However, there was talk of him using a much better system for the type of development the kernel does (bitwize or something like that, which is designed much closer to the development model of the kernel). However, I haven't watched the kernel group lately, so don't know where that discussion went.

    This is not a desktop calculator, it is best that all patches get reviewed, not just added and debugged later.

    -- Keith Moore
  • Why? Because NT is such a huge, lumbering beast, its patches are too.

    SP4 for the Alpha was a 48Mb download - and the file is a self-extracting exe. If you have Macs, you need another 3 or so hotfixes, and it still won't work quite right without trashing the MacVolume index and rebuilding it. (They re-wrote SFM (Services For Macintosh) to use the disk for the index instead of memory) That 48Mb basicly replaces a full 75% of the OS - exes, dlls, drivers and other such important items. Also, in order to be Y2K ready, you need 2 other "patches" - Internet Exploder 4.01 and Data Acess objects 2.0. It's a scary thing.

    My personal experience with SP4 is mixed - killed a few machines, saved others. I would like it better if I could just patch what was needed.
  • "Furiously, she stood and glared at me, her jiffies totally overflowing."


    --
    Moodle http://moodle.com/

  • You can patch it but you have to get the 2.2.2, 2.2.3, 2.2.4, 2.2.5, 2.2.6, and the 2.2.7 patch and patch them in order...

    Still better than d/ling the whole thing though

    -Lee
  • i was reading /. when my boss snuck up on me. i totally overflowed my jiffies!
  • Just in the process of upgrading from 2.2.6 now. Its 0:46am over here in Reading, England (near London), but what the hell? I want to be the ***most*** up to date Linux user I know :-) (my copy of RH 6.0 will arrive on Monday - the day of Release! :-)

    periscope.
  • And why they dont release ntkernel updates frequently ? or security fixes ?

    It is because NT doesnt have bugs ? Certainly not... ;)
  • How come NT has had 4 patches in 3 years and we're on 2.2.7 already?

    Because we are better than that. We release security and bug fixes. :)

  • Yeah, that would work too...

    :)

    but i need the tcp/ip fix from 2.2.4.
  • I thought stable versions ended in even numbered digits. Or am I just dumb?

    "Software is like sex- the best is for free"
  • If you send mail to Majordomo@vger.rutgers.edu with "subscribe linux-kernel-patch" in the body, you will automagically receive the latest patch, as soon as it is posted by the Linus.

    If the patch is too big (> ~400k) you will receive a notification and the diffstat output.

    The truly wicked could setup a procmail rule to feed mail from linux-kernel-patch directly to uudecode and then invoke /usr/src/linux/scripts/patch-kernel on the resulting file ....

  • Automatic kernel patch distribution (Score:4)

    by LucaL ( 25364 ) on Wednesday April 28, 1999 @02:30PM (#1911614) Homepage
    If you send mail to Majordomo@vger.rutgers.edu with "subscribe linux-kernel-patch" in the body, you will automagically receive the latest patch, as soon as it is posted by the Linus.

    If the patch is too big (> ~400k) you will receive a notification and the diffstat output.

    The truly wicked could setup a procmail rule to feed mail from linux-kernel-patch directly to "patch -p1 -d /usr/src/linux" ...

  • Small Patch to get USB to compile (Score:5)

    by LucaL ( 25364 ) on Wednesday April 28, 1999 @02:12PM (#1911615) Homepage
    I have just posted it to linux-kernel:

    --- drivers/usb/usb.h.bak Wed Apr 28 21:59:45 1999
    +++ drivers/usb/usb.h Wed Apr 28 22:08:08 1999
    @@ -363,8 +363,8 @@
    void usb_audio_interface(struct usb_interface_descriptor *, u8 *);
    void usb_audio_endpoint(struct usb_endpoint_descriptor *, u8 *);
    #else
    -extern inline void usb_audio_interface(struct usb_interface_descriptor *, u8 *) {}
    -extern inline void usb_audio_endpoint(struct usb_endpoint_descriptor *, u8 *) {}
    +extern inline void usb_audio_interface(struct usb_interface_descriptor *interface, u8 *data) {}
    +extern inline void usb_audio_endpoint(struct usb_endpoint_descriptor *interface, u8 *data) {}
    #endif

    #endif
  • actually... my first question is how do I post post a reply to the main topic, rather than sombeody else's reply?
    Back to the ac patch thing. I'm pretty sure "ac" stands for Alan Cox. But are these patches solely his work? Or does he gather all of the patches and combine them under his initials? Are they included in the 2.2.7 release?
    I'm just an idiot, so forgive me if I don't understand how it all works.
  • That definately needs to be added to the jargon file :)

  • not that thats a bad thing: i think this is a good counterpoint to FUD

  • Linux uptime? (Score:4)

    by Znork ( 31774 ) on Wednesday April 28, 1999 @04:31PM (#1911624)
    Well, a bit on the silly side, I just went through a horrible horrible bug in Linux 2.0.30. After somewhere around 500 days uptime... IT WRAPPED!
    According to last, it went up at:

    reboot system boot Tue Dec 16 22:51

    (That would be 1997).

    Now:
    11:40pm up 22:21, 1 user, load average: 0.01, 0.04, 0.00

    Suddenly I was back at 0 uptime, and my logfiles went through a spasm with kernel error messages ending with this:

    Apr 28 01:56:07 dream kernel: Call Trace: [do_gettimeofday+34/68] [sys_gettimeof
    day+44/112] [system_call+85/124]
    Apr 28 01:56:07 dream kernel: Code: f7 f1 ba 10 27 00 00 89 c1 31 c0 f7 f1 a3 dc
    fa 1a 00 89 c3

    Oh, well, still it didnt panic, altho a load of apps including X went haywire.

    It sorta puts life into a perspective too. Last time I rebooted this machine I was 25. Now I'm 27. Agh.

  • Re:NT is only on sp4 (Score:4)

    by Kenneth ( 43287 ) on Wednesday April 28, 1999 @01:40PM (#1911636) Homepage
    Linux has far more releases than any closed source OS, because the philosophy behind closed source code is to never let your customer see a bug. Obviously this hasn't worked with NT. Open source on the other hand, releases every even marginally working piece of software. This software is then viewed and 'fixed' by a large number of people, then resubmitted to whoever is maintaining that code.

    With closed source software, about 10% of the time is spent writing the code, then the rest is spend debugging. Debugging takes many many worker hours in order to do correctly. This is why there are alpha and beta versions available for some things, but they are still being tested.

    With open source, such things are not worried about. The user has access to the source code, and those thousands or hundreds of thousands of 'eyeball hours' looking at the code for problems, can be accomplished in a matter of days or even hours.

    There have been times during the development of linux that Linus released a new kernel version more than once a day. I would imagine that this happens more with the unstable releases. For discussion on how to tell, see earlier posts.

  • As soon as I load up Netscape 4.51 and hit the bookmarks button it locks thats running with KDE 1.1, YMMV
    I'm back to 2.2.6 :(

Slashdot Top Deals

Term, holidays, term, holidays, till we leave school, and then work, work, work till we die. -- C.S. Lewis

Close