Slashdot Mirror
Linux 2.2.1
Myrdraal writes "As the subject says, 2.2.1 is released. It just
has bugfixes for the showstoppers in 2.2.0. I'll
have a changelist on Cutting Edge Kernels in
about 10-15 minutes." Linus announced it as the
"Brown Paper Bag" Release. Cute.
I'd like a Chicken Salad to go with that kernel please!
MAKE IT STOP MAKE IT STOP.
I've recompiled every $#^*@$ kernel since 2.2pre1
I'm getting fucking sick of it =).
I haven't even compiled 2.2.0 yet!!! Now I have to download it again. Oh well, it took me less then a minute on my cable modem...
I love rapid bug-fixing. :-)
What do you know... another buggy kernel release. When will the kernel developers get a clue and actually do a little testing before a mojor release? At least the folks in Redmond have their QA in place. When was the last time you saw the NT kernel patched a day after release?
Heh.
*cackle*
...um, you were being sarcastic yes? If that wasn't a joke then it isn't even worth the effort to respond to that comment.
it just doesn't patch them.
So ... because "the folks in Redmond" take months or years to get bug fixes out (if at all), as opposed to days for the Linux kernel, they have better QA?
Actually that's less to do with QA (for both systems) than it is to do with the frequency of bug fixes. Linux does considerably better (a couple days turnaround!) in that regard.
I'm... Stunned. The 2.2.x series is stable for one... And. You. Are connecting Microsoft with good programming. Or even good QA. Okayokay, fess up, get me some of your drugs & everything will be happy.
-jeff Gondek, gondek.1@osu.edu
j00 aR3 Tru3lY 3l!T3 W1Th J0uR Cabl3 M0D3M! W3 @1l Ph3Ar j0U!
Thank You.
And get back under your bridge, troll.
'at least the folks in Redmond have their QA in place'
Hahahahahahahahahahahahahaaaaa!
'When was the last time you saw the NT kernel patched a day after release?'
Bwahahahahahahahaaaaaa!
This has made my day....
Nice to get kernel fixes within days instead
of months, you dumb flaming troll!
To all you whiners: Live and learn.
Better to sit back and wait
for before always jumping on the very latest
release. Unless you really need a feature or
bugfix in the new kernels, better to site back
and wait. Personally I'm not exactly tripping
over myself to get "better Amiga PCMIA support"
when I'm running an Intel box.
Think before you download.
I thought the rumour was that 2.2.1 would hail the correction of the cycles_t compile-time problem for SPARC builds.... (it still appears to fail on a sun4m)
Is there any additional news on this?
Forget the clue stick, you need the clue anus scope. Stick it up there far enough and you might just find a useful comment. Then again, maybe you won't.
many times
The reason you dont see NT being patched is because the redmond geeks are too lazy to fix their bugs. I have seen bugs last months... some of them were NEVER patched... hmm... Yeah talk about buggy code, eh?
I downloaded the bziped patch and noticed that:
bunzip2 -dc patch-2.2.1.bz2 | patch -p0
did not patch the include/version.h. uname still says 2.2.0-final!
Actually he collects patches too so he is probably better organized than Linus - not technically superior...
So do I. Ph3Ar m3 2!
Holly cow, I wish Microsoft would release bug fixes in a matter of days. At my previous job, the server software we developed was useless on SMP NT4.0sp3 systems due to killer bugs in the COM dlls. And it took them what, only a few months to get around to fixing it in sp4?
It was frustrating for me (until sp4 "magically" fixed the bugs, it was "obviously" all my fault), but it was utterly awful for the customer who's system couldn't keep up with only one CPU.
Our mistake, I suppose, for trying to craft enterprise solutions on Microsoft software.
I think the kernel developers have done pretty damn good, considering that they hardly have the
budget for testing that Microsoft does. WE are the people that test the kernels. If you want to run something stable, run 2.0.36 and wait until 2.2.x is mature enough for the big leagues. Granted you won't have the improved speed, performance, and device support of the new kernel, but you'll have 'stability'. And you're entirely correct, I've never seen an NT patch put out a day after a major release. You're lucky if you see patches for critical systems put out months after a major release. How long did it take Microsoft to get their shit together for OOB, land, and the teardrop family of attacks? Now how long did it take the Linux community? You might want to rethink some of those comments.
to add to the previous response:
It only happens with glibc, althogh the bug itself has nothing to do with glibc, its just that the glibc version of ldd happens to aggrivate the kernel bug.
Seriously, you guys need to think a bit. Trolls aggrivate people because its fun for them to do so! If you only let a couple of people reply to a troll, the troll will go away (in fact, if you do that with most posts, the troll problem IMHO will wither away. Now here's a test:
>Linux.2.2 supports few drivers, and the ones it >drivers it uses are buggy as hell. You cannot
>do anything usefull it, except for talking about >something really cool that you can do while you >jack-off with your friends.
Now if you know you can refute that, then reply..If you can refute it much better than anybody else. But if a number of you should refrain. Take the approach of that stupid Vulcan from that wannabe Sci-Fi TV show (Hint, Hint). Common now, I know you can do it!!
Yes but, it's not that a x.0 release of NT doesn't NEED a patch, it's just that you wont see it released for a few months.... also, as the testing front goes, I seem to remeber a highly promoted Dos 6.0 causing lots of data loss and even a class action lawsuit, this just goes to prove, Redmond has never been all that good at testing either... I'd also like to add, anyone with any experience knows, .0 is always painful.
... and at the same time introduce more bugs and call them features
More to the point: why *don't* ms release a patch for their
OS as soon as a serious bug is discovered?
With Open Source you get the fix almost as soon as the
bug is discovered - with MS you get to wait months while
crackers have a go at your systems
I'm a newbie. I'm running RH 5.2 and have never compiled a kernel myself. Still, I saw lots of news here about the 2.2.0 preX releases, including Linus BEGGING people to test it. Sounded like everything was going cool, only minor fixes needed. Then he's finally satisfied enough to release a final and the shit hits the fan. Now, did Linus add in a major bug after the last PreX? If not, where were all these people, who downloaded the final IMMEDIATELY, during testing?
We REALY should not blame Linus if 2.2.0 was a buggy release. It's only our fault if we didn't tested the pre serie enough.
--- MAINTAINERS.vanilla Fri Jan 29 00:54:12 1999
+++ MAINTAINERS Fri Jan 29 00:54:21 1999
@@ -324,7 +324,7 @@
M: arobinso@nyx.net
L: linux-kernel@vger.rutgers.edu
W: http://www.nyx.net/~arobinso
-S: Maintainted
+S: Maintained
Linux panoramix 2.2.1ac1 #1 SMP Thu Jan 28 12:12:55 CET 1999 i686 unknown
12:57am up 3:01, 9 users, load average: 1.08, 1.17, 1.21
nt
duh!
Yeah but is it doing anything useful. I had several NT Servers just running IIS on approved hardware and ended up having them reboot every night.
include/version.h is generated by the makefile
I thought a troll was a qt programmer :)
After seeing enough taglines with "Linux is the only REAL O/S, everything else is a joke", and the ilk, and comments about how every other OS (funny how some people think that windows is always the "other" OS) isn't as stable, i begin to wonder why.
My BeOS system is doing fine:
Uptime: 9 days 7:49:35 IDLE: 4 day 7:11:50 (46.0%)
But hey, feel free to attack me on irrelevant issues to BeOS not crashing, since that's the style everyone seems to enjoy using in replies.
Is it not trawling?
I took it that the troll came up from under the bridge to make a comment that wasn't thought out
I would have said they were trawling for comments.
Listen, folks. He's trolling. Calm down. It's a troll. He doesn't really believe what he's saying. That's because it's a troll. He's much smarter than you 10z3rz are because he's having a good laugh at what a huge war he started with hundreds of people thinking they're posting something funny and intelligent and creative by saying "NT4 sux!" Calm down. Do something else for a while. No wonder there's so little subtle comedy on TV these days -- none of you loser geek k1dd0z would get it. It's a TROLL. A fucking troll. Get it?
And, yes, I am an Anonymous Coward today -- I didn't feel like taking the time to make an account. So get over it. My e-mail address is below.
P.P.S. If you reply to this, you're falling for a troll. Oh, and CmdrTaco? Enough with the geek shit, okay? We're all impressed with what a Geek you are living in your Geek house with your Geek friends and how you're so Nerdy and stuff. It's not funny anymore.
notatroll - grady@ibm.net
Why don't you shut up and stop yelling against linux. If you love NT so much then don't come here anymore...where people have open minds!!!!!
Yes, and if you'd like to clean up your trillions
of patches, take that clean 2.1.35.tar.gz,
untar, apply all patches and tar it into a 2.2.1.tar.gz.
You'll get all the benefits of downloading a new clean source tree without the pain of
downloading. (Sure, it'll take some time, but
at least no-one else will suffer from you taking up bandwidth).
One more reason is because 'Bull Gate' doesn't want to pay for the work fixing the NT bug!!
Uhhhh.... one word: Teardrop.
Linux had this fixed within a week.
Microsoft took months and months.
I'd rather download 10 kernel patches per day, than get caught bent over with my pants down waiting for any script kiddie with a couple packets to spare to give me a *surprise*.
Reply to "Compile and then talk": If you like being arrogant, please use a different OS, not Linux.
So BeOS can't have problems with USERS rebooting your computer.
I would patch my kernel up everyday if I could. I say bring on the new releases, bug fixes, etc, at least there free.
I have seen a lot of CLOSED-minded statements from Linux bigots here on Slashdot, but this one takes the cake. So, those of us who are trying to learn aren't wanted? Fsck off.
Just so you know, that pretty blue screen with the funny numbers on it?
not a screen saver....
Now, does anyone know why the mirroring system is
sucking so badly? I had to go to russia to get these, they weren't on us or uk mirrors.
Someone needs to invent multicast FTP to get all mirrors updated at the same time.
Er... is it just me or did you just fall for a troll? Wasn't that my whole point in the message you replied to? That trolls are WANTING you to reply to them. Flamebait. Trolls. Calm down.
grady@ibm.net
Oh yeah, physicists are all stupid.
(Yes, that was a *_TROLL_*! Starting to catch on now?)
if your sp4 box runs like shit maybe you should take the time to learn how to use it.
runs on 486 100/64 meg here as quick as my linux box.
quit yer whining and learn how to set-up.
Look at Win98, that needed a service pack the day it was released, and MS knew it. Same thing with 95. They rush buggy software to the market and want you to buy it! Bill needs another Maserati! Buy Windows January 99 now!
Gotta love M$'s marketing, though.
I always thought that linux was based on mutual use and that weird sence of humor we hardcore users understand. if it was not obvious that it was meant for humorous purposes, you need to go out for air!! just sit back, read the messages, laff...ENJOY LIFE, LOVE, LINUX
FreeBSD oWNZ yEW
Windows NTSP4 had a nextday for to correct spooling problems (on some machines)
Change the poll cmdrtaco! :)
---
You could always try using the full mirror names... that way you can find one you like that gets the kernel quickly and keep going back there. I've had good luck with ftp4.us.kernel.org. ftp2 also seems to get them early, but only as .bz2, and I haven't got enough RAM in this machine to decompress those in anything like a reasonable amount of time, so I don't go back to that one any more.
This also solves the ncftp weirdness you get when two hosts with the same alias don't have the same directory structure.
Unfortunately it looks like your patch broke. Look for *.rej files around in your source dirs, and if that fails, just download the whole tarball (from a mirror!). After that use patches. Unfortunately sometimes things go wrong.
Posted by oNZeNeMo:
Well, it's a good thing to have a bugfix released a few days after the kernel. At least the kernel people keep up with problems whereas MS will sit for a few months on bugs in their OSs. I can't complain about the frequent updates. How often has Win95 been updated, once every 6 months or so? It's not like MS OSs have been particularly stable. In fact, most of my devel kernels have been far more stable than any windows machine I've put together.
Posted by Myrdraal:
You need to delete arch/i386/lib/checksum.c and arch/i386/lib/checksum.o. You also may want to upgrade patch which should have deleted checksum.c for you.
-Myrdraal
Posted by Drizuid:
./core or something, in 2.2.0ac1 it has to have filepermissions 600 to work, get more info on it, @ http://www.genocide2600.com/~tattooman
There is a bug in 2.2.0 and 2.2.0ac1 (under certain circumstances) 2.2.1 is not affected by this bug. The bug would allow local users to reboot your machine by attempting to run
I know I'll be patching asap!
Can Linus take the bag off his head now?
Ooh, a sarcasm detector. Oh, that's a real useful invention.
And I thought, gee what if I kicked over the rock here and discovered 2.2.1.
Well there is already a cd in the oven with 2.2.0, so it will have to do.
Its a happy upgrade cycle, because its free.
no worries...
^~~^~^^~~^~^~^~^^~^^~^~^~~^^^~^^~~^~~~^~~^~
ABORTED effort:
Close all that you have.
See http://roadrunner.swansea.linux.or g.uk/clue.html
If you have a arch/i386/lib/checksum.c remove it. That became arch independant in the 2.2.x series IIRC
NcFTP keeps track of the last *real* directory you were in. If you cd to "/pub/linux/kernel", that might just be a symlink to someplace like "/.1/mirrors/kernel", which is what NcFTP ends up saving.
When you reconnect using one of the mirror hostnames, it's actually picking an IP out of the list randomly, which means the "real" directory structure might differ.
Mirrors aren't always updated the instant something is released. It could take 5 minutes or an hour or a day. You're not likely to get any sort of decent download rate since everybody and their grandmother is trying to download the latest full kernel (as opposed to, say, the PATCH)... Might as well wait until things quiet down.
I've never seen an NT kernel patched the day after it was released. Too bad really, 'cause it needed one.
-Derek
P.S. It took me a day to recover from the blue screens of service pack 4!!
I'm not sure whether to go with 2.2.1 or 2.2.0ac2 - the ac patches are bigger.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
Maybe I'll stick with -ac2, unless there's something critical in 2.2.1 that's not in the ac patch.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
do a little testing before a mojor release?
I guess their mojor isn't working....hoho...uhh..mojo....mojor....umm...I'll get back to work now...
dylan_-
--
Igor Presnyakov stole my hat
I'm sorry, it seems we only have Penguin Salad. Fiddlesticks.
You don't get it. When your servers are susceptible to DoS attacks you need a fix NOW not months from now. This has nothing to do with being lazy, it has everything to do with people like linus staying up to make sure YOUR isp does not get hacked into.
What exactly were the problems? 2.2.0 works fine for me, so I'm not touching 2.2.1... I know there was that sound compiling problem... What else?
No, it takes Microsoft months to admit there are bugs, more months to investigate them, and then they decide it's more profitable to add cute animations instead of fixing them, and then they charge for it and call it Win 98.
Just quit bothering with it and wait until Red Hat, Debian, SuSE, Caldera, etc. release a distribution with it on the CD. Otherwise, quit yer complaining and keep compiling them...
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
Well... I'm stalking around on ftp.us.kernel.org as we speak...
.in.linux-2.2.1.tar.bz2.
-rw-r--r-- 1 root root 8654848 Jan 28 21:55
Almost there... I'm sweating... waiting... I can feel the bits flying my way already... anxiously awaiting my next kernel compile...
If I'm always compiling
a) I never have to do anything else, like work
b) I don't accumulate uptimes, so I don't feel obligated to do things like try to move my computer while running by keeping it on the UPS while riding the elevator down to the car, then plugging it in to the lighter....
What was fixed in 2.2.1?
I think you've missed the point of OSS completely. Let me fill you in.
...
1) The main developer or team put together a release. They hammer the crap out of it as best they can, but they are relatively few, and catch only some of the bugs.
2) The rest of the community downloads it and uses it. We rarely do testing as such, but there are bazillions of us, so a lot more bugs get flushed out.
3) The bugs get reported, fixed, and 'round again we go.
It's very simple and very effective, and we're damn lucky to have primary developers that respond so quickly to bug reports and issue fixed releases with an efficiency unparalleled in industry.
And you're complaining that the process works so well. Sigh
"The question of whether machines can think is no more interesting than [] whether submarines can swim" - Dijkstra
30 seconds.
;-)
C-ya
Get my e-mail after a captcha test in: http://tinymailt
When was the last time you saw a buggy TESTED NT kernel release? When was the last time you saw a NT kernel release? When was the last time bugs were fixed in the NT kernel the next day after they were discovered?
Open your mind or go away.
looks like the sound might be a little easyer to configure for soundblaster now though i havn't tried it yet
After compiling 2.2.0 last night, I'd like to warn of another little caveat.
If you use Solaris x86 in a partition before your linux partitions, be prepared to reboot to single user mode (or a previous kernel) to fix your system. I'd guess this only applies if you compile in ufs support and x86 partition table support. What you get is some extra partitions, so I have:
Obviously, hda5 used to be / for linux; now it isn't so linux wouldn't boot. Took a little work (especially as / was mounted ro..) but it's booting now and working fairly well. Now to get sound working again...
Hmm, after watching my friend's computer with NT4 SP3 bluescreen twice, fail to run executables until it was rebooted three times, lock up the whole system five times, and all in the same day, I'm very happy I'm running Linux.
So you can take your NT super kernel and shove it where the Sun Microsystems doesn't shine.
I've seen too many blue screens in my lifetime, and I'm only 19.
And at least have the guts to not be an "Anonymous Coward"
For the love of god...
We "upgraded" to VFP 6.0 last September, where I work, and yesterday a 16MB service patch (spc2 - there wasn't an spc1) was made available for downloading. It only took 5 MONTHS to attempt to fix some bugs, but the major bugs are still chewing holes in the apps. It also seems that it wasn't enough that M$ forced one to isntall their latest version of Internet Explorer (to read the HTML docs) when VFP 6.0 was initially installed, this service pack installs M$'s version of the Java VM. Just more pushing and shoving of the sheep into the M$ corral for more shearing.
Running with Linux for over 20 years!
Let's see. SP2 out the door one day... _HUGE_ problems found (eg. bye-bye hard drive). It was pulled months ago and they still haven't rereleased it.
We are all in the gutter, but some of us are looking at the stars. -- Oscar Wilde
The original poster has a point. We are the last link in the kernel development cycle. If we don't test and report bugs, they are not going to be caught as quickly. As with any product, OSS or not, *.*.0 is at least a bit iffy. If you can't afford to take any chances, wait a while. If you can then by all means *test* and *report*. Note to the original poster: You don't have to be part of some elite crowd to have a valid opinion. If you run into any problems compiling 2.2.1 let me know; I'd be glad to help. Instructions are in your RH5.2 manual if you have the boxed set.
-Steve Bergman
steve@netplus.net
Sure, it takes work to ride the edge, but nobody is forcing you to. This is exactly how OSS improves so just stay at 2.0.x or whatever until things simmer down.
Massive peer review is good, I'm glad these things get fixed.
I use SP3 on my work laptop and it hasn't blue-screened, but it does screw up every few days and require a reboot, and apps frequently lock up and require re-installing (e.g. Outlook yesterday). Even as a workstation OS, NT still has a long way to go - I don't subscribe to the view that it blue-screens all the time, though that can happen on some hardware, but it is not really stable yet.
Needless to say, my home systems, running Linux, have yet to crash, except when a disk controller blew up...
Well try to get a braincell and work out that you don't need to download absolutely every kernel. Get one that works for your hardware and *use* it.
Someone is going to have to change 2.2.0 to 2.2.* in the poll now.
Fixing for instance a DoS hole doesn't take months of testing. What takes Microsoft so long is that they seem to consider it unworthy to release service-packs with a size less than 10 MB's, and thus they have to gather bugs for a couple of months.
We kernel developers at least realise that patches are needed as soon as possible. A server just can't stay down a month because it's vulnerable to some exploit that there is an unreleased fix for.
Ok, so I'm an idiot and i tried it.
It will reboot before you get your finger off the key. Ugly......thank you 'who ever the hell wrote it' for fsck
OMG!!! 2.2.1 is out!!!! I better flame the hell out of the entire kernel team!
Sheesh if you whiners would shut up for 1 second you'd realize that the kernel team is doing an EXCELLENT job!
I think that a release this quickly is awesome and proves that OSS and Linux can kick the rumpus out of any commercially made OS on the planet!
Linus and the Kernel team... You guys are GODS!
Do not look at laser with remaining good eye.
That's why theyre there! Don't keep downloading the whole kernel. You're just wasting bandwidth.
No, it takes Microsoft months to admit there are bugs, more months to release a patch, and then they charge for it and call it Win 98
I tried to compile 2.2.0, but got an error in checksum.c it turns out (from my limited knowledge of c) that one function is declared twice. Anyone know how to fix this....is it in the -ac2 patch or 2.2.1?
thanks
-Mark marks@NOSPAM.magpage.com (you know what to do)
-mark
If your computer says LINUX, run...computers can't talk! [unless you have text-speech software]
And I just finished d/l 2.2.0!
Why you say you no bunny rabbit when you have powder puff tail? --The Tasmanian Devil
We should use "Brown Paper Bag" as our themesong for this kernel. (It's a lot cooler than "start me up")
Thank you for not thinking.
In order for opern source projects to work, people need to grab the source as soon as possible and test it. The developers can't do all the coding and testing by themselves, and testing gives the people who haven't ever written a single line of code a chance to participate.
The more people who grab it and test it, the fewer bugs that there will be in the next release.
So, wait for a later release if you feel the need, but I say grab it and go with it.
They give us good laughs...
-matt
I don't know. I d/led 2.2.0. I had initial problems compiling and some other problems with a bad util-linux rpm (couldn't log in b/c of incompatible shared libraries) but once I returned to my old version of util linux, compiled the kernel with auto kernel module loading or whatever, I've had nothing but improved speed, etc., and no crashes yet. I upgraded from 2.0.34 and I am definitely pleased with the results. The only reason I'm going to get the .1 patch is so I don't have that pesky ldd core bug. :)
-Y
"There is no culture in computer science, only cults." - M. Felleisen
A little jealous are we?
And lose the damn d00d speak -- it's freaking annoying.
And at least have the guts to not be an "Anonymous Coward"
Oooh, is that why you use Hotmail? Is your last name really M?
(NT == no testicles == eunoch != UNIX)
-- Blame any errors on your own stupidity. All wrongs reserved.
> Maybe we should be the ones wearing the brown
> paper bag... after all **WE** are
> the ones who failed to identify the bug
> prior to 2.2.0, eh?
Heh. I guess so, though by that logic EVERYONE IN THE WORLD should be wearing a paper bag right about now.
It might be interesting to see all those people bumping into each other, but I guess it wouldn't add that much to the Linux development effort.
Aw, who am I kidding. Bring on the bags!
Even if NT 4.0 did what it was billed to do, and even if you could claim that MS gets it's bug fixes out (I remember an extremly buggy 95 original and 3-5 MONTHS for a service release, NT sevice pack 1 took several months and even at SR4 it only tenitivly makes the C2 rating it claims out of the box). Your forgeting, IT'S FREE, and the people who support it do it out of the goodness of their hearts, for us and The Industry as a whole.
Spyder
So when do we get a jpg of Linus wearing it ;-)?
ps. I don't think Linus deserves anything but praise and thanks for 2.2. Linux is moving into 1999 with a bang!!!
>When was the last time you saw the NT kernel
>patched a day after release?
When was the last time you heard of Microsoft being ABLE to fix a bug within 24 hours?
-ck
-- This sig is only a test. If this were a real sig it would say something witty. --
Q. How many Open Source Software developers does it take to change a lightbulb ?
A. Maybe SIX BILLION. One to change the lightbulb and then everyone else on the planet can check
if it was done it right !
M$ and others should be scared of this.
You know, if you would just take a gamble and type your messages straight onto the board, instead of typing them in Word and running the spell check, it might be easier to read. Hehehehehe...