Linux Kernel 2.5.19 Released 201
The Pi-Guy writes "It's that time again, yep, another kernel update - 2.5.19 is out there, including lots of drivers ported to the new API, and lots of ia64 and PPC32 fixes! Grab it from your local Kernel.org mirror, or if you're feeling mean, you can grab it directly from kernel.org here. The changelog is also at kernel.org."
Bluetooth in 2.6 will be something to wait for (Score:5, Insightful)
Re:Bluetooth in 2.6 will be something to wait for (Score:2, Informative)
Re:Bluetooth in 2.6 will be something to wait for (Score:2)
state of USB support (Score:2)
Input devices like mice/tablets generally work fine too. The main fiddly thing at this point is USB storage -- my USB Zip drive is usable, but it's a little more fiddly with 2.4.14 than it should be.
Re:Bluetooth in 2.6 will be something to wait for (Score:2)
<clip> in 2.5.14 [kernel.org] Bluetooth support (no longer experimental!) (Maxim Krasnyansky, Bluetooth team) <clip>
You might also want to look at this alternative (Score:2, Informative)
Affix Bluetooth Protocol Stack for Linux [sourceforge.net]
kbuild (Score:2, Interesting)
Ah well.. It will probably end up like kgi.. Great project, but dies because some finnish guy doesn't want to include it into the kernel...
Re:kbuild (Score:5, Informative)
KGI: what sort of example is this?? KGI was nowhere even near complete enough to be added to the kernel. These guys gave up before they were even half-way done. See the ruby project (the great console re-org) for kgi done right. Linus was correct to deny KGI.
A better example would be CML2. It was pretty close to ready. ESR just needed to fix some bugs and make the interface closer to what we've been using for years. Did he? No, he spent all his time adding stuff like autoconfig that nobody would use anyway and flaming up a storm. He completely ignored the grievous xconfig bugs and performance issues. The design of his CML2 language was weird and needlessly complex. Using Python was motivated more by language advocacy rather than design decision. Linus was correct to deny him.
Look at how long it took RL to get the preemtible kernel patch in, or Jaroslav to get ALSA in. Persistence pays off, stability reigns.
If you want kbuild 2.5 now, it's easy to apply the patch...
Re:kbuild (Score:2)
Does this mean we don't get the kernel configuration adventure game?
Re:kbuild (Score:1)
Welcome to the KGI Project
[...]
News 2001-07-03 new snaphsot released
Hmm.. doesn't look like somebody's working on it anymore...
Shouldn't this be prefaced by "Developer..."? (Score:1)
Oh, wait, I shouldn't expect consistency...this is Slashdot.
Re:Shouldn't this be prefaced by "Developer..."? (Score:2, Insightful)
If it was the 3.x.x series, it would be a bit more important. But this is the standard 2.x.x series. This isn't news for most of us.
Re:Shouldn't this be prefaced by "Developer..."? (Score:3, Interesting)
Re:Shouldn't this be prefaced by "Developer..."? (Score:1)
Re:Shouldn't this be prefaced by "Developer..."? (Score:3, Interesting)
I read this as "Linux Kernel 2.4.19 released," and was about to download it and install it on my box as a replacement for 2.4.16. Then I went to get a copy, and couldn't find it... Only then did I realized that it was the 2.5 series, not 2.4
Re:Shouldn't this be prefaced by "Developer..."? (Score:1)
Re:Shouldn't this be prefaced by "Developer..."? (Score:1)
You're right, that's not the sound of a stable release. A stable release has more oomph. "DEVELOPERS! DEVELOPERS! DEVELOPERS! DEVELOPERS!"
:)
LOL (Score:5, Funny)
<rml@tech9.net>
[PATCH] Robert Love likes leather and chains
> Hmm. That patch does not compile. "p->cpu" does not exist, it's
> "p->thread_info->cpu". Tssk.
Ouch, I am bad. Sorry.
Make the ChangeLog entry something really defamatory.
Robert Love
Re:LOL (Score:2)
I thought I would fill in the story
I sent Linus a scheduler patch of some sort and - since I am working on 2.4 and 2.5 O(1) scheduler - I guess I pulled it from my 2.4 tree and it would not compile on 2.5.
Linus sent me the "tssk tssk" email you see and I told him to log the fix under a dematory ChangeLog entry (as you can see) - the result is the above...he has a good sense of humor.
Re:LOL (Score:1)
Worryingly low number of "beta" testers? (Score:5, Interesting)
Today, the Linux Counter system statistics [li.org] show less than 1% of users using the "development" kernel.
Is this a worrying aspect of the Linux community's development cycle?
Re:Worryingly low number of "beta" testers? (Score:1)
Is this a worrying aspect of the Linux community's development cycle?
Or is it an indication that the numbers of professional linux users (who absolutely need stable kernels) is increasing?
Just because the ratio of stable:beta kernel users is decreasing, it doesn't necessarily mean that the number of beta-version kernel users is waning.
I'm encouraged by the lower stable:beta kernel ratio.
Re:Worryingly low number of "beta" testers? (Score:3, Insightful)
Re:Worryingly low number of "beta" testers? (Score:1)
Re:Worryingly low number of "beta" testers? (Score:1)
opps! my bad, please mod parent down. (Score:1)
That is unscientific (Score:2)
The other thing is that the people running that machine-update script are self-selecting. For any statistic to be accurate, it is important that the samples are chosen accurately (for example, a random sample or a stratified random sample). The Linux Counter sample selected themselves to participate, and that is of course not accurate.
So I don't think the statistics presented there are accurate at all.
Re:That is unscientific (Score:1)
It calls itself a poll, right? How many polls question every single member of a group? its always a small segment, of which you then extrapolate the general statistics. Thats just...how they go...you know?
Re:That is unscientific (Score:2)
And even if 2,401 machines is enough.. this group is *self-selecting*. This means this group chose to participate because they are interested in it. Again that is not representative of the entire Linux population, because not all in the Linux population are interested in running that machine-update script or this poll.
In a real survey, the researcher would first calculate the sample size that would be big enough to get accurate results. Next, figure out a proper sampling method. There are many ways to obtain your sample.. I'll just describe one here.. let's say, random sampling.
In the random sampling method, the researcher would go randomly to any Linux user, and ask them questions like what kernel they're running and so on. In this case, we reach a proper number of Linux users.. whether or not they're interested in the poll or aware about it. They would continue doing this until they obtain the required sample size. This is in contrast with the Linux Counter poll, where the entire sample consist of only those who are interested in the poll. Thus, it is inaccurate.
Random sampling is just one sampling method to make a fair survey with less errors.. there are other methods too but I won't go into their details unless you really want me to.
That's the major problem with all online polls. That's why Slashdot polls have that line that says that Slashdot polls are:
"wildly inaccurate. Rounding errors, ballot stuffers, dynamic IPs, firewalls. If you're using these numbers to do anything important, you're insane."
And that's also why the QuickVote polls at CNN.com says:
This QuickVote is not scientific and reflects the opinions of only those Internet users who have chosen to participate. The results cannot be assumed to represent the opinions of Internet users in general, nor the public as a whole.
I hope this clarifies my points.
Re:That is unscientific (Score:2, Informative)
Then don't "feel"... Calculate. You'll see that 2,401 is enough for just about any significance level you can think of. Statisticians deal with this stuff all the time. A sample of 2,401 is HUGE for this. Most Gallup polls only use a sample of 1,000 or so to calculate the opinion of the entire adult population of the United States. See: this [gallup.com] You're correct about the Linux poll probably not being a representative sample, but even that isn't known for sure. Care to take the data from this poll, data from a random sample and preform a simple hypothesis test? (Hint: It's expensive, time consuming, and in the end they might be the same.)
SO SIGN UP WITH THEM THEN! (Score:1)
Re:That is unscientific (Score:2)
But - if anything, the geekiness of the counter should select FOR the kind of people who participate in Linux kernel development. If even this clearly biased sample shows such a small percentage of Linux users spending time getting beyond the 2.4 series, does the Linux community have breadth enough to "continue the show"?
Harald, alarmist
Re:Worryingly low number of "beta" testers? (Score:1)
Is this a worrying aspect of the Linux community's development cycle?
In the old days, linux use was FAR less for production servers than it is now. Development boxes, and the occasional oddball's desktop.
Now adays, its used as a desktop in corporate environments, as a production-level OS, and etc. Due to this, the % using the development kernel is going to be FAR less.
Re:Worryingly low number of "beta" testers? (Score:1)
I try to run the bleeding edge and used all but a few of the 2.3 kernels. About half of 2.5.x series doesn't compile. Right now raid stuff doesn't work.
Not really worrying (Score:1)
Assuming the counter is accurate, even 1% of the user base is still quite a lot and probably includes those most capable of actually contributing to the development of the kernel.
Still, at least personally my eagerness to run the latest kernels has been on the decline. I've been running Linux for over 10 years now. There was a time when I used to immediately patch in anything posted on the kernel mailing by Linus, and spent a considerable time testing and writing patches myself.
These days all of my hardware has been supported by Linux for ages. The new kernels just don't seem to bring so many cool new features to my life as they used to, so I tend to go for a bit more sedate update pace now (of course, there is the occasional memory system rewrite, which makes one want to bite the bullet again). I suspect many of the old timers have the same experience. Besides, Linux performs so good I don't even have any reason to upgrade my hardware.
Re:Worryingly low number of "beta" testers? (Score:1)
User Prefs... (Score:1, Offtopic)
Re:User Prefs... (Score:1)
I, and others, like to read about the progress of the kernel, even if we don't download and compile every version. You're making a mountain out of a molehill here; my advice is to chill out, grab some alcohol from the fridge, and move on. It's obvious you need to relax.
Re:User Prefs... (Score:1)
when will we see 2.4.19? (Score:1)
Also, I'm curious.. is there a "safe" way to play with the development kernels, if you don't have the time to debug filesystems or recover from disasters? are the file-system bugs as bad as they sound, or do they only affect non-ext2/ext3 file systems? does anyone have any "tricks" that they'd like to share? are the 2.5 kernels going to support using GCC 3.1 for greater speed?
Re:when will we see 2.4.19? (Score:3, Informative)
Good news for PowerPC users (Score:3, Informative)
PPA, the girl next door.
Re:Good news for PowerPC users (Score:1, Funny)
>
> PPA, the girl next door.
The _girl_ ? Do you really exist?!
If this is true you have shattered my view on the world.
May I have your phone-number, please? I'd like to propose ASAP.
Re:Good news for PowerPC users (Score:3, Funny)
>
> May I have your phone-number, please? I'd like to propose ASAP.
Dude, she told you she lives "next door." I know that many slashdot geeks shy away from personal contact, but you may find walking one house over much more effective than using the phone.
Wow! (Score:1, Redundant)
This is great news for Linux (Score:1, Insightful)
1) Keeps user skills fresh on how to tar zxf; cd; make; make install
2) Frequent full source downloads lets developers, hosts and users all make sure their net connections are optimized
3) Announcement on non-kernel-related sites keeps Linux, OS of the Gods, in the forefront of everyone's minds
4) Announcement on non-kernel-related sites also ensures that stock analysts, mechanical engineers and daycare workers all get the latest source code to make "all bugs shallow"
5) Let's slashbots who wouldn't know gcc from herpes simplex talk about "the kernel" as if they knew thing one about programming or Linux
Not for normal users (Score:1, Insightful)
Old news (Score:1)
If you want a lot of 2.5 features without sacrificing to much then there is allways the jp kernel patches located here [infolinux.de]. It includes xfs, ext3 fixes, low-latency, alsa and a lot of other things.
Re:Old news (Score:2, Informative)
As was foretold lo, these 3 days ago... (Score:1)
by Lethyos (tofuchute@hotmail . c om) on Saturday June 01, @09:13AM (#3622181)
(User #408045 Info | http://slashdot.org/ | Last Journal: Thursday May 09, @03:25PM)
Haven't you editors heard yet? BSD is dying! Get with the program and post another Linux 2.5 patch level increment announcement. Thank you!
It's about time! 3 days? What took so long?
this is a fantastic coincedence (Score:1, Funny)
Did Linus earn his salary :) (Score:3, Interesting)
(29-May-2002)
version: changelog entries
2.5.19 : 7
2.5.18 : 10
2.5.17: 6
2.5.16: 8
2.5.15: 4
2.5.14 : 2
2.5.13: 3
2.5.12: 5
2.5.11: 5
2.5.10: 10
(24-Apr-2002)
Re:Did Linus earn his salary :) (Score:2)
Linus is salaried to work on the Linux kernel. That's what he does at Transmeta (who have their interest in having Linux run on their chips).
Re:Did Linus earn his salary :) (Score:2)
*bonk* wake up *bonk* it is not necessary to defend your friend (I know everyone knows Linus
What's new in the 2.5.x kernels compared to 2.4.x? (Score:2)
OK, here's my stupid question of the week.
What does the 2.5.x development kernel do that the stable 2.4.x kernel does not?
What are the flashy new bells & whistles?
No flames intended, I'm genuinely curious.
Alsa sound drivers! (Score:3, Informative)
This is quite a big move and I'm glad, the old kernel sound drivers were getting horribly out of date and lacked many features, such as hardware mixing support to name one.
I've been compiling the alsa drivers as modules for the entire 2.4.x series, and am *really* glad I won't have to be doing that any longer.
Re:What's new in the 2.5.x kernels compared to 2.4 (Score:2)
Re:What's new in the 2.5.x kernels compared to 2.4 (Score:1)
From the brief glance, it seems like there are a number of rewrites of the experimental hacks in 2.4 ("Bluetooth no longer experimental", framebuffer rewrite, Video 4 Linux redesign, Generic ACPI, etc).
Mostly stuff I wouldn't use in a hurry. After all, I use Linux for stability over Windows, and really cutting-edge stuff is secondary to rock-solidiness.
kernel.org silliness (Score:5, Informative)
I did, and created a simple PHP script that will redirect you to a random mirror site. (My list of hosts that it uses is everyone with something greater than a T1.) I have very little bandwidth, and am not really supposed to run a webserver on my cable modem anyway, but perhaps someone with a "real" server, with PHP support, could put this up. It's probably not the best-written program, but it works. If someone wants to put it up on their high-bandwidth site and post links, I think it would be a great way to offload traffic from the main server, especially for when "bigger" things are released that cause the main site to hover just below 100 Mbps...
<?php
$hosts = array("http://www.internap.lkams.kernel.org/pub/l
//www.netnitco.lkams.kernel.org/pub/linux
l.org/pub/li
idco.lkams.kernel.org/pub/linux/", "http://www.ftp-orst-edu.lkams.kernel.org/pub
/l
on.lkams.kernel.org/pub/lin
ams.kernel.org/pub/lin
", "http://www.secsup-org.lkams.kernel.org/pub/linux
$max = count($hosts);
$site = rand(0, $max);
header("Location: $hosts[$site]");
exit;
?>
Note that this code uses VERY little bandwidth, since all it sends is an HTTP redirect.
Re:kernel.org silliness (Score:2, Informative)
http://pwhite.myip.org/linux-mirrors.php [myip.org]
Why no one is using the development kernels (Score:2, Interesting)
VIA Chipsets (Score:2)
Re:VIA Chipsets (Score:1)
I'm using a VIA K7T266 motherboard currently. I have no freezes whatsoever. Using 2.4.18 w/ preemptable patch on gentoo.
If you still have crashes, select yer bootimage and append mem=nopentium and you will no longer get crashes.
Re:VIA Chipsets (Score:2)
Sorry, but that doesn't solve all problems. I'm having problems with DRI and OpenGL acceleration on my Matrox G450. It's not the Athlon/AGP bug, but it is quite probably a VIA chipset timing issue.
I'm a little hesitant to file this on the LKML, as I have already filed it in the DRI bug reporting system [sourceforge.net]. I may post it to the LKML eventually, as I have heard the DRI guys are rather busy, and if it really is a VIA chipset issue, it is not a DRI bug, but a general kernel bug, right?
MartRe:VIA Chipsets (Score:2)
Re:VIA Chipsets (Score:2)
at any rate i've found that when all else fails there's ususally a workable alterantive. i've used cdex, and it works just nicely.
Re:VIA Chipsets (Score:2)
This is what annoys me so much, that when I report a problem with a widely used chipset, that the response I get is "Buy a better chipset", or no response at all. The VIA/Lost Interrupt problems have been with the 2.4.x series since 2.4.9, searching the kernel mailing list brings up plenty of people having the same lost interrupt problems, and 99% of them getting ignored, or ridiculed for having a VIA chipset.
Re:VIA Chipsets (Score:1)
Wow, you're a kernel developer, right? Well, its more helpful than the response I got from the kernel mailing list...
well i've tried it.. (Score:3, Interesting)
That's a big change from 2.3's, they would almost always build for me.
While this is the topic... does anyone know if 2.4.19 will ever come out of RC stage?
my goodness! (Score:2)
uh.. not quite. As I write this I'm having to disable scsi support because "Error: please fix me". I'll do w/o my burner to test it anyway.
Thanks magical /.
Now I'm off to post some complaints about other stuff so it should all be fixed in a minute. No need to thank me.
Somewhere near the transmeta headqts... (Score:2)
Oh no! 2.5.19 is full of trivial bugs for which I have patches, they will all fail in the compile stage and give up *frown*.
Quickly Linus, make a new release!
Rusty Russell (Score:1)
-------
[PATCH] DIE "Russel", DIE!
My name is *not* GPL: you may not derive works without approval.
Rusty.
PS. I've also applied for a patent...
--------
I'm guessing he's referring to 'Rusty Russell'? Just thought I'd post it for anyone who missed it, seems a strange thing to put in the changelog.
2.4.19?! (Score:2, Interesting)
I am getting increasingly uneasy about using a kernel in which development on it appears to have STOPPED! I mean, the whole world uses this kernel branch pretty-much...
linux is unstable (Score:1, Interesting)
2.5.20 is out... (Score:4, Informative)
2.5.20 (Score:2, Insightful)
Re:Tell me this (Score:1)
Uhhhh... it's a development kernel. It's not supposed to be stable. Don't go complaining that it destroyed your file system; you were the one who opted to install it.
Besides, if you don't like it, you could dig around in the source to fix it... or, you could apply just a little bit of effort and file a bug report.
Re:Linux Kernel? (Score:1)
Kernel updates matter to nerds
News for NERDS. Stuff that matters. (Score:1)
Re:Don't Complain! (Score:3, Interesting)
Yes - but the same people can easily visit www.kernel.org [kernel.org] every day.
If they did visit kernel.org regularly, they'd have been running the 2.5.19 kernel two days ago.
Re:Don't Complain! (Score:2)
1. linux kernels [kernel.org]
2. debian news [debian.org]
3. kde news [kde.org]
4. mozilla news [mozilla.org]
on slashdot, if you disagree with any of the above 4 topics, the bot-moderators automatically flamebait your ass into oblivion...
Re:Don't Complain! (Score:1)
Dev kernel posts are reasonable (Score:5, Interesting)
There is always going to be something on Slashdot that you aren't interested in. Kernel releases, Star Wars, anime, whatever. There are too many different tastes to please everyone. But you have preferences that allow filtering, and article summaries and headlines to help you decide whether to read an article. I'd rather see more material on Slashdot than less, and decide what's interesting myself.
Finally, redundancy complaints aren't really reasonable. Yes, you can get anime news on an anime site, world events on BBC, linux kernel releases on kernel.org, etc. But because the Net is so large and provides so much information, there's redundant sources for amost all types of information. The point of Slashdot is to provide a nice selection of interesting information to browse at an idle point in your day. Including more information and then letting people filter seems to achive that goal well.
Now, maybe more fine-grained filters w
Re:I know why they posted this article! (Score:1)
Re:I know why they posted this article! (Score:1)
Re:Oh god! Not another kernel update article! (Score:1)
Bitching about it here really isn't going to get you anywhere.
Re:When will Mandrake upgrade their kernel? (Score:3, Insightful)
Re:When will Mandrake upgrade their kernel? (Score:3, Interesting)
On the other hand, if you are not able to come up with this suggestion on your own, it's probably safe to assume that you should stay away from installing it until it moves to something known to work without any problems (after all, unstable code always moves to stable at one time or another.)
Oh boy, I still didn't make time to watch AOTC.
PPA, the girl next door.
"Try it out. Kick the tires." (Score:2)
Perfect opportunity for Linux users to consider switching back to Windows, is more like it.
- A.P.
Re:"Try it out. Kick the tires." (Score:2)
PS> I have a spare comp sitting around that I do a daily data-sync with. Cheap, fast, effortless backup!
Re:"Try it out. Kick the tires." (Score:2)
For example, if you install it on an IDE system that doesn't come up with UDMA ...
Re:"Try it out. Kick the tires." (Score:2)
I love watching the people with sub 1000 user ids. They sound more and more jaded as the years go by.
Wakko's always been like this (Score:2)
Re:Who cares... (Score:1)
Personally, the 2.5.x kernels are of academic interest onlz. 2.4.y remains the bread and butter of what I run and I guess most other people.
Re:Who cares... (Score:1)
Re:WTFC (Score:1)
Re:BFD (Score:2)
Re:BFD (Score:2)
Re:BFD (Score:1)
So instead of useless complaining, might I suggest to move on with your life and deal with it, instead of filling up this article's comment with worthless complaints.
I agree with you though, there should be a "linux software" category at the very least.
Re:BFD (Score:1)