Slashdot Mirror

Back in early November 2001, I started following a discussion between two factions of the Linux kernel community ... There were two main factions, the preemption patch faction and the low-latency patch faction. Both groups were very passionate (i.e. vocal) about the superiority of their solution.

Back in early November 2001, I started following a discussion between two factions of the Linux kernel community ... There were two main factions, the preemption patch faction and the low-latency patch faction. Both groups were very passionate (i.e. vocal) about the superiority of their solution.

Back in early November 2001, I started following a discussion between two factions of the Linux kernel community ... There were two main factions, the preemption patch faction and the low-latency patch faction. Both groups were very passionate (i.e. vocal) about the superiority of their solution.

Linus doesn't believe that designing software is that important.

"I _am_ claiming that the people who think you "design" software are seriously simplifying the issue, and don't actually realize how they themselves work. " - Linus

You should read a letter than Linus wrote a few months ago.

That's hardly the only place where Mr. Raymond has been making an ass of himself these days. He apparently spends more time stirring up flamewars on lkml than he does actually coding his magical kernel configuration machine. He has a tendency to make it sound like his pet project is the most important thing to happen to linux since TCP/IP; the huge threads he spawns would be amusing if he wasn't wasting the time and inciting the anger of the people who actually work on the kernel itself.

That's hardly the only place where Mr. Raymond has been making an ass of himself these days. He apparently spends more time stirring up flamewars on lkml than he does actually coding his magical kernel configuration machine. He has a tendency to make it sound like his pet project is the most important thing to happen to linux since TCP/IP; the huge threads he spawns would be amusing if he wasn't wasting the time and inciting the anger of the people who actually work on the kernel itself.

Yes, but it's all rather confusing. Read this thread in the Linux kernel mailing list if you're really interested. (WARNING: You won't understand any of it unless you know how the x86 virtual memory mechanism works.)

That's OK, Linux has Hot Swap CPU support.

:)

The solution to the problem is really quite simple. As Larry McVoy, who maintains the powerful but non-free BitKeeper RCS system and knows a thing or two about patches, has hinted towards kernel.org may be better off not providing a tarball for each release, instead providing some kind of utility that downloads the latest available full kernel, but only if necessary, plus patches. I'd be all for it. In the meantime, there are a number of incremental patching systems for the Linux kernel that automatically download patches, verify their signatures and patch the kernel which may be worth looking into to save time, bandwidth and resources:

• dlkern
• buildkernel
• lkpatch, which has fallen into disrepair
  

Of course, it goes without saying that everyone should still use their local mirror, particularly as kernel.org will only be accessible to mirrors for the forseeable future.

The solution to the problem is really quite simple. As Larry McVoy, who maintains the powerful but non-free BitKeeper RCS system and knows a thing or two about patches, has hinted towards kernel.org may be better off not providing a tarball for each release, instead providing some kind of utility that downloads the latest available full kernel, but only if necessary, plus patches. I'd be all for it. In the meantime, there are a number of incremental patching systems for the Linux kernel that automatically download patches, verify their signatures and patch the kernel which may be worth looking into to save time, bandwidth and resources:

• dlkern
• buildkernel
• lkpatch, which has fallen into disrepair
  

Of course, it goes without saying that everyone should still use their local mirror, particularly as kernel.org will only be accessible to mirrors for the forseeable future.

well, they do have massive filesystem corruption bugs.

no, wait- the stable series has that as well.

An autoconfigurator would come in handy for CPU-optimization. For example Pentium4 runs significantly faster on some benchmarks when specially compiled software is used.

I don't know of course how much the gain would be, but I think if it's more than 5% it's worth it.

Hehe!
I liked O Xymoron's enthusiastic response:


On Fri, 4 Jan 2002, Ingo Molnar wrote:

> this is more than 6 million context switches per second!

Everyone knows scheduling is boring.

:)

Second link: hats off to those who don't run wu-ftpd. I would never claim that all linux systems are alike. As for those who do (run wu-ftpd), arguing over whether a given windows or linux combo is worse is almost pointless.. they're both buggy POSs and it takes one crack to ruin the box.

As for the third link..

A design flaw, rather than a true "bug"

I do program, and where I come from, design flaws usually count as bugs. Usually they're the hardest ones to fix

There is absolutely NO evidence that this vulnerability has ever been exploited

You could apply that statement to MS's latest problem, and you'd be equally foolish

HOW LONG was it, after the design flaw became known, that the flaw was fixed and new releases made to fix it. A day or two?

This is just from the searching I have done, and it's so ridiculous I actually don't believe it myself; somebody *please* correct the errors here, but AFAICT: Flaw published1/4/01, apparently fixed by redhat 4/10/01 (debian nailed this on 4/16). Somebody noticed on the kernel mailing list 7/24/01 that there was still a problem (improved exploit perhaps?), and this was fixed by redhat on 10/09/01. I can't find a second debian fix; maybe they got it right the first time. Anyway to answer your question, not quite.

Rather it is for the direction sending (perhaps receiving, but less likely) the traffic? The I2 maps shows arrows with two different lines so you can clearly see it... that's my guess with the color split halfway.

Plus, the Internet2 backbone is moving to OC192 in the near term. Saturate that...

Code is findable at namespaces-S2.gz ; it looks like there are a number of newer versions, as we get gradually closer to 2.5.x properly forking off...

Well, I think there's a few more inconsistencies we might not have realized in this story:

1. Slashdot was founded in 1997. The Taliban seized control of Afghanistan in 1996. I know that Slashdot is cool, but do you think that it spread by word of mouth to Afghanistan?
2. Linux was ported to Commodore in 1998. Was this guy infatuated in Linux without ever installing or seeing it?
3. The state of Texas initated the Microsoft antitrust debacle in 1997. Even if that (and details of it) spread by word of mouth, do you think this guy would have had time to form a rational opinion? ("I thought they were going to get Microsoft.")

As a side note, I feel sorry for anyone using any of the search engines that they remember from 1996.

He once virtually ran "like hell" from Tove...

(...)
Trust me. The people who came up with MAP_COPY were stupid. Really. It's an idiotic concept, and it's not worth implementing.

And this all for what is a administration bug in the first place.

In short: just say NO TO DRUGS, and maybe you won't end up like the Hurd people.

Linus

The key word there is "sometimes". Stephen Tweedie recently commented on the Linux Kernel mailing list that resierfs is significantly faster on empty filesystems, but slows down as the filesystem approaches 90% usage (which is pretty typical for a production box).

Beware of the evils of HTML formatting !

Only belive in the one and true Linux Kernel Mailing List and cut'n paste from here into a file.

Then this should do it:

cd linux patch -p1 < /path/to/patch-file

Anyways thank you for the attempt to use this miserable forum, here is a link to the kernel list archive and the post with the patch.

http://www.uwsg.iu.edu/hypermail/linux/kernel/0109 .3/0599.html

Hello Linus? Wake up? How about some reacting on feedback?

This change is present in kernel 2.4.13 (and yes I looked at your corrected link).

Jeff

There were some benchmarks posted recently, but I seem to recall that the subject lines weren't particularly "on topic." Makes them harder to find.
In the meantime, here's one post..
Success report..
aah.. here's the one I was thinking of:
VM benchmarks..

There were some benchmarks posted recently, but I seem to recall that the subject lines weren't particularly "on topic." Makes them harder to find.
In the meantime, here's one post..
Success report..
aah.. here's the one I was thinking of:
VM benchmarks..

Hi, I just noticed a post on the linux-kernel mailing list from Andre Hendrick regarding this. It appears that the "Linux ATA Development has a legal signed NDA for the proper development of the complete and correct FastTrak(tm) open sources driver."

He even got a e-mail from FastTrak which included the beta binary driver, back in November 1999.

Odd. (The relavent e-mail should be here by tomorrow).

That link hasn't worked for a while now. Try
http://uwsg.iu.edu/hypermail/linux/kernel/.

http://uwsg.iu.edu/hypermail/linux/kernel/0012.3/0 538.html

> I've read everything that I can find regarding support of the Highpoint
> controllers RAID functionality under Linux, and I understand what the issues
> have been. The one promising bit of information that I dug up in this process is
> that the 'pseudo' RAID functionality of the Highpoint and Promise IDE RAID
> controllers is now supported in FreeBSD (4.2-RELEASE and 5.0-CURRENT). My
> question is, can the new BSD code be leveraged to add support for these
> controllers to the Linux kernel, and could we reasonably expect to see such
> support in the near future?
>
> (I think that most all of the relevant/important bits are in ata-raid.c and/or
> ata-raid.h. In
> any event, the IDE/ATA guy over on the FreeBSD side is Soren Schmidt
> (sos@freebsd.org), and he
> wrote all of the stuff for this.

Not only do I doubt he would say that

http://uwsg.iu.edu/hypermail/linux/kernel/9906.2/0 906.html

It may or may not still blow (I haven't tried it), but there have been substantial changes. Andrea Arcangeli did a large rewrite in pre11 ("major VM merge.")

If you're looking for a stable VM, I've heard a handful of good things (like this) about 2.4.9-ac10. Alan's been much more cautious than Linus about merging VM changes.

i believe it's a matter of "cat 0 > /proc/sys/cpu/1/active" (Or something similar)

You would probably want to use echo, not cat. :)

More info about it can be found here:

link to mail archive.

Drepper certainly isn't known to mince words. Here's one example from l-k. Judge for yourself, who is more/less rational...

However if this makes it in linux 2.6(?) i will be missing the slashdot "linux now runs on Risc-Q" see the Boot log here on the /. home page.

http://www.uwsg.iu.edu/hypermail/linux/kernel/0105 .1/0334.html

Speech recognition, transcipting, and subtitling of audio-video content helps all of us, particularly the deaf, blind and sportsbar drinkers. Unfortunately speech recognition is not perfect. Good speech recognition could save the CIA a pile on FBIS. Searching text transcripts of a/v files, is only the start.

Internet2, a gigabit network for education and research (see PDF map), has a major future use as an audio-video storage library and distribution network. Video-napster? CMU's Internet2 Informedia Library project researchers are designing visual-video search software for faces, on-screen text, images and shapes. Computers finding on-screen people, text and similar programming... scary.

Check out this presentation with screen shots about Internet2, and its cool tools, uses and experiments. Slide 36 shows Facial Recognition and Optical Character Recognition (OCR) at work. It works so well, it finds text (bottom right) on the U.S. Capital's dome columns... whoops. Slide 37 "Similar Shapes/Content" shows examples of similar content of female news anchors, and soccer / football.

remove the nofreakinspam. to e-mail me.

Speech recognition, transcipting, and subtitling of audio-video content helps all of us, particularly the deaf, blind and sportsbar drinkers. Unfortunately speech recognition is not perfect. Good speech recognition could save the CIA a pile on FBIS. Searching text transcripts of a/v files, is only the start.

Internet2, a gigabit network for education and research (see PDF map), has a major future use as an audio-video storage library and distribution network. Video-napster? CMU's Internet2 Informedia Library project researchers are designing visual-video search software for faces, on-screen text, images and shapes. Computers finding on-screen people, text and similar programming... scary.

Check out this presentation with screen shots about Internet2, and its cool tools, uses and experiments. Slide 36 shows Facial Recognition and Optical Character Recognition (OCR) at work. It works so well, it finds text (bottom right) on the U.S. Capital's dome columns... whoops. Slide 37 "Similar Shapes/Content" shows examples of similar content of female news anchors, and soccer / football.

remove the nofreakinspam. to e-mail me.

There's a known problem with 2.4.x and sun4c sparclinux users. The problem was first mentioned here. Basically the kernel cannot correctly map esp's dvma. I got a one line solution from uzi @ #sparc @ irc.openprojects.net .
Here it is:
edit arch/sparc/mm/sun4c.c
go down to sun4c_mapioaddr()
after the "unsigned long page_entry;", add:
if (!(physaddr & 0x0f000000)) physaddr |= 0x08000000;
That should get you booting.

Your criteria is "small, fast, and compatible" not to mention "freely hacked"

Yet BSD is a BETTER choice based on your criteria.
Small - PicoBSD is a fully functioning Unix OS on a floppy.
Fast - BSD on benchmarks runs FASTER than linux. 10-30% faster.
Compatible - The result of the X86Open group (creation of a common binary format for X86 unixes) was a Linux ELF format. Thusly products like FreeBSD can run Linux binaries. And, FreeBSD can run Quake III linux binaries FASTER than Linux does. And, it is well known and accepted that NetBSD is ported to more platforms.

And, "freely hackable", the BSD license allows the user of the code to commericalize their product without the worry of having to release the IP the company uses in the product. The code is MUCH more free to do what a human wants to than other licenses.

Looking at 2 'linux' projects that are willing to admit they used the pioneering work of NetBSD are the dreamcast port (acknolodge the boot code is NetBSD inspired) and Linux on WinCE

Linux may 'work well', but in the world of computer code, the BSD tree is BETTER!

Given that you're posting around here, I'm guessing you have a Linux box handy. Here are some of my favorite sysadmin tools:

• dig - This is a more advanced tool for seeing what's going on with DNS.
• nmap - A great tool for probing your server to make sure you haven't left anything open.
• Apache Bench (ab) - This simple but effective benchmarking tool comes with the Apache server. It's great to see how your site will perform under load.
• wget - a tool for remotely getting web pages; it's very versatile -- you can even use to save a copy of your whole site, just in case.
• Ethereal - Having trouble figuring out what's going on between the browser and your server? This will capture all the packets and decode them into a nice conversation for you.
• vmstat - want to know why your server is slow? Get used to watching the vmstat numbers while it's fast, so you can see what's different when it's slow. It's raw numbers that are hard to interpret, but it's worth getting to know. Maybe this should be another Ask Slashdot question?
• Netsaint - this is my favorite automatic monitoring package. Once your site is in production, you can set this up to patrol things and make sure everything is working. That lets you get on with other stuff, knowing you'll hear about trouble pronto.
• MRTG - A tool that makes excellent long-term graphs of bandwidth use.
• IPtraf - Where MRTG gives you the broad overview, this gives you the second-by-second nitty gritty.
• perl - Last but most is Perl, a Swiss Army chainsaw of languages. If you'll be doing any web stuff, pick up a copy of Learning Perl and spend a little time with it. Once you learn the magic of regular expressions, you will never again say "that's impossible!" to a problem.

As far as non-sysadmin stuff goes, here are some of my other favorites:

• Bugzilla - this is a free and flexible bug tracking system. Highly recommended, especially for those people who don't think they need a bug tracking system. Our designers thought it was silly to start, but even they use it all the time now.
• CVS - Like bug tracking, most web sites don't think they need version control. Most web sites are wrong! CVSweb is also recommended.
• HTML Tidy - bad HTML in, good HTML out.
• WebTV Simulator - Sure, you and I don't use WebTVs, but a lot of people do. Browse your site with this to see how the other half surfs.
• VMWare - Along similar lines, VMWare is a Windows box emulator. I use it to keep a bunch of synthetic windows machines with a variety of OS versions and browser versions. It makes QA much easier.

And if there are particular tasks that have you stumped, come back and ask again. 'Round these parts, we have big toolboxes.

• It's so ridiculous that it should make people think
• It's owned by a non-U.S. corporation. This should make the traditional patent-mongering U.S. corporations think
• It shows exactly why the whole idea of software patents hurts society instead of helping it

And remember what he said about Y2K?

I admire and like Eric--he's an uber-hacker--but I think in his zeal to sell "Open Source", he's become too confident in his theories.

His main problem (besides his tendancy to overreact to things) is that he's coming from developing for NetWare and Win2k to Linux and he doesn't yet understand the Unix philosophy.

He sometimes posts comments to linux-kernel which are completely foreign in idea (let's put fsck in the kernel, etc.) and than promptly gets attacked by a bunch of random tactless people.

Whatever his background is, he appears to be very competent, if a bit misguided and misunderstood.

And his comment about his brain being more electrically active then everybody else didn't help either...

His main problem (besides his tendancy to overreact to things) is that he's coming from developing for NetWare and Win2k to Linux and he doesn't yet understand the Unix philosophy.

He sometimes posts comments to linux-kernel which are completely foreign in idea (let's put fsck in the kernel, etc.) and than promptly gets attacked by a bunch of random tactless people.

Whatever his background is, he appears to be very competent, if a bit misguided and misunderstood.

And his comment about his brain being more electrically active then everybody else didn't help either...

Check this and this

What would happen if someone said "Let's add 2 new data pins to RS232"?

They already did:

pin 14 STD Secondary transmit data
pin 16 SRD Secondary receive data
(also pin 19 SRTS Secondary RTS, pin 13 SCTS Secondary CTS, etc.)

These pins can be used to double the amount of data sent through your RS-232 cable, which would be useful if you decided to (say) switch from 8-bit characters to 16-bit characters.

It's not an RS-232 cable unless it has all 20 wires!!! (-: (-:

• You can either "mount" the resultant FS publicly, so that anyone on your system can see it, or
• Each and every program that wants to go after encrypted data needs to individually explicitly include cryptographic support to read keys and access/update encrypted data.

As is nicely outlined here , the Plan 9 operating system provides the ability to do this Another Way, using the concept of namespaces. The basic idea is that the "file tree" is no longer a global thing, but is, instead, localized to processes.

In effect, I might, from my shell, run the command:

mount -t crypto /home/cbbrowne/encrypted-stuff /mounts/plaintext

This results in the encrypted stuff in my home area getting attached to the file tree under /mounts/plaintext

The two clever things about this are that:

• /mounts/plaintext is visible to the shell that invoked the mount, as well as to its children.
• /mounts/plaintext is not visible to any other processes.

  Entertainingly, perhaps not even to root.

Why should this be considered relevant to Linux? Because there has been considerable discussion of namespaces in relation to the Linux kernel. It won't be in 2.4, but it's the sort of thing Alex Viro is liable to consider experimenting with in 2.5 or some such version.

Looks like the author of the article is confused about the difference between networks and applications. Research and Education networks, such as Internet2, are there to facilitate existence of advanced applications, such as various data grids, teleimmersion, LBE bulk data transfers, etc.

Appearence of new applications reinforces the need for advanced networks, not the other way around. In fact, we (Internet2) work with the U.S. counterparts of the described European project

Perhaps a lot of students don't realize this, but all traffic between Internet2 participating Universities goes over Abilene (Internet2 backbone).

More information about Internet2 and its activities can be found at:

• Internet2 main page
• QBone (web site runs Slash)
• Abilene NOC site

Many many motherboards allow you to install more RAM than their L2 cache can handle. My Asus P5A can only handle 128MB, for example. Add more RAM and it's acessible, but not cached. So things get dramatically slower.

The slow ram can, however, be used as swap. Much faster than a swap partition on a hard disk, though obviously not as good as real cached RAM.

I've got an old VLB 486 motherboard. And I've got a cacheing VLB IDE controller which takes 30-pin SIMMS. I figure I'll keep the OS on a SCSI drive, and put in an old 40MB IDE disk with a 16MB swap partition and 16MB of cache RAM. Should be a pretty fast swap disk. I just need to find the time to set it up. :-)