Slashdot Mirror
Wonderful World Of Linux 2.4 - Final Candidate
EngrBohn writes "Joe Pranevich has posted the latest & greatest Wonderful World of Linux 2.4 at LinuxToday. 'In the beginning, there was Linus and his 386 ... Several years and many thousands of lines of code later, Linux 2.2 was released ... Linus (and company) continued to hack away at the Linux OS and the 2.4 release of the Linux kernel is nearing completion. Submitted for your approval, this document describes some of the new features in Linux 2.4.'" Helps sort out rumors, half-truths and innuendo. I hope Joe's estimate is conservative on getting CDs shipping with point-four; usually it seems like the distro houses are pretty swift to incorporate.
On the bright side, though, those guys in debian, they have the sexiest asses. Whooo! Yawp!
As a fully qualified IT consultant working on a report into the freeware phenomenon exemplified by the Linux operating system, I am heartened to see that Linux is still riding the crest of the wave of its recent popularity. Despite what many people in my industry have thought, it continues to garner market share, and its popularity as a server is slowly growing.
With the new features that this release brings, Linux is coming ever closer to being ready to handle mission-critical enterprise applications. While I personally think it is almost there, there are several things which are missing before it is truly ready to compete with superior platforms such as Solaris and Win2K.
Scalability is the key - Linux needs to increase its ability to adapt to the constanly growing demand with enterprise servers are placed under, and unless its ability to scale is improved then it will never truly suceed in this growing market. Another important aspect which is missing is increased support for such emerging protocols such as XML and SOAP, which from what I have seen so far, are sadly lacking in version 2.4. Once these minor oversights are corrected, then I truly believe Linux will be able to compete successfully against the might of Win2K, and maybe even win.
Anyway, well done to Linus Torvelds and Richard Stallmann and all the other Linux developers. Thank you.
Actually, Solaris does have PnP support.
no, we don't. and your argument is incorrect, as has been explained to you repeatedly in excruciating detail by people who know signifigantly more than you about all things unix-like.
Linux's handling of OOM situations is perfectly fine by both POSIX and ANSI C (since ANSI C only refers to programs, while POSIX defines behavior for processes).
brk/sbrk's behavior is documented as increasing your address space. There is no promise that the memory will be there when you zero fill the page on first page fault.
for the VERY LAST TIME, the system call you want is mlock, its behavior is PRECISELY what you have been whining for. brk/sbrk have certain semantics, which you don't seem to like. that doesn't make them any less correct.
so far everyone has been polite in explaining why you are flat out wrong, but since the first 4 times it was explained seemed to go right over your head, expect flamage if you continue to repost your drivel without even correcting the parts people have poked giant holes in (which is about 95% of it by now).
``````````````:+;.```````````````````````````````
``````````````````,WBR:``````````````````````````
````````````,;:.``.;;:`....``````````````````````
````````````:+,.``````+RBBX,`````````````````````
````````````````.;;:.`iIII;;IItIi.```````````````
```````````..```:ii:``````:WMMMMiiYYYII+.,:;++,:+
``````````;t:```````.:::,`.:::::;MMMMMWiBMMMMRiRM
``````````````.,,.``;iii,.,,,,.`YWRRRXiRMMMMMIRMM
`````````````.+ii,.``....+IIIi,.......IMMMMMXYMMM
```````,YY:```...``...``.:;;:,.ittti,+YYYYYY:IYYY
``````.+ti.```..`.iIIi,..````.:tttt++BMMMMMItIIII
```````````;XWY.`.,,,..iItI+.`.,,,,:RMMMMMWitttII
`````.,,```:ii:```````:ttti,.+iii+,IXXWWWYitttttt
`````;I:```````iVXVI.`````..iYIIY+tMMMMMMIitttttt
`````````.,:.`,VWWXi;IIIIi..++++;;BMMMMMXittttttt
`````````;YI:.`````:RMMMBi++++i+,IXWXXWYiIIIIIttt
``,tt,````..`.:;:.`:iiii;IMMMMMXtBBMMMRiRRRRRWXXY
``...```````.iYYi.`....`;RMMMBR:BMMMMBIRMMMMMMMMM
``````.+i;```````.+YII:`.,,,,,.VMMMBBXIBBBBBBBMMM
+t+```````.iII+.```..`.iYYYYiiBMMMMWttItIIIIIYYXB
````+YI,```,,,..+iit;.`,;;;;:BMMMMBttiiitttIIIYBM
```.IVY:``````.+YYIt..++++;.IWWWWXiiii+iiiiitIBMM
````````;VVVt``......+YYYY;tBMMMMB+Iii+++iiitXMMM
```````,YXWX,+tttti`.++++:;RMMMMMiYIItiiiiiiIMMMM
````````````;BMMMB+:;;;;:,YRRRRRtIYIIIttti+iBMMMW
````````````;tttt;iMMMMMYYBBBBBXIBBBBBWXXYIWMMMBt
`````````````````;BMMMMXtBMMMMM+MMMMMMMMMMMMMMMW+
`````````````````.::,:,:WMMBMMIWMMMMMMMMMMMMMMMMM
```````````````````````,;:,.::,;++iYYXWRBMMMMMMMM
````````````````````````````````````````,:+tXRMMM
``````````````````````````````````````````````,+Y
`````````````````````````````````````````````````
`````````````````````````````````````````````````
`````````````````````````````````````````````````
`````````````````````````````````````````````````
thank you.
Right, the s3 VIRGE uses TCP/IP over PPP over AGP to transfer command buffers. This way if you suspect your video card has died, you can simply ping /dev/agp/s3virge thanks to Richard Gooch's new DevFS.
Quite an exciting development for s3 owners.
I bought an HP C200 not so long ago, and it comes with a serial connector. It's cheap, takes nice enough pictures for me, and works with gphoto. I'm no "photog" by any means, maybe more expensive ones only have USB. But there are some still left out there that work just fine in Linux.
Please, please, please, please don't go looking at any kernel.org mirrors. We need such insightful comments as yours, and no one wants that to stop by actually visiting the mirrors and seeing 2.4 as it exists currently. "Vaporware" refers to something not existing for users to view, commonly just a list of features in press releases. So we can't have you seeing that it has existed for nearly a year.
Thank you, and good day.
It still pales in comparison to an early release of a Linux kernel. Between an NDA for a beta program (I've done a few for Windows games), and the EULA for the product itself, you basically aren't allowed to say anything about it to anyone. This is the current corporate mindset that drives companies that operate in closed source and Windows worlds. They don't want any bad press, or a leak of their sacred intellectual property.
Back in the days before Windows 95, I knew of a fellow who had some dinky web site about it on his ISP account. He didn't trash it, or anything too bad (in my opinion), but still got letters from MS, the ISP, etc to remove it. So, I use language like "sell your soul" to describe this behavior, because it prevents you from doing what comes naturally, talking. It is over the top, even absurd, as you don't literally sell your soul to them. But as Mr. Limbaugh says, it's "pointing out absurdity by being absurd."
The thing about Linux though, is one piece can go to hell, and leave the rest of the kernel running. There have been times when working with a sound driver or something and it craps out. I can't rmmod it because it's marked as used. The end result, sound is gone but everything else is still working. The kernel is still IP Masq'ing, I'm surfin the web, etc. Compare this to kernel integrations in Windows, and any one piece will bring down the entire machine. So, let them integrate simple daemons like this, not only can I choose to not install it, but if it goes to hell, I know it doesn't (necessarily) bring the entire thing down.
Ok, you can try Linux 2.4 legally anytime. You can get yourself on lists to beta test software for Windows and the like, but you must sell your soul to them first (NDA). You can download it from IRC, newsgroups, etc, but at any given moment MS can sick the feds on you and take everything you own as "evidence." The chances of it happening, probably low if you don't give it to others or use it in your business (they like making examples of those sort of cases). But it is the law, and if you do it, you accept the consequences. When my truck could go over the speed limit, I did so accepting that if I were caught, I'd pay the $100 (or whatever) fine. With the risk comes responsibility.
:)
Even if I cared about trying Windows early, to me, it's just not worth the hassle of jail for a few days, years in court, hundred-thousand dollar fines, having to perform community service, etc. They always go overboard with those sort of trials.
Thank you. I hate it when people go and read the first couple of paragraphs of an article before asking their questions here. What, are they too good for those of us who react on impulse? Who died and made them king? Keep up the good work, my friend. Don't let The Man keep force you to read something before responding.
Now go back under your bridge.
Faking sigs. How mature.
Want to learn about race cars? Read my Book
Now that we're getting into the 2.4-pre series, it's time for ever Linux user to do their bit for kernel development:
- Go grab the latest tarball (use a mirror please)
- Configure, compile, and install it.
- Make a report on linux-kernel about what worked and what didn't
Remember: "given enough eyeballs, all (kernel) bugs are shallow" Linux requires the massively parallel bug-finding that only widespread use and testing can provide.
Keep that paper bag off Linus' head! Find and report them bugs! Do your bit!
Want to learn about race cars? Read my Book
Regardless, truly "new" stuff doesn't magically appear in commercial programs. New concepts are generally developed by research teams, either funded by corporations or working for academia or the government. By the time an idea makes its way into a mainstream program, it's usually been around for a while in one community or another, and someone can point to previous work and say, "See! It's not *new*!"
Still, on the newness theme, the article also has this to say about Linux:
The bit about raw disk access implies that Linux's scheme for raw device access that doesn't require a double for every block device node in /dev is a new thing, too.
You also say
I want a free, robust system that allows me to get my work (programming, research, school work (I attend college in the evenings)) done. Linux does all of that for me. Windows? Not free and far more unstable, anyway. Solaris? Not free, and the Intel version sucks performance-wise anyway. (And no, I'm not going to go buy a Sparc.) BSD? Free, but without the application base Linux has (although most of my work is done with Emacs, LaTeX, CMU-Lisp, and gcc).--Phil (And yes, I'm using 'free' in the 'libre' sense.)
355/113 -- Not the famous irrational number PI, but an incredible simulation!
Note that devfs is currently an *optional* feature, though there are inklings that in the future (i.e. 2.5/2.6) it may be optional in the sense that procfs is, i.e. you don't have to have it configured, but you loose functionality if it's not.
Richard Gooch has been working on getting devfs included in the kernel for ages, it is nice to see it actually show up.
-- Have you ever imagined a world with no hypothetical situations?
It does a little explaining, though I'm not sure there is an absolute justification given. Again, as you said, if you don't like it, don't use it.
-- Have you ever imagined a world with no hypothetical situations?
I'm sorry but it seems you looked too far:
/usr/src/linux/REPORTING-BUGS
See, it was on you hard drive all along.
Thats funny. I wrote it seriously at the time, but then again any time one 'explains' Debian freedom so that people outside debian can understand it Debian people get offended. Go figure. Maybe instead of just moderating you could tell me where I am wrong or "flamebaiting"?
Or does flaimbaiting mean something else in Debian terms, maybe 'something out of the regular authorized mantra.'
I do however ammend a little bit to the definition of stable. They don't want to change it anymore, but they will for security or other emergencies. But they are still rather tired of it and would rather not touch it any more.
^~~^~^^~~^~^~^~^^~^^~^~^~~^^^~^^~~^~~~^~~^
ipfw allow ip from any to any via lo0
allows all traffic via the loopback device, and
ipfw allow tcp from any to any 80 in via fxp0
ipfw allow tcp from any 80 to any out via fxp0
ipfw deny ip from any to any via fxp0
allows all external traffic (via my ethernet card, named fxp0) to my webserver and nothing else. I haven't messed with natd on FreeBSD much, but the huge advantage it has over Linux's ipchains is that it handles NAT in both directions, so hitting your firewall with a connection to port 80 can be redirected to your internal webserver. On Linux 2.2, you have to use ipchains for masquerading (which is a separate kernel option) and ipportfw (which is yet another kernel option) for port-forwarding, aka reverse NAT.
Right now, FreeBSD's got a few advantages over Linux 2.2 in firewalling:
- state monitoring for TCP (can allow all established TCP connections on any ports, but not allow incoming connections, etc.)
- arbitrary, and clearly labeled numbering of rules (counting down the list just to figure out the insertion number for a new rule is inexcusably stupid)
- an intelligent natd that handles masquerading in both incoming and outgoing directions
The good news is that all these "disadvantages" are fixed with the new Linux 2.4 iptables setup, which is a vastly more intelligent, though slightly more complex, way of doing things. There's even a iptables module to handle FTP (which is absolute hell for firewall designers) intelligently. Good stuff; one of 2.4 features I'm really looking forward to.All the mirrors seem to only have up to pre3.
Got a URL for pre4?
Many thanks!
Fialar
I think the fsck example was just wrong. What kind of distro are you running that won't fsck an uncleanly unmounted drive on boot-up?
What is PPPoE ? (I know, I've been living under a rock lately.)
Do not read this
He meant to say ECP and EPP, not UDMA. UDMA is for IDE hard drives, not parallel ports. EPP (Enhanced Parallel Port) is an io port on the parallel port controller which writes into a FIFO, and the pport controller does the handshaking while sending from this buffer. ECP is the same, but with DMA. AFAIK, these were supported by earlier kernels. You can pass io= and dma= on the command line to the parport_pc module, and you can echo 7 > /proc/parport/0/irq to set lp0 using interrupt-driven operation with IRQ 7. I've sent the author an email to let him know about the error, so don't flood his mailbox!
#define X(x,y) x##y
#define X(x,y) x##y
Peter Cordes ; e-mail: X(peter@cordes ,
> rewritten networking - BSD's still better
What is *BSD's net code better for, and how and why? 2.4 isn't out yet, so we don't know exactly what its networking code is like. (Of course, it will be pretty darn close to the code in pre-2.4 releases, so you could have looked at those; Did you? What did you find?)
I admit I haven't looked at either of them, but I'm not going to believe you unless you give me a reason. (threadedness, correctness, efficiency, code clarity/cleanliness, or something. Stuff isn't usually just plain "Better", especially when it is as complicated as networking code.)
#define X(x,y) x##y
#define X(x,y) x##y
Peter Cordes ; e-mail: X(peter@cordes ,
most people have only two arms, and one head. This means it is hard to use more than one computer at the same time. In any case, most people don't have 3 or 4 good computers to switch between to do their different tasks. Divide and conquer is exactly the way to go with user space software, no question about that. Unix command line tools prove that for the things they are good at doing (file/text manipulation, scripting, pretty much anything that can be batch-done :). For operating systems, it doesn't work. Are you going to turn your chair around and say "I'll use my Windows machine now because it has plug & play support for my video decoder card.", then say "ok, now I'll start downloading this video at streaming speed in real time with OpenBSD, because it has kick ass networking code", then realize that you want to be doing more things at once, so you turn around again to your Solaris box. Yeah, works for me like a hole in the head. multi booting on the same machine is even more annoying.
I sure as hell don't want to have to wait for some other OS to boot every time I want to do something different. You basically need a single OS that does all the stuff you want to do, unless you are willing to be annoyed constantly.
#define X(x,y) x##y
#define X(x,y) x##y
Peter Cordes ; e-mail: X(peter@cordes ,
Do your homework, glibc-2.1.3 is already in there. Regarding XFree86, RedHat 6.2 doesn't include it either. And I think that it will take half a year or so before any of the large distributions will include it (Mandrake will of course use it within a month). I can't say anything about apache, but you can use debs from woody if it's so important. Or compile it yourself of course.
Linux 2.4 seems very unstable to me. Remember 2.2.x, it wasn't stable until 2.2.10 or something.
So should they release now or wait for another half year, I think not.
All this stuff (and more) will be included in the next release of Debian (woody), which is currently (quite appropriately) marked unstable. Even RedHat didn't include XFree 4.0 in 6.2 because it's still too unstable. You can't just stick the latest and greatest stuff in the distribution. You have to let it mature a little.
Distributions with kernel 2.4, XFree 4.0, Apache 2.0, etc. will come out by the end of the year. And if you are really desperate, you can just install all this stuff yourself -- nobody is stopping you. But I prefer to wait until everything has been well-tested. Remember, Debian is considered to be the most stable distribution because they freeze it for at least 3 month and spend a lot of time on testing. It comes out rock-solid in the end. And that's why I love Debian. That's the only distro I'd put on a server.
___
___
If you think big enough, you'll never have to do it.
khttp really bothers me. The obvious question is why? Just to get a few more points on benchmarks? (or shall I say "benchcrafts"). And risk the reliability for the sake of a little bit of speed? I *know* it can be un-compiled, but that's not the point. The point is that Linux kernel developers have fallen for the Mindcraft's "benchcrafts" and decided to speed up http by putting the web server in the kernel (a la Microsoft). Everyone knows that the major reason IIS is so unreliable is because it runs in the kernel mode. Haven't we been bashing it all along just for that? And believe me, in the "Enterprise", reliability is much *MUCH* more important than speed.
So, what do we want to put in the kernel today?
___
___
If you think big enough, you'll never have to do it.
Linux ipchains model does not have separate chains for input and output. (I got a bit confused too when I first looked at it). Every packet arrives at the input chain, then passes through the forward chain (if the box forwards/masquerades packets), then leaves from the output chain. Each chain applies its filters to the packets. If a packet in accepted in the input chain, it can still be denied downstream (in the forward or output chain). However, if a packet is denied in the input chain, it does not continue to forward, etc.
Why would you need 3 chains? Well, it's actually *very* convenient. All packets arrive intact to the input chain. If they are accepted, they go to the forward chain where they get forwarded/masqueraded. Forwarding changes the interface of the packet (e.g. eth0 may be external interface while eth1 internal. An incoming packet from the outside will change its interface in the forward chane from eth0 to eth1). Also, if you are using masquerading, packets get masqueraded/demasqueraded in the forward chain. So, when packets arrive to the output chain, they have already been forwarded and their headers have been rewritten by masq!
I find it extremely convenient. On my home firewall (which is also a samba/nfs server for the lack of another box), I can block the smb and nfs traffic with just 1 rule (for each), inserted in the output chain.
Oh yeah, also ipchains allows you to match packets by interface. For example, if eth0 is your external interface, you know ips 10.0.0.0/8, etc. cannot come from there. ipchains allows you to block these addresses coming from eth0, but allow them if they come from eth1 (where eth1 is the local interface). I'm not sure how FreeBSD handles it.
Also, does FreeBSD support port forwarding?
___
___
If you think big enough, you'll never have to do it.
This is a known issue, with a known solution and people are working on it. It is on Alans 2.4 Jobs List
I think that a mention of the new agp support in the kernel is important to show the advance linux is making towards the gaming market.
Its spelt "L-I-N-U-X", but pronunced as "Free Beer"
looks like he converted a bmp to an xpm.
Ever looked at an xpm file with a text editor?
support gun control: take guns from cops
Yeah. What if I want all those things?
That is ... LINUX.
support gun control: take guns from cops
then why does Solaris scale to 128-way SMP, while Linux craps out above 8?
Solaris doesn't scale that high on PC hardware.
Solaris does on SPARC since Sun build their own hardware -- a HUGE advantage.
it's just another thing the kernel developers have been wasting time on instead of catching up where they need to catch up. This project needs some direction.
Linux isn't a commercial development house. If someone wants to add a feature that is useful to a certain subset of Linux users, why not, so long as it's stable and well designed?
Maybe has more features, but last time I checked, BSD was FAR faster and scaled better.
Now your true troll colours come out.
BSD is better at certain things than Linux, and worse at others.
Scales better? BSD barely does SMP at all.
Hate to burst your bubble, but if 2.4 has anything to do with fsck'ing the drives after an unclean shutdown then it would be *seriously* overstepping the bounds of what it's supposed to manage.
Sanity checks on the filesystems is the responsibility of the init scripts, not the kernel.
Compiling debian source pages is stooopid easy. Thus you can run any 'unstable' package on your 'stable' Debian box as long as you can compile from source (so the resulting executable is linked against the proper versions of libaries you have on your stable box.)
Most of this software, especially Xfree86 4.0 and Linux 2.4, will be made availible somewhere somehow in pre-packaged form for stable (potato) Debian systems. Or you can just compile it yourself.
Feel free to contact me directly with any additional questions, or just ask the kind folks in #debian on irc.openprojects.net.
cheers!
This entire article details the stuff in the new kernel.
-- Give him Head? Be a Beacon?
-- Give him Head? Be a Beacon? :P)
(If you can't figure out how to E-Mail me, Don't.
...what does this kernel have in it? Where can I go to find out? Can some kind soul post a link or a list? Inquiring minds want to know....
Ive been wondering this too, for a while. I can eliminate alot of other authors, why not AC's? Currently, I have it set to I dont see anything less than a Score: 1, so I dont view them, but thats more like a workaround than a fix.
Can't you add those things into a regular Debian box? I ask this because I'm just getting into Debian and I really like it. But I'm hoping that I can add the goodies when I want to, which will probably be before they are available in a regular Debian release.
If tits were wings it'd be flying around.
There's also Storm Linux, which is Debian-based.
--
the fact that bigger *NIX variants already have these features will just make it easier for Linux to penetrate the places it's bigger cousins have traditionally held sway over.
but think about it. Solaris and BSD are both descended from the "real" unix, and have had many, many years and lots of money thrown behind them to get the level of performance, stability, and acceptance they have now. Linux is quite new to the playing field, even more so when it comes to the enterprise market. Linux is starting to penetrate this market, even lacking the features of the upcoming 2.4 kernel. The release of 2.4 with the features it adds is only going to speed this adoption. This will get more eyeballs and more dollars pointed at Linux and opensource software, which is a good thing.
Or would you be happier if Linux failed and Solaris took it's place?
--
Of course, you can just recompile the stuff you want, but that is messy on a Debian box.
--
Microsoft and Intel's "PC99" spec has actually banned the PC GamePort. (However, a MIDI port is still "optional".)
So, expect one of the last untouched remnants of the original IBM PC to go away RSN, if it hasn't already.
--
Business. Numbers. Money. People. Computer World.
Well, comparisons to a complete kludge like Win98 are a little unfair to Linux, so let's look at "business-class" "modern" OS, Windows NT:
More fine-grained locking - Windows NT has only had this since Service Pack 4 - only a year ago or so.
USB support - Only shipped last month for WinNT
P&P - Only shipped last month for WinNT
WinModems - Many are still not compatible with WinNT
khttpd - When did IIS ship?
raw I/O - Still not in WinNT (or if it is, who uses it?)
rewritten networking - SP4 again. Still behind in firewalling and other network infrastructure type things.
--
Business. Numbers. Money. People. Computer World.
Can't really be sure of that one, but I bet that when it is out, it'll be SuSE who has it out first. That said, I think I'll wait for Slackware to upgrade to it - 7.0 has been out for a while, so I figure it's only a matter of time.
Know ye not that ye are Gods???
Not soon enough :-)
However, you can always set threshold to 1.
Take a look at Dell PowerEdge 8450
It is an 8-way xeon machine. Not overly expensive if you take the fact that it's a server into consideration.
--
Leonid S. Knyshov
Network Administrator
Leonid S. Knyshov
Find me on Quora
1) You can go back - just unmount the ext3 fs and mount it with ext2. That's all. Oh, it's *designed* to do this. But yes, you get your fscks back.
2) There will be patches for 2.4, just like there are patches for the current 2.2 kernels (including 2.2.15preX). SCT also promised some RPMs with the RH6.2 kernel plus ext3 patches in it. Probably you want to check the linux-fsdevel mailinglist.
There is a quite good patch for kernel support at
http://www.davin.ottawa.on.ca/pppoe/.
the creator Jamal Hadi Salim is actively working on it (last update March 30)
He has made a proposal on netdev (archive) about it, and Michal Ostrowski who wrote another implementation in kernel space has shared the discussion. Read the long thread in the archive.
Jamal writes somewhere in the readmes they'll plan to merge at pppoed 0.5 and it seems they're actively pushing for getting it into the kernel.
Unfortunately I wasn't able to find some information lately, but the fact that the pppoed is being updated gives me hope.
Suse has incorporated Jamals pppoed in their 6.3 kernel and Suse's Andi Kleen had his hands on that code (modularization).
The final goal seems to be to create a generic pppox (x=ethernet/atm/whatnot) device in kernel space and to incorporate pppoed (the userspace part, doing the discovery) in pppd.
I have to say that pppoed on linux is far superior to every implentation on other os's I have seen (winpoet and friends suck ass). There is one driver for win2000 made by a volunteer which seems very good, but only linux already has the pppoe-server.
And they have a fix for the mtu-problem on the clients when connecting a network to the internet with pppoed.
While you're completely right, I don't see how you can call most drivers user-space.
I've been in th is discussion about Linux vs FreeBSD firewalls on comp.unix.bsd.freebsd.misc - basically I've been using FreeBSD for assorted network torture, and had been looking at how Linux does firewalls, and they're completely different.
:)
The eventual conclusion was that the FreeBSD 'single chain' model was more powerful than the linux model (especially with the tee ports in 4.0), and that Linux has this killer hole in the masquerading stuff. But someone did point out that a lot of this was changing under 2.4.
Any news? Any sign of tee ports (a divert port that then drops its' output back into the firewall chain)?
Dave
I write a blog now, you should be afraid.
Those who complain about Microsoft's delays, and shout about "vaporware", are just a vocal minority, and don't represent the community as a whole. I don't care about Microsoft delaying Windows 2000 (even though I have to make strategic decisions based on Windows releases), and I don't mind 2.4 being later than promised (since I can always use the development kernel and patches, most of which are stable enough for my needs).
So, no, we are NOT all hypocrites, though, undoubtedly, there are a few among us.
"It's overkill, of course. But you can never have too much overkill." - Anonymous Slashdot Coward
>Linux isn't Windows 9x. Does Solaris have it? >BSD? DOS? Windows 3.1?
> No, but DOS and Windows 3.1 are dead, and Solaris is THE premiere solution if you need scalability on the server side. If you want
> something to run at home, use Windows. Neither of those OS's are good at everything, but the combination of the two really leaves
you no reason to run Linux anywhere.
Does Windoze NT have USB support?
Informative, my @55
What information is contained in this post? The fact that the poster couldn't get a development kernel to compile?
Gotta love moderators!
I've been trying to use MS Netmeeting 3 across a masquerading Linux box (kernel 2.2.13). There is a kernel module which works whith Netmeeting 2, but with netmeeting 3 only the outgoing video/sound works, while I don't see/hear anything from the other guy.
I didn't see anything on this in the article so I'm asking it here:
Does anyone know if an updated module is in the works for 2.4 (or already in the dev kernels)?
Alternatively, are there good free beer alternatives to netmeeting under windows (or Linux, since I dual-boot... but the webcam is USB - a Phillips Vesta PCV675)
I strongly believe that trying to be clever is detrimental to your health. -- Linus Torvalds
Ok nothing against SuSE, it's a great (gigantic!)distro. But I found it notoriously difficult to upgrade to the latest version tru FTP. If I were a little paranoid, I'd say that, indeed, they're trying to force you to buy the next version (at least they got me to do so for 6.2->6.3).
Well, maybe I'm clueless, since I've only begun to use SuSE recently (before it was mainly RedHat), so if someone wants to correct me.
I strongly believe that trying to be clever is detrimental to your health. -- Linus Torvalds
I just perused the 2.3.99-pre3 source, and I didn't see any sign of a pppox.c driver (not that I necessarily know where to look, I checked in drivers/net...). Anyway, I had heard that 2.4 would include pppox support, but maybe it will show up in the later releases (hopefully before 2.6...).
I have Bell Atlantic DSL (which has recently switched to PPPoE and dynamic ip's), and I use
pppoed and the pppoe kernel patch by Jamal Hadi Salim (2.2.14 patch and pppoed source available from: http://www.davin.ottawa.on.ca/pppoe). It works great for me. I have heard others have success with the Roaring Penguin pppoe userspace client.
PPPoE is PPP over Ethernet. Seems like many DSL providers are using this instead of DHCP or static IP's. I personally think it sucks, but until there is a fast access alternative here in central jersey I am stuck with it (I don't live in comcast's area, and so I can't get 2-way cable modem). Here's a portion of RFC 2516 on PPPoE:
PPP over Ethernet (PPPoE) provides the ability to connect a network of hosts over a simple bridging access device to a remote Access Concentrator. With this model, each host utilizes it's own PPP stack and the user is presented with a familiar user interface. Access control, billing and type of service can be done on a per-user, rather than a per-site, basis.
I just read some of the netdev and linux-kernel mailing list archive. I think that pppox support is in the 2.3.99-pre3 kernel, in drivers/net/ppp_generic.c. It seems that specific implementations of pppox will be handled by plugins to pppd.
BeOS may be awesome at multimedia creation (never tried it personally, except my TV card doesn't seem to work, a wintv no less!) But the media playback for standard formats is horrible. I own beos 4, got shipped 4.5, got the patches and even have 5 PE installed, but I cannot play many movies. Sure, older avis/movs play great, no sorenson (don't blame them) and avis with mpeg audio don't work. Also, I still don't understand why the MPEG codec they have is so inefficient. I mean, the least efficient stuff under linux( i.e. no hardware overlays, no use of assembly code) Takes about 50-60% to play, the most streamlined (still without hardware support, just a lot of MMX colorspace conversion/IDCT) 20-40%, but under be, Pulse goes through the roof and audio and video skips on a 400 Mhz machine playing a 352x240 MPG-1 systems stream! BeOS is snappy in so many ways, but if it is so much of a great multimedia OS, then why does playback seem to lag behind in support and efficiency?
XML is like violence. If it doesn't solve the problem, use more.
Cycon wrote: "The more corporate-oriented distributions will quite likely wait until 2.4 has had a chance to stabelize a bit longer, especially the distributions targetted at servers - Redhat for instance, and probably Caldera and Corel as well."
/some/ distributions make his estimate conservative" or similar.
True enough -- some distribs. do emphasize stability over speed, and Mandrake is a good example of one that will probably feature New Toy X first. I probably should have said something like "I hope that at least
timothy
jrnl: http://tinyurl.com/c2l8yr / foes: http://tinyurl.com/ckjno5
I2C device stuff wasnt mentioned..its in 2.3.x right now (hardware monitoring etc) and it should be in 2.4 as well.
Maybe we should have "not before [...]" instead of "by [...]", hm?
Laws do not persuade just because they threaten. --Seneca
Maybe someone should write:
Utilizing Linux in a Windows 2000 Environment or 'How do I work around all those intentional problems with the Windows 2000 boxen my boss insists we buy'
Will in Seattle
September 17, 1991.
And if you want to take a look at it, here's a link.
Thanks for the correction, but it's sorta weird that the link is wrong. I just double clicked in Netscapes address bar, to highlight the whole thing, and then immediatly pasted into the reply box. Weird. Maybe slashdot hates me or something : )
>> i love Debian, but they have the worst
... where it'll be properly thought of ... so, what's
... so they
>> timing...
All this stuff will be available in woody,
no doubt
as "unstable" in the Debian sense
the big deal?
Kernel 2.4.0 should be stable unto itself.
glibc, apache, XFree86, PAN, xv, etc may all be
stable unto themselves
Debian is concerned with everything
would be hard-pressed to declare the whole
system stable without giving the configuration
a while to prove itself.
As my father lik@(munch munch)...
While Red Hat may be targeted at servers, it has, at least in the past, not been the most secure. Corel is definitely not targeted at servers. They say clearly that they're for desktop users. They're even less secure than Red Hat, with more root exploits than you can shake a stick at. There's nothing wrong with Debian waiting to be secure and stable in their releases. This ensures that people can know that a Debian release is stable and secure. Anyone can upgrade their kernel if they wish.
Chris Hagar
"The price of freedom is eternal vigilance." - Thomas Jefferson
P.S: I believe vaporware is a software that not only is delayed and hasn't been released, but has not been seen by the public or has been seen so little only in self-running demos.
Chris Hagar
"The price of freedom is eternal vigilance." - Thomas Jefferson
The thing I wasn't expecting reading the article was the change to the /dev/ directory. In order to more uniformly identify connection types, a cleaner way of identifying devices is being established so that ide drives will be identified more like scsi have been (i.e. ide0 ide1 ...).
/dev/ directory will only be populated with devices that were actually detected, which should help in navigating the directory, especially for newer linux users such as myself.
Also, the
Work for Change & GET PAID!
Where can I get 6.4? I didn't see it on their site.
is that how Bell Atlantic allocates dynamic IP's? I have a static IP from Bell Atlantic. Perhaps that's because I got it before they changed over to dynamic ip allocation.
Hates people who have stupid little sigs
Actuallly, I live in Northern New Jersey. I've been using their DSL for 8 months now. They are the most incompetent boobs I've ever run into. I can't wait until Cable rolls out internet service in the next couple of months.
Hates people who have stupid little sigs
Actually, that is a pretty stupid reason to call it GNU/Linux. There is a lot more to Linux than tar and grep. BeOS uses the GNU suite too. Does that make it GNU/BeOS? Hell no. Face it, the GNU stuff is pretty minor these days. Back in the day, maybe tar was worth something. These days tar and cp aren't exactly a major part of the OS, (unlike GNOME, or X, or KDE, or OSS, ad nauseum.)
A deep unwavering belief is a sure sign you're missing something...
Actually there's a lot more to GNU than tar and grep.
The GNU/Linux thing is more predicated on the likes of gcc and glibc than it is on tar and grep (great though they are).
Not that I really care about the GNU/Linux thing, 'Linux' is just a label in a lot of ways, and given most people's are running a 'Something Linux (Redhat, Debian, SuSE or whatever) this is even more the case.
I know that this is getting OT, but I have to agree. I think SuSE is the best distro out there. Its great to have 6 cdroms with every tool known to man, and to have a nice default desktop envirment. But I dont think your parinoid, I have never seen an ftp site slower then SuSE, there free version is very cutdown, and poorly done, and now it looks like there going to time there releases to make people buy multiple copies.
To bad SuSE 6.4 was released last weekend with 2.2.14 . I hope they didnt do that on purpose to try and sell 6.5 in 3 months.
is at http://www.cs.columbia.edu/~ezk/research/software/ cryptfs/
Don't like to correct you, but the correct link is this.
Sorry, couldn't help it. Feel free to moderate this to off topic.
*poing*
Is this post not nifty? Sluggy Freelance. Worshi
Answering your subject, not the actual content of your post... yes. USB works in Linux. I have successfully managed to sync my Visor with the 2.3.99 series. Still a few glitches, but it does work. Swing by http://usbvisor.sourceforge.net to get more info. Greg (the author) is rather helpful in resolving any problems you might encounter.
--------------------------------------------------
I was actually referring to Webcams/video cams
If you are sure it is a real bug, you could post to the linux-kernel mailing list.
I know responding to trolls is stupid...this time
I just have to...
Yes... Linux is not GNU. It is Linux. RMS never
said any differnt either. Linux is also useless
on its own. The "Linux OS" is mostly GNU, with the
Linux kernel.
[sjc@brake sjc]$ cp --version
cp (GNU fileutils) 4.0o
[sjc@brake sjc]$ ls --version
ls (GNU fileutils) 4.0o
[sjc@brake sjc]$ tar --version
tar (GNU tar) 1.13.17
[sjc@brake sjc]$ df --version
df (GNU fileutils) 4.0o
[sjc@brake sjc]$ grep --version
grep (GNU grep) 2.4
Need I go on? Sure lots of tools are not GNU, but
the most common ones, the ones most people use
most of the time are GNU. (unless there are linux
distros that wrote ALL of their own tools from
scratch? is there one? Maybe a BSD toolset?)
"I opened my eyes, and everything went dark again"
Exactly. I am getting pretty angry about nvidia not owning up to their PR. With 2.4, and DRI I should be getting all $250 worth of my GeForce, but no. Now, back on topic, I am glad to see 2.4 progressing. I compiled 2.3.44 a while back, and it ran great. USB support is really needed, especially with some computer manufactures selling computers that have the keyboard and mouse connected through USB. Can't wait until the next 2.4 based distro's come out!!
Hey someone was telling me at some point that the new kernel (to be 2.4) HAd more specific optimizations for more processors, and have found this to be true. HOWEVER I was told that the K6-2 was more or less 10 or soo smaller RISC based chips that had a scheduler and a decoder for x86 compatibality and that the new kernel now sent the processor RISC like commands instead of x86. I thought this was very cool, consiering that this would be a VAST speed improvement, but now with kernel 2.3.99-pre3 (latest I've found) the speed improvement brought on by checking the "k6-2" option in the kernel is negligable, noticable but not like 10 smaller RISC chips working through things. Was wondering if anyone knew anything about this? Thanks. - J
Hiya
Im particularly interested in the 'raw devices' that 2.4 adds. Im currently locked in an arguement with our Oracle DBA, who believes that its the best thing since sliced bread.
Im wondering, since Oracle accesses these raw devices directly (implementing its own minimal 'filesystem' (although my friend here still maintains it is not a filesystem)), would it not be more efficient to write a minimal OracleFS driver for Linux, and have Linux service the Oracle calls? I can see arguments for both sides, where Oracle avoids Kernel overhead by accessing the device directly (although the kernel must allow this to happen, and hopefully would have some for of check in place), and on the other hand having the FS driver running in kernel space would have to be more efficient that Oracle's user space 'Filesystem' driver, but this adds another level of complexity.
Has anyone looked into this? Theoretically, which would be the most efficient approach? I know that this 'Oracle Filesystem' driver doesnt exist, but would it be theoretically possible to implement?
Views/ideas anyone? Please respond.
Simon
The real linux_penguin has Slashdot ID 101961. Anyone else is an impostor. Including Bruce Perens.
Anyone know if 2.4 lifts the 1000-ish sockets per process limmit? I am trying to construct a server to handle many thousands of people simulaneously (spelling) connected... --Ach! Mein liebe! (Random)
rewritten networking - BSD's still better
I won't deny that FreeBSD is faster then the current Linux Kernels at networking. But with 2.4 not even being released yet how can you claim that BSD is still superior? Do you have any bench marks I can look at anywhere? I would do it myself but I only have one computer and repartitioning now would be a real pain in the ass.
Molog
So Linus, what are we doing tonight?
So Linus, what are we going to do tonight?
The same thing we do every night Tux. Try to take over the world!
One possible security issue would be some kind of DoS attack. Maybe by saturating the machine with requests, the kernel would freeze up or something. As with any new thing, you should test it first.
I am not a lawyer.
2.3.99* wont even boot on a via VT86c686a based motherboard. 2.2 works fine, where is the improvement?
BTW, where is the official bug report page/email for the linux kernel? I've looked all over but have found no documented way to report bugs. =\
Webmaster www.streetrodstuff.com
I wonder if there is a single line of (non trivial) code from 0.01 that still exists in 2.4... probably not :-)
Nvidia are being WHORES about thier 3D cards and XFree86. No DRI for you. No 3D acceleration for you. No OpenGL for you.
Feel free to write to them on how you feel about thier decision.
Lars -
Quick question - in what ways are the BSDs, especially NetBSD (which seems to have the best support for advanced IP features like ipv6) - incompatible with the letter of the IPv4 specification?
I'd love to see an URL on this. Thanks!
--N.
The thing that makes Linux a pleasure to use is, by and large, not the kernel. It's the vast userspace that surrounds it that is of interest, especially the packaging.
[Extreme opinion] At the CORE OF THE OPEN SOURCE MOVEMENT is gcc. It is incomparably important, and I would argue significantly more crucial to the future of free software than the ol' piddly linux kernel. Why there is not more attention paid to the dozens of changes and influences on this project is absolutely beyond me.
[/Extreme opinion]
Other crucial things:
The twenty odd shells in use
perl/python/expect BLAH BLAH BLAH... Basically the thousands of languages and scripts that make everything go
Netscape/Mozilla
dozens of filesystems
DRIVERS DRIVERS DRIVERS!!!
The poster children samba/apache and their ilk
glib / readline / texutils / binutils / GNU GNU GNU
The BSD-licensed variants of all the above
Documentation, support, people, IRC, usenet, et al
Themes and skins and shit galore
All the new XP multimedia APIs
XFree86 and all the other implementations of X
DRIVERS DRIVERS DRIVERS
ALLL the toolkits and bindings and ALL THAT SHIT that flings by on freshmeat
TRILLIONS of things for manipulating mail and network monitors and ALL THAT SHIT
Extremely fundamental: make, makealikes, all the packaging and dependancy systems
The THOUSANDS of standards and APIs and reference implementations for EVERYTHING
And most crucially, wanda the fish
XXX
All i'm whining about is, userspace is vastly more important than piddly dumbfuck kernel, as nice and interesting as it may be.
Duh. I made an obvious. Now feed me moderation!
/me picks nose.
-troll taker
Good thing they're not the ones doing kernel development. When you submit a bug, they're not the ones who get it. There's a difference between doing a distrobution and working on kernel devlopment.
no no no..you're all wrong. you've jumped the gun. you should have poured 2.3.99pre3 bowls of hot grits down your pants.
So does this mean that 2.4 is released? I checked kernel.org, and it still has 2.2.14 as the most recent stable kernel.
segfault@bellatlantic.net
Yes, I get a dynamic from Bell.. Do you live in Northern VA? Possibly it's dynamic in some places and static in others.. However, I am sure that bell would give me a static if i wanted to shell out a few bucks :o)
segfault@bellatlantic.net
Yes, if you check out Bero's Post from an earlier thread, it does appear that RedHat wants to wait for 2.4.5 (which makes perfect sense to me too)
segfault@bellatlantic.net
What I'd like to know, is when will it be likely to see a distro with 2.4, x 4, and kde 2 ?
I'm not competent to add those components manually yet, but I'm itching to try em out.
thanks.
I looked for linux because I didnt like what I saw anywhere else.
The following post brought to you by the letters V, I, and M
I know. What I was trying to say is I hope USB now works so it would help bring down the learning curve so the masses can enjoy Linux too. The fsck example was just that...an example of why Linux is simply to hard for them right now.
To answer your question I use Mandrake 7.0.
/dev/hdx is having problems.
Yes it will scan with fsck on it own. It doesn't always fix it though. What it does do is promt you for the root password for maintainence. The average user cannot work from the command line let alone figure out which
Cool. Just the news I've been waiting for. thanks for the link.
Whenever they get the drivers done (soon i hope), I'll be able to do hardware accelerated X and OpenGl on my cheap ass $40 ATI Rage Pro card. Yeah! at least i didn't hand over $100+ for a fancy Nvidia card that will probably never do either... cheap hardware and Linux, what a combination :)
All that we see or seem is but a dream within a dream.
"Aamzingly," this comment is moderated up to the top of the comments (yeah, I know - I logged out to see why this post gets past my +1 threshold, sue me) and is rated as +3, Informative.
I'm sorry, but I don't see how this is anything but a troll. Yeah, Linus and all the others are getting ready to promote a development tree to 'stable' that doesn't even compile, give me a break. Yeah, there are glitches that need to be worked out - that's why there is a development tree.
Moderators picked the wrong day to start/stop sniffin' glue.
There is much cruelty in the universe, John.
Yeah, we seem to have the tour map.
Does anybody knows if pppoe will make it in the 2.4 kernel, or if it will ever be?
---
guillaume
give me all your garmonbozia
Plug and Play - Windows has had it for how many years? Now, how many years has it had a reliable one? The first answer is about 4, the second is still zero. Among gamers, it's referred to as Plug and Pray, because that's how you install things in Windows.
WinModems - you get what you pay for. If you need a winmodem support, set up a w98se machine with internet connection sharing, and tell your linux machine you're on the net. Or go buy an external (serial, not USB) modem that you can use your standard Hayes commands on.
Don't compare Linux to a half-baked system like Windows. Even though Linux has almost no support for modern hardware like PCI sound cards and some things, it will always have its place in my house - as a dedicated Half-Life server, word processor, and a way to show off that I am better than my W9x using friends.
"Assume the worst about people, and you'll generally be correct"
However, since it is still under construction, you can expect minor glitches that you would also see if you were using something that was not finished - i.e. a car without spark plugs, an airplane with 'BETA' wings that somebody gave you for free, or a boat with a hull constructed by somebody who found that a similar hull worked on their boat. You're taking risks with situations like those. However, when you shell out money for an OS, you expect it to be done, not half-baked like Windows. It seems that the raw eggs they used never quite got the salmonella cooked out of them. When you get raw cookie dough, like Linux, you are taking the risk of poisoning knowingly. Nobody gets salmonella poisioning from Chips Ahoy!(tm). Nobody should crash their computer with Windows.
However, I found your post to be correct on several issues. I have, in fact used Linux only for 4 months, and still am trying to get all my hardware to work. But I didn't buy it expecting all my hardware to work, or expecting to do a lot of programming, or because I know anything about the inner workings. But I do want to accomplish those in the future.
"Assume the worst about people, and you'll generally be correct"
I've been monitoring Alan Cox's diary on his site (www.linux.org.uk) and it's been a long time coming. Seems the trees are merging and him and Linus are working on the final touches. Will be great when it's stable.
Even the samurai
have teddy bears,
and even the teddy bears
get drunk
So you have testdriven an IP stack that isn't even released yet? Or, if you have tested the 2.3.99-pre3 and the latest BSD IP stacks, could you post (a link to) your test results?
-><-
Grand Reverence Zan Zu, AB, DD, KSC
At least in terms of BSD, this is not true - in fact, the licsensing debacles of past years means the *BSD had to fight from behind.
The fact that *BSD's excel at their given tasks has little to do with money or time, but simply hard work.
"An engineer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away." IMHO, there is nothing at all more important than keeping new and unique features out of the kernel unless they are absolutely necessary -- even this seems like too much (well, really just khttpd; I agree with you that it seems like an exceptionally silly way to whore Linux to the small web-server market.)
An app revision can be partially judged on the basis of a bullet-list of what it does now; a kernel revision, not at all. Experience tells me to place blind faith in there being good new stuff in there that I'll never know about, because it is good. This may be what prevents M$ from ever producing a good OS -- they're embarrassed to ship a product if all they can say about it is, "It's faster and it crashes less. It's better."
Coming soon: ksendmail! kbash! Anything that's on the console would be better in the kernel, right? Maybe we should put the web browser in there, or at least let it handle a few API calls... ;-)
About BSD/Solaris/Linux: I must accept your statements on faith, as I have never used them.
However, DOS isn't dead (for me, anyway), as DOS is my asm-coding and plaintext-editing environment of choice. And it is still floating around beneath Windows 95/98. Also, DOS/Windows3.x are all over around here, as people pick up an old {2|3|4}86 at a garage sale for their kids to play with. And it's what surfaces in the computer-repair class at JCC: who cares if an XT board goes up in flames because you connected P8 and P9 backwards? You still lose points, but it's cheaper than blowing up your neighbor's board.
Linux, AFAIK, can be pretty small, until it gets bundled with <sarcasm>Microsoft Office for Linux</sarcasm> or the millions of things that people like to put in distros. One I looked at required 600 megs of HDD space, and recommended 1.5 gigs! And they say Linux runs on the "old iron"----I'd like to see how Mozilla fares on an ancient Digital at 20 MHz.
~~~LXT~~~
Life is like a computer program: anything that can't happen, will.
USB support - who doesn't have it?
AmigaOS handles USB!? Wow! I didn't know my good ol' Workbench 1.3 was that futuristic!
khttpd - this is just stupid. Are we asking to get hacked?
Then don't compile it. Then you'll have an HTTP-incapable server! (See also: MS Windows.)
P&P - Windows has had this for HOW many years?
WinModems - ditto
Linux isn't Windows 9x. Does Solaris have it? BSD? DOS? Windows 3.1?
Basically, no matter what it comes down to, there's always a better choice. Want to run consumer software? Windows. Want scalability? Solaris. Want a good server? BSD. Notice Linux anywhere in there?
So to get all of these at once, I have to run 3 operating systems... simultaneously! What fun!
~~~LXT~~~
Life is like a computer program: anything that can't happen, will.
khttpd - this is just stupid. Are we asking to get hacked?
IBM includes a kernel level web server in AIX. NT pulls some kernel level tricks. If you don't like it you can always disable it.
I have been waiting for this, between this and X4.0 we finally have Direct Rendering!!!! Take advantage of my Ge-Force please!!!!!!
Remember the Pentium bug was even less likely to affect you, but people didn't accept that becouse the bug was built in, and that made the thing upredictable under normal circumstances.
I'm posting as an AC because I'm about to get flamed... but...
I remember late in 1999 when Linus and Cox were promising kernel 2.4 before the official release of Windows 2000. That day has come and gone, and we still don't have 2.4. In the past, we as a community flame Microsoft for such promises, calling their products "vaporware", yet we did not do the same with our beloved Torvalds when his dates passed.
Now, is this due to the nature of corporations - whose promises are presumably under their control and payroll vs Open Source projects that seem to be improved in spurts - or are we all hypocrites?
That being said, it should be known that I use Linux solely [1] on an SMP machine and I am dying for the new 2.4 kernel.
[1] Well, I did boot w98SE yesterday - in VMWare.
http://kurt.andover.net/Reiser -filesystem-HOWTO.html
Reiser driver and utils are still in development, and there's been some flamage on the kernel-dev list over it, but hey, works for me!
Kernel developers cannot fix unreported bugs....
Want to learn about race cars? Read my Book
The kernel httpd stuff is very simplified. It's not like they're merging Apache into the kernel source here. Since it is very lean and limited in functionality to providing simple, static content, the potential for instability is minimized. For people using their Linux systems for serving up static web content, this definitely makes things faster and more efficient.
And you don't have to "un-compile" it. You have to explicitely turn it on if you want to use it. If you don't want the feature, leave the damn thing turned off when building your kernel.
Redhat is slower than anybody, in that I can't remember the last time I got in to ftp.redhat.com. I usually use ftp.freesoftware.com (formerly ftp.cdrom.com) although they don't have redhat 6.2 yet. They are usually up to date, available(5000 users) and fast(1Gbps pipe to the internet).
--
"L'IT c'est moi!"
2.3 boots just fine on my laptop, but when I move it to the iopener, it fails to find the partitions on
Running an Iopener will be a major hurdle to the acceptance of 2.4. At least in my house.
Anybody had this same problem, or better yet, a solution?
---
Get back to me when 2.4.20 comes out :)
I guess it all comes down to the long-term goals. Mandrake and RedHat, being for-profit companies are driven by the market they cater to to provide the latest stuff in their distribution. Debian has the goal of providing a free, stable Linux, even if it's not up-to-date.
But look at the "current" stable Debian distribution. It's so horribly outdated, it's beyond the "less flashy" area and getting into "less useful". There are many things that I _need_ to be able to do with my Debian boxes that I have no choice but to run potato or some other distro.
--
XFree86 4.0 has been released
glibc 2.1.3 has been released
apache 1.3.12 has been released
Linux 2.4.0 might get released
i love Debian, but they have the worst timing...
--
How easy is it to migrate from 2.2.x to 2.4 (well, 2.3.99pre, in this case)? Can I just download source, compile and go like I can with current 2.2 kernels? Or are other crucial pieces of software going to have to be upgraded as well?
Jay (=
As is obvious to everybody with one or more neurons firing, the previous poster was not Linus Torvalds, but YADTAST (Yet Another Dummer Than A Stump Troll).
/usr/include.
That having been said, the original poster is probably referring to the 'THIS_MODULE' undeclared here error message a number of us ran into, the solution to which is here.
In short, you need to #include linux/modules.h just before the first #include asm/.. statement. An annoying buglet in the pre-release code, but easy enough to find the answer at deja.com, altavista.com, or by grepping on the contents of
At the very least, a bug report gets (at this point, has gotten) the bug fixed.
The Future of Human Evolution: Autonomy
It would be nice if Linux got some efficient, simple user-space file system support. Until then, I think the best bet is still to put things into the kernel.
I'm not an expert, but from my experience, there will be a few important utilities that you'll want to upgrade to get a stable system. I believe the README has a list of required software; if you check it against what you have you'll see what you would need to upgrade. It shouldn't be very complicated, but you might be better off sticking with 2.2 until your distro makes the switch.
Gates' Law: Every 18 months, the speed of software halves.
Maybe I should just break down and replace my 8-year old drives, but in the meantime software RAID-5 is an economical way to weather the frequent drive failures.
Unfortunately, it does not seem to be included in the pre3 kernel. Anyone know what's going on with this?
How old is Linux?
cpeterso
When will that be implemented :)
Thank you SO much. You just saved me hours of work. I was compiling the code as I read your info. The work that I am doing is using Sys V IPC heavly. Thanks again!
/etc/fstab. Because 4.0 uses SHM. Here is a printout of an ipcs call on my Linux box...
By the way anybody using XFree86 4.0 should include your adition to the
[rreich@orcana rreich]$ ipcs
------ Shared Memory Segments --------
key shmid owner perms bytes nattch status
0x00000000 98305 root 644 4096 5 dest
0x00000000 131074 root 644 4096 2 dest
0x00000000 1376259 root 644 4096 2 dest
0x00000000 1409028 root 644 4096 3 dest
0x00000000 1441797 root 644 4096 2 dest
0x00000000 5210118 root 644 4096 6 dest
0x00000000 5242887 root 644 4096 10 dest
0x00000000 5275656 root 644 4096 6 dest
0x00000000 18710537 root 644 4096 2 dest
0x00000000 18743306 root 644 4096 2 dest
0x00000000 18776075 root 644 4096 2 dest
0x00000000 18808844 root 644 4096 2 dest
0x00000000 18841613 root 644 4096 2 dest
0x00000000 19169294 root 644 4096 2 dest
0x00000000 19202063 root 644 4096 2 dest
0x00000000 19234832 root 644 4096 2 dest
0x00000000 19365905 root 644 4096 2 dest
0x00000000 19300370 root 644 4096 2 dest
0x00000000 19398675 root 644 4096 2 dest
0x00000000 19922964 root 644 4096 2 dest
0x00000000 19955733 root 644 4096 2 dest
0x00000000 20578326 root 644 4096 2 dest
------ Semaphore Arrays --------
key semid owner perms nsems status
------ Message Queues --------
key msqid owner perms used-bytes messages
... all of the SHM entries are for XFree86 4.0. They don't show up if you don't include the stuff in the fstab.
If at first you don't succeed, skydiving is not for you.
Actually, BeOS uses gcc was well. RMS isn't on Be's case.
A deep unwavering belief is a sure sign you're missing something...
Want to work at Transmeta? Hedgefund.net? AT&T?
Can your IM do this?
Re: USB
Try buying a Camera for your PC that ISNT USB nowadays.. not to mention that within a couple of years i can see USB replacing the Gameport Joystick for a joke, along with PS/2.
"this has got to be a late april fool's prank -- if not, add this to the heap of reasons not to execute binary-only code..."
Or don't run as root.
I definately agree with you, though. I really hope that you cannot write your own microcode. Screw up and you can go buy a new processor.
------
Oh in general I agree. I usually refer to the
OS I run as just "linux". However...I don't think
its silly to use GNU/Linux either.
However, I am a definite supporter of what GNU
and the FSF stand for. While I don't see the need
to make everyone call it "GNU Linux", I do think
the GNU proect deserves some credit.
As far as being major parts of the OS...I don't
use GNOME or KDE (blah). I spend most of my days
in an ETerm typeing into vi and pushing files
around. I think all of those things are fairly
major parts of the OS.
In any case...It doesn't really matter what you
call it. I just favor GNU/Linux for the reasons
that it gives credit to the FSF, and it makes
the note that Linux is just a kernel...there is
alot more to the system than that.
Is it silly? I don't think so. Is it silly to
argue about? yes it is. (tho...I kind of enjoy
arguing over stupid things)
"I opened my eyes, and everything went dark again"
This is a note/plea to the users out there who are reluctant to submit bug reports. First, you must all realize that you are not dealing with Microsoft, or any other large corporation for that matter. While it may seem to you like submitting a bug report will be a waste of your time ("they'll never read it / do anything about it"), it is not. Bug reports are acted upon if they possibly can be, and everyone's bug reports are extremely valuble. The kernel development people (and anyone developing software for linux, for the matter) are very receptive to bug reports, and really need them to help them improve their software. 2.4 is only (relatively) close today because of peoples' bug reports. Submitting reports is not a waste of your time. You'll be helping yourself (when the bug is fixed), and you'll be helping all of us. So please, don't think it's a waste, or that nothing's going to be done. While calling MS with a bug report may be a waste of your time, in this case it isn't.
I know it won't be included in the kernel for 2.4, but is there a version of it ported to 2.3.x? I took the available patches and tried applying them, but too much of the kernel fs internals have changed. So, before I sit down for several hours trying to understand everything about filesystems in the kernel, has anyone already done it? I really want to try 2.4, but I just can't go back to ext2. Once you've tasted it, there is no other. ;)
I use CFS - Cryptographic Filesystem, personally.
Admittedly, this still leaves you vulnerable to the script kiddie that gets in and can get at /crypt; that is probably still nearly as protectable as kernel-based approaches...
If you're not part of the solution, you're part of the precipitate.
We created some Wonderful 2.4 HOWTOs:
Linux Volume Management - or 'How do I grow my filesystem by buying more disks'
Linux Advanced Routing and Traffic Shaping or 'How do I run my internet exchange with nothing but Linux and keep bandwidt for myself
Any mention of filesystem crypto? Those of you with near terabyte collections of mp3's *ahem* might be interested in this.. that means that "the man" will have a hard time getting into your goodies when the plug gets pulled, or when joe skript kiddie decides it's time for some fun with the latest edition of Redhat.
So, what's the story? Easy filesystem crypto? Where easy == transparent..
Kudos!
..don't panic
If it does maybe then people will take Linux seriously on the desktop. If it doesn't work soon Linux is doomed to a life among us geeks. I love Linux, but right now I would NOT want to even dream of having Linux on my mom's computer. I have enough trouble supporting all the stupid questions I get for Winows. Can you imagine telling Id10T users how to log on as root to run fsck because they turned the power instead of typing halt. Maybe 2.4 will help the learning curve in some ways.
My current workstation:
/etc/fstab: /var/shm shm defaults 0 0
/dev fs, there's many (interesting) files to play with, the cooles imho being /dev/microcode. You can basically rewrite the binary data built in to your processor. While i dont think 'cat /dev/random > /dev/microcode' would be such a good idea, there's a good page with info on this here
nooky:~$ uname -rmpv
2.3.99-pre4 #4 SMP Thu Mar 30 13:14:58 PST 2000 i686 unknown
I've been using the 2.3 series since 2.3.32, and have had tremendous success with them. Several things to note that _weren't_ in the article:
1) sysvipc has changed. if you use anything that depends on shm, you'll need to add this to your
none
2) Along with the new
--BlueLines "The cost of living hasn't affected it's popularity." -anonymous
This is a free minix-like kernel for i386(+) based AT-machines. Full
:-)
...)
source is included, and this source has been used to produce a running
kernel on two different machines. Currently there are no kernel
binaries for public viewing, as they have to be recompiled for different
machines. You need to compile it with gcc (I use 1.40, don't know if
1.37.1 will handle all __asm__-directives), after having changed the
relevant configuration file(s).
As the version number (0.01) suggests this is not a mature product.
Currently only a subset of AT-hardware is supported (hard-disk, screen,
keyboard and serial lines), and some of the system calls are not yet
fully implemented (notably mount/umount aren't even implemented). See
comments or readme's in the code.
This version is also meant mostly for reading - ie if you are interested
in how the system looks like currently. It will compile and produce a
working kernel, and though I will help in any way I can to get it
working on your machine (mail me), it isn't really supported. Changes
are frequent, and the first "production" version will probably differ
wildly from this pre-alpha-release.
Hardware needed for running linux:
- 386 AT
- VGA/EGA screen
- AT-type harddisk controller (IDE is fine)
- Finnish keyboard (oh, you can use a US keyboard, but not
without some practise
The Finnish keyboard is hard-wired, and as I don't have a US one I
cannot change it without major problems. See kernel/keyboard.s for
details. If anybody is willing to make an even partial port, I'd be
grateful. Shouldn't be too hard, as it's tabledriven (it's assembler
though, so
Although linux is a complete kernel, and uses no code from minix or
other sources, almost none of the support routines have yet been coded.
Thus you currently need minix to bootstrap the system. It might be
possible to use the free minix demo-disk to make a filesystem and run
linux without having minix, but I don't know...
------------
a funny comment: 1 karma
an insightful comment: 1 karma
a good old-fashioned flame: priceless
this sig limit is too small to put anything good h
I find this has everything to do with the particular distribution's target audience.
For instance, with Mandrake you see shipments of not only the latest stable kernel, but pratically all of the available patches towards the next stable release already applied. Who is Mandrake's target audience? Desktop users - or at least the Desktop/Workstation sector is where you'll find many of the Mandrake installations. Mandrake is known for it's ease-of-use and execellent default window manager settings. I would predict that many of the desktop/gamer/cutting edge-targetted distributions will be shipping with 2.4 shortly after it becomes available.
The more corporate-oriented distributions will quite likely wait until 2.4 has had a chance to stabelize a bit longer, especially the distributions targetted at servers - Redhat for instance, and probably Caldera and Corel as well.
I further predict that the distributions that aren't shipping with 2.4 running by default will at least include a 2.4-compiled kernel available for install at some point during the installation process - not only so that they are not perceived as "behind the pack" but also to get some feedback from their userbase as to where problems have cropped up.
Also bear in mind that XFree 4.0 was released recently, and at this point can't be considered fully stable/complete since many video card drivers have not been released yet, and there has not be enough time to throughly test it. Many distributions will wait for their next point-oh release until after 2.4 and XFree 4.0 have stabilized somewhat
And of course, some distributions like -ahem- Debian will simply wait until the next stable kernel is just around the corner before a release in say... q1 2001 (c:
Ouch! Bruce... put down that stick! I was just kidding... ouch!
-Cycon
Your Brain + EEG + LEGO Robots = Brainstorms
One of the things that has worried me about this 2.4 freeze was a number of features that were added last minute (during the 2.3.99 series). One of these such features was a multilink implementation by Paul Mackerras.
What was disturbing about this feature add is that several people had been working on other implementations of MLPPP for Linux for some time (including, but not limited to: The PPP Multilink Protocol (MP) for Linux, Chris's Multilink PPP for Linux, and Babylon (by Spellcaster)).
I am particularly touched by Chris Pascoe's e-mail:
What disturbes me is that I wrote Paul, the present PPP driver maintainer for Linux, asking him if he would like help testing his new MLPPP code before 2.4 final and if he has a new PPP daemon capable of using his new driver. I, and Chris Pascoe are still waiting for the reply. This seams very counter-productive to the open-source model of development. The last time I mailed Chris he said that we was going to get one of his friends to go knocking on Paul's door at the college where he resides to see if he is still alive...
Anybody heard more on PPP?
Anybody seen anymore recently added code during freeze time?
-AP