Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
Linux Software

Linux 2.3.46 Released Unto the World 234

Posted by Hemos from the letting-it-loose dept.
jschauma writes "I just saw on freshmeat that Linux 2.3.46 is out - thought I'd share the news. Freshmeat also has the changelog online. " One step closer to 2.4.
This discussion has been archived. No new comments can be posted.

Linux 2.3.46 Released Unto the World More

Linux 2.3.46 Released Unto the World

Comments Filter:

  • Woo Hoo! (Score:2)

    by pb ( 1020 )


    Oh wait, I could have just read freshmeat.

    What's supposed to be in 2.4, anyhow? SMP updates? Maybe my TV card will work better...
    ---
    pb Reply or e-mail; don't vaguely moderate [152.7.41.11].
  • I suggest running a 2.3 kernel.. It's fun. A lot of people don't seem to want to run devel kernels for fear of them crashing - but I've had very, very good luck w/the 2.3 series.

    --
    blue
  • I partially agree here. But there is a reason it's called a beta kernel. If you've just installed linux and don't have too much experience (ie i've never even RECOMPILED my kernel), then DO NOT use a 2.3 kernel. However, if you are not worried about crashes, or if you have some need to run 2.3 (USB, better TV support, etc) then go for it. But don't cry when it doesn't work :)
  • I got burnt after a 2.1 series kernel that scrambled a filesystem with my mp3s on. Recovered most of em, but some of em got mixed up, so I got about 10secs of each mp3 in the directory...
  • Boy did the Win2000 get ignored today or what?

    Cool that a Linux release came today to spoil the "party".

    Although this should have us worried:

    http://www.tpc.org/new_result/ttperf.idc

  • Right, Windows 2000 escapes^H^H^H^H^H^H^His released tomorrow, the kernels are close to completion. Hmmm... [ntk.net] imminent death of the internet?
  • your TV card may in fact work better.... Gerd has done a lot of work on the bttv driver, but many of the changes aren't merged yet... go to the xawtv site and grab the latest 7.x version of bttv - you may be pleasantly surprised.

    (It will work with 2.2.x as well, with some patches)
    ----

  • Mention that it's a development kernel (Score:3)

    by CentrX ( 50629 ) on Thursday February 17, 2000 @02:05PM (#1263975)
    If only because a huge number of naive Linux users out there read slashdot. Apparently, they get confused into thinking that Linux isn't stable because the kernel isn't stable because they're using a development kernel. Aside from arguing the merits of this being posted on slashdot. I do think that it should at least be mentioned that it's a development kernel and so is not as stable as a stable kernel.

    Chris Hagar
  • ...2.3.978a-ac-dc is far more important news indeed. Gee. Dozen of changes to ancient network drivers and to devfs. I am thrilled.. Come on. THis is business as usual. Win2K launch is not (as much as I dislike the system). Have a sense of perspective.

  • The big news... devfs (Score:5)

    by Chris Pimlott ( 16212 ) on Thursday February 17, 2000 @02:09PM (#1263977)
    devfs is in.

    devfs = Device File System. Instead of makedev and having dozens and dozens of device files in /dev/, 99% of which you don't use (ya sure, I have 20 ide partitions. And 5 sound cards. And 9 SCSI CD-ROM drives...), you mount /dev as devfs, and only see the devices you have. Simplifies life a great deal. More info at the devfs overview [lwn.net]. Devfs has existed as a patch for a good time now but Linus had issues with it.

    Congratulations to Richard Gooch on his efforts over all this time and his tireless dedication to getting in into the kernel. Hats off to you.
  • i ahd the same problem when i tried to compile it. took out all the unnecessary block drivers, and it still wouldn't compile. i pseudo-traced it down to the generic pci-ide controller driver, but hey, i could be way off. i can't use this kernel without pci-ide, so i guess i'm stuck at 2.3.42, a GREAT 2.3-series kernel as it were. 2.3.42-45 had trouble with my sound/ethernet, and 2.3.46 won't compile. oh well. i love the new Voodoo3-frame buffer driver in these kernels. it's so cool. there is firewire/usb as well, and the usb works pretty well, although since i planned ahead i have no USB devices to use regularly(if that can be called planning ahead...) oh well. if you're looking for a good 2.3 series kernel, definitely get 2.3.42, probably the best one i've found out of the lot. 2.3.39 was also decent.
  • I submitted this as a story, but it was rejected: An article in the New York Observer about dot.com rip-offs [observer.com] that is very harsh on the VA-Andover merger. Quote: "Prior to the I.P.O., [Andover] insiders had only paid a grand total of $15.7 million in cash into the company. Now they are being handed back $60 million as compensation. What a deal. As for investors in VA Linux, they are getting hosed."

    I would be interested to see some opinions on that. I know this is off topic, but isn't the big argument against censorship: "Slashdot is the readers, not the VA corporate control".

  • (Oh yes, and it's labeled experimental) (Score:3)

    by Chris Pimlott ( 16212 ) on Thursday February 17, 2000 @02:13PM (#1263982)
    Forgot to add that it's listed as experimental, so you won't see it in the options if you don't have experimental turned on.

  • This new kernel release actually *IS* news. (Score:5)

    by jonathan_ingram ( 30440 ) on Thursday February 17, 2000 @02:13PM (#1263983) Homepage

    Normally, I agree that announcing the latest development kernel on Slashdot is a little silly - after all, if you're running the devel kernels, you know where to look for them.

    However, this kernel release IS newsworthy. Why? Well, take a look at this posting to the linux-kernel mailing list:

    [PATCH] devfs v158 available [kernelnotes.org]

    If you can't be bothered to follow the link, here's the important sentence from that posting: This is the patch that was sent to Linus and included in 2.3.46-pre5. That's right boys and girls, DevFS is now part of the standard Linux kernel. This is wonderful news, and amazingly hasn't yet sparked off any great flamewars on the mailing list (those of you that read the list will know that mentioning DevFS on it has seemed akin to posting about atheism on an evangenical Christian newsgroup). For more information about DevFS, have a look at Richard Gooch's kernal patch page [csiro.au].

    I'm still amazed that this has happened.

  • It was released today... My office got it's first shipment. Frightening.

    --hunter

  • Did W2K not launch? (OT) (Score:4)

    by acarey ( 34175 ) on Thursday February 17, 2000 @02:14PM (#1263985)
    Did I miss something? Did Microsoft cancel the W2K launch at the last moment? Did the spooks cover it up? No, everybody else is reporting the news except /.

    If the W2K launch isn't "News for Nerds", then I'm afraid I don't know what is. A Linux kernel update is "News for Nerds", but the most anticipated OS release in the past two years isn't?

    Today is a new low for /. Congratulations. Long live media bias!

    (And don't flame me with "but this is a Linux news site" - the site specifically says "News for Nerds", not "News for Linux Nerds".)

    Open soure. Closed minds. We are /.

  • Freshmeat does certain things well. (Score:3)

    by Hrunting ( 2191 ) on Thursday February 17, 2000 @02:15PM (#1263986) Homepage
    This is a rant, but this is not a troll, flamebait, nor is it off-topic. So read with open ears.

    I am sick of seeing linux development kernel upgrades posted on Slashdot. I think if Slashdot is going to get in the business of announcing minor software updates, they should announce all software updates. I recognize the need for Slashdot to mention major software upgrades, such as GNOME hitting the 1.0 plateau or KDE hitting the 2.0 plateau, but announcing every single minor development kernel revision is ridiculous. That's why we have places like Freshmeat, and that's why we have things like Freshmeat slashboxes. It's that simple.

    But, I can understand how this might be of some value to people who can't figure out how useful Freshmeat is or even know it exists or just plain don't like it. I like people to be constructive, not destructive, so I propose that Rob et al develop a new Slashdot topic like 'kernel-development-update' and make it real specific to development kernel announcements. I like reading about major proposals to the kernel, so that shouldn't be in there, and I certainly don't want to filter out all Linux related news, so Linux development kernel updates shouldn't be under that heading. Give it a cute kernel icon, like a corn kernel or something. It's just inane to make these announcements every week or so for something that is in development. Yes, it's the road to 2.4, but let's wait until we get a 2.4pre kernel or something and the end is in sight. With Linux development kernels having a history of getting into the hundreds in minor version numbers, we don't need these. Freshmeat's good enough.

    And for those who are going to say that the universe doesn't revolve around me (and I'm sure you're out there), Stephen Hawking postulated that the universe could be expanding from any point, and so right now, I'm designating that point as me. Call it the Hrunting Corollary.

    *wheeze wheeze*

  • It's a typo (Score:4)

    by sighup ( 1594 ) on Thursday February 17, 2000 @02:16PM (#1263987)
    Just edit drivers/block/ll_rw_blk.c, and change line 237.

    back_merges_fn should read back_merge_fn (IE, remove the 's')

    This has been posted to linux-kernel as the fix, and it works for me.
  • Yeah I'm sure that the Win2k camp is crying their eye's out today because the xxxxxxxx version of a dev release came out..
  • I really don't like lists of changed files. Whatever happened to the good old changelog that says *what* was changed? I much prefer that. Of course, It is more work, and if it really bothered me that much I could do something about it.
  • My favorite change going into 2.4 is the halving of the filesystem caching structures. Instead of a read buffer and a write buffer, there is a single buffer. That makes the buffering data files use far less memory. Systems like mine [starshiptraders.com] will benefit greatly from this because the key to good performance at my site is getting all the game files in memory to avoid the awful penalty of disk IO. ;)

    Other changes are detailed in a story [linuxtoday.com] over on Linuxtoday [linuxtoday.com].
  • According to linux-kernel, try this:

    diff -u linux/drivers/block/ll_rw_blk.c.orig linux/drivers/block/ll_rw_blk.c
    --- linux/drivers/block/ll_rw_blk.c.orig Wed Feb 16 20:15:56 2000
    +++ linux/drivers/block/ll_rw_blk.c Wed Feb 16 20:45:56 2000
    @@ -253,7 +253,7 @@
    INIT_LIST_HEAD(&q->queue_head);
    q->elevator = ELEVATOR_DEFAULTS;
    q->request_fn = rfn;
    - q->back_merges_fn = ll_back_merge_fn;
    + q->back_merge_fn = ll_back_merge_fn;
    q->front_merge_fn = ll_front_merge_fn;
    q->merge_requests_fn = ll_merge_requests_fn;
    q->make_request_fn = NULL;

    Well, you get the idea even if it looks like crap on slashdot.

    (Original linux-kernel post [kernelnotes.org])
  • OK, so maybe Freshmeat does post this stuff. Ask yourself, how many servers do you have in your company? Have many drives in the array? Redundancy is OK. If two sites do the same thing, then what's the harm? Slashdot reports on stuff that is on other news sites, but no one says, "Hey, News.com already posted this. Why don't I just check there?"
  • No Linux system?

    It's a price for performance... no linux system??

    This looks bogous to me.
    --------------------------------
  • Yeah... I definetly don't like the idea that Microsoft magically appeared at the top of the list using software that's yet to be released. It does seem rather fishy, to see it there and here nothing of it from them... Is this part of a stealth marketting effort on microsofts part? Or is it them doing what they've accused Oracle Sybase Informix and IBM for quite a while now? (Doctoring their applications to perform the TPC benchmarks exceptionally well).

    It just doesn't seem convievable that an 8CPU system from Microsoft could compete head one with one with 64 processors from sun or 128 processors for SGI... ESPECIALLY given the later two's 64 bit operating systems and Windows 2000's decidedly 32 bit addressing.

    I'll stop now...WAY OFF TOPIC, I'm sorry!!!

  • There is a one line typo in ll_rw_blk.c. The name of a variable had an 's' added to it accidentally. See this linux-kernel message [kernelnotes.org] or just apply this diff:

    diff -u linux/drivers/block/ll_rw_blk.c.orig linux/drivers/block/ll_rw_blk.c
    --- linux/drivers/block/ll_rw_blk.c.orig Wed Feb 16 20:15:56 2000
    +++ linux/drivers/block/ll_rw_blk.c Wed Feb 16 20:45:56 2000
    @@ -253,7 +253,7 @@
    INIT_LIST_HEAD(&q->queue_head);
    q->elevator = ELEVATOR_DEFAULTS;
    q->request_fn = rfn;
    - q->back_merges_fn = ll_back_merge_fn;
    + q->back_merge_fn = ll_back_merge_fn;
    q->front_merge_fn = ll_front_merge_fn;
    q->merge_requests_fn = ll_merge_requests_fn;
    q->make_request_fn = NULL;

    (modulo the way that Slashdot mangles quotes, of course.)

  • How about Kernel Slashboxes? (Score:5)

    by razvedchik ( 107358 ) on Thursday February 17, 2000 @02:24PM (#1263998)
    That way, if you really care, you can just go into preferences and turn on the box, so you can see the latest, greatest kernel version.

  • Weeeeellll... since VA and Andover are both publicly held, their administration can't do anything that doesn't benefit the stockholders financially. If they *do* do that, they'll get their pants sued off, and lose their jobs.

    That would be no good. I don't have any personal opinion on the merger, except that VA now owns both sourceforge and server51, so the only non-VA free development platform-type-site that I know of is openprojects.net. But that's a different point altogether.

    Oh, and themes.org and Slashdot are now owned by the same people...isn't that exciting? But, of course, the Andover/VA Linux staff has no say in what gets posted and what doesn't.

    Oh wait, roblimo is an editor of AndoverNews. Hrm. He's also been with them for a long time, I do believe.

    So are we really to believe that Andover (through roblimo) doesn't have any say in what gets posted on slashdot?

    And...what about the guy who posted to the original merger (VA-Andover) thread, from valinux.com, who got an automatic +4(!), without any moderation. Hmm.

    Some of it seems a bit suspicious. But that's the way I am sometimes, eh?

    -ed fisher.
  • ... either.

    And for one, I don't get it. NASA news is news for nerds, and its generally about stuff that matters.

    But, see, Slashdot stories are selected by Rob and Hemos and the gang. Its stuff that's interesting to them. Thus, Slashdot is a cult of personality - the personalities that are Hemos, CmdrTaco, etc.

    So, tough.

  • Re:tulip drivers (Score:1)

    by Anonymous Coward
    I would like to know this as well, unforunately I cant seem to find any info on this. If someone can help, mail me [mailto]. My NIC is unsupported in the 2 year old tulip driver used in 2.3.45. If anyone has any info or suggestions on who to inform please tell me. I would LOVE to be using 2.3.xx on both of my boxes(which both use this card, Linksys 10/100TX w/ Wake on Lan Header).
  • I am sick of seeing linux development kernel upgrades posted on Slashdot. I think if Slashdot is going to get in the business of announcing minor software updates, they should announce all software updates. I recognize the need for Slashdot to mention major software upgrades, such as GNOME hitting the 1.0 plateau or KDE hitting the 2.0 plateau, but announcing every single minor development kernel revision is ridiculous. That's why we have places like Freshmeat, and that's why we have things like Freshmeat slashboxes. It's that simple. But, I can understand how this might be of some value to people who can't figure out how useful Freshmeat is or even know it exists or just plain don't like it. I like people to be constructive, not destructive, so I propose that Rob et al develop a new Slashdot topic like 'kernel-development-update' and make it real specific to development kernel announcements. I like reading about major proposals to the kernel, so that shouldn't be in there, and I certainly don't want to filter out all Linux related news, so Linux development kernel updates shouldn't be under that heading. Give it a cute kernel icon, like a corn kernel or something. It's just inane to make these announcements every week or so for something that is in development. Yes, it's the road to 2.4, but let's wait until we get a 2.4pre kernel or something and the end is in sight. With Linux development kernels having a history of getting into the hundreds in minor version numbers, we don't need these. Freshmeat's good enough. And for those who are going to say that the universe doesn't revolve around me (and I'm sure you're out there), Stephen Hawking postulated that the universe could be expanding from any point, and so right now, I'm designating that point as me. Call it the Hrunting Corollary. *wheeze wheeze* [ Reply to This | Parent ] Yeah, well, there's no NASA news hardly ... by torpor (Score:2) Thursday February 17, @07:27PM EDT How about Kernel Slashboxes? by razvedchik (Score:1) Thursday February 17, @07:24PM EDT If you talk about Freshmeat.... (Score:2) by blogan (slashdotter(at)network(dash)geek(dot)com) on Thursday February 17, @07:19PM EDT (#41) (User Info) http://www.Network-Geek.com/ OK, so maybe Freshmeat does post this stuff. Ask yourself, how many servers do you have in your company? Have many drives in the array? Redundancy is OK. If two sites do the same thing, then what's the harm? Slashdot reports on stuff that is on other news sites, but no one says, "Hey, News.com already posted this. Why don't I just check there?"


    Nathaniel P. Wilkerson
    NPS Internet Solutions, LLC
    www.npsis.com [npsis.com]
  • Well, I've had no trouble compiling 2.2 series kernels with egcs.. but that doesn't mean that I go around recommending the use of anything but gcc 2.7.2.3
  • No... I actually am not 100% fond of Linux... I'd much prefer Macs any day of the week for anything that requires me to actually interact with a computer.

    It seems a bit off to me because even if Microsoft did produce the most streamlined OS in the world... I simply doubt that Sun, IBM, Oracle, and all the other "enterprize" class vendors would be unable to to beat Microsofts results using what they have available to themselves -

    Multitudes of more CPU's, each of which is more powerful than anything Intel offers,
    64 bit memory addressing = enough memory to conceivably hold the database in RAM.
    Years more experience in that arena...

    No... SOmething seems greatly wrong, to me. I KNOW linux can't match any of those numbers, simply because it doesn't run, or isn't optimized to run, on those classes of machines, and the software isn't available for them at that level...
  • I suspect (from your comment) that you've already read this, but, for the rest of us... ;-)

    For a nice summary, see http://linuxtoday.com/stories/15936.html > [linuxtoday.com]

    It may be updated periodically, thus rendering this link obselete, tho'.

    Enjoy. (Yes, TV Card support may be easier, the article says... ;-)

    _______________________________

  • Weeeeellll... since VA and Andover are both publicly held, their administration can't do anything that doesn't benefit the stockholders financially. If they *do* do that, they'll get their pants sued off, and lose their jobs.

    I think that is a quite naive statement. The history of Andover so far is: They made a bunch of deals, went IPO, and hugely profitted from selling the company. It is not quite clear, if people who actually bought the shares profitted so far, since the stock price went downhill from the start [smartmoney.com], with a slight bump upwards from the sale to VA Linux.

    The question, if something benefits the shareholder is quite murky. The article states the opinion that the Andover deal hurt VA Linux shareholders. VA Linux has a different view on this for sure. Assumed I am a VA Linux shareholder, I guess I have barely a chance to successfully sue VA Linux. IANAL, though.

  • Alright, the list time /. posted a kernel release on the board, I defended it saying that an occasional heads up was important for some projects. But this is going a bit far. I was fine with 2.3.41 because it had been a while since the last post, but here's another only .05 releases later? Sure a lot of dev releases are important, but here are a few suggestions that could fix the kernel dev release problem.
    1. Definatly announce full releases.
    2. Announce and pre- builds like 2.2.0pre9
    3. Post about major feature additions.
    4. Announce the occasional dev build just to keep people aware of the project.
    It would also be more helpful if, instead of saying, here is another dev kernel, people would post about some of the stuff in the kernel, bug-fixes, new feautures, articles about new subsystems, etc. Not everyone is on the kerneldev mailing list you know.

  • Yes, I have to agree. Same goes for the BSD releases, and everything else. If it's big news, /. readers need/want/must know about it. But if it's little routine stuff like development steps, and even patch levels of stable releases (unless some major and important fix included), it shouldn't be mainline headlines.

    I would like to see a little box that shows what the latest/greatest releases of major OpenSource freeware is, be it Linux, distributions of Linux, BSD, KDE, Gnome, or whatever big popular software it might be. In our preferences we could then pick and choose what software titles we want the /. chipmunks to put in the box for us. Then add some code that highlights the changes we haven't seen, yet

  • > It's a price for performance... no linux system??

    Actually, the link supplied isn't price/performance, it's flat performance. That's what makes it especially relevant. And if you'd care to look at the numbers, this $4mil Compaq/Microsoft box is chunking nearly two-thirds more transactions per minute than the drastically more expensive IBM and Sun servers located down the line. Oh, and this is an independent scientifically audited benchmarking site, so this ain't no FUD. And to specifically answer your question, if SMP under Linux and MySQL worked at all well, you'd see them on the other list, here [tpc.org] where the list is not blanketed by performance, but rather by price. Eat your heart out.

    Jake

  • I would strongly recommend that you avoid this kernel unless you know precisely what the consequences may be. The kernel is in the middle of huge changes. The entire networking layer has been yanked out and replaced. Some of the network drivers are not updated yet. The block device interface has changed, and this is still rippling through the source. RAID code is still landing, and doesn't currently compile.

    Aside from that I think it is very clear that these kernels do not undergo even the most cursory testing. The typo in the ll_rw_blah_blah.c means that nobody even tried compiling this kernel before release. If they had that typo would have been caught.

    -jwb

  • I was curious about this, so check some numbers last night from the official filings of Andover. Our good old friend CmdrTaco currently owns 111,111 Andover shares, and made a cool half a million from the Andover sale in cash payments alone. His profit from the Slashdot sale in total is probably around $10 million. Did he deserve that? Compared to the workers in Asia that assemble the machines on which this wonderful revolution is running on and get a buck per hour probably not. But compared to what he could have made, if he would have been really only after the money, it's probably meager.

    The biggest shareholder of Andover is Bruce A. Twickler, the CEO who owns close to 2 million shares. He made $15 million from the cash payment of the sale alone. He must be a happy man.

    And then there is the random invester who bought shares in Andover because it is such a cool Linux company. Well, the stock went downhill and he is the one who brought in the actual money that is passed out around here. The random VA Linux invester is in the same boat, that stock price is currently $118 from a $320 high.

  • I was there. Here are a few highlights:

    * Celebrity guests were Patrick Stewart and the guy who played Peterman on Seinfeld (predictably, there were lots of jokes about "enterprise" and "engage!"). The Peterman guy played a venture capitalist who made a deal with the demonstrator (playing a startup founder) because he was able to show him his business plan on the plane using the IntelliMirror function to replicate his desktop from the server back at the office.

    * Lots of focus on plug-and-play, use of the Infra-red port on laptops to transfer files, USB compatibility, and Firewire to transfer files from camcorders. Also focused on DVD capabilities.

    * There was a chart with the results of a third-party stress test, which showed that the average uptime for Windows 95 was 2.5 days, NT 4 was 5.4 days, and Windows 2000 was 90 days (and counting...the test machine was still running)

    * Gates announced two new TPC-C numbers putting Windows NT/2000 in the #1 and #2 position for transaction processing performance for the first time (the second one, announced today, used a cluster of 12 x 8-way Compaq servers to get over 227,000 tpmC, the highest number previously was IBM with about 150,000 tmpC, at four times the cost of the Windows 2000 system)

    * There was a demo of a massive web server cluster running Windows 2000, supposedly capable of handling 1.2 billion hits per day. To prove it, curtains were raised around the auditorium to show that the walls were literally covered with desktop systems, all of which were banging on this cluster.

    * The main prop on stage was a giant (i.e. 40' high and wide) laptop. The show closed with the bottom of the laptop lifting up, and underneath was the band Santana, who then broke out playing.
  • i know for one thing, the 2.2 kernel is now including the new AND the old tulip drivers since there's a bug affecting different cards for each version. i really hope that gets fixed before 2.4...

    -l
  • BTW, regarding your development sites - www.netpedia.net is a free dev site and works great for me.
  • I just downloaded 2.3.46 and I can't find any gui in there at all, never mind buggy ones. Which buggy gui are you talking about? the make menuconfig curses gui interface works pretty good since 2.0!
  • The two Compaq systems that appear on the list have 96 and 64 CPU's (Intel P III Xeon's), respectively. The IBM system in 3rd place (just behind #2) has 24 CPU's.

    Conclusion: Microsoft is very inefficient.

  • It's newsworthy because... (Score:3)

    by grytpype ( 53367 ) on Thursday February 17, 2000 @03:44PM (#1264035) Homepage
    ...we're all anxiously awaiting 2.4. Every devel kernel release is getting us closer. It's like a countdown. And there's often something about the release that merits discussion (like devfs in this release). Slashdot is really a discussion forum, it's not a news forum.
  • I think that it is News for *Discerning* Nerds.

    Perhaps it does show some bias, but I think it is (if then by accident!) an act of extraordinary clear mindedness to simply ignore what was essentially a media-manipulation exercise. /. has posted many things on W2K in the past, and both sides had their say, and both sides were accordingly moderated up (and down!). I hope and expect that healthy discussion of That Other OS will continue in the future.

    But really, what can one say about a product launch? I can see it now. "Oh yeah, and That Company finally really officially actually really said that W2K is out, after we've already been discussing the Gold Release for about two weeks." ;-)

    ________________________________________

  • But really, what can one say about a product launch? I can see it now. "Oh yeah, and That Company finally really officially actually really said that W2K is out, after we've already been discussing the Gold Release for about two weeks." ;-)

    Well, amusingly, even after it went Gold and RTM'd, there were still lots of people here claiming it was vaporware because "it wasn't in stores yet".

    A better use of a clue-by-four, I couldn't find.

    Simon
  • My favorite change going into 2.4 is the halving of the filesystem caching structures. Instead of a read buffer and a write buffer, there is a single buffer. That makes the buffering data files use far less memory. Systems like mine will benefit greatly from this because the key to good performance at my site is getting all the game files in memory to avoid the awful penalty of disk IO. ;)

    What!? There wasn't a single buffer all this time? That's scary. Yes, this should speed up disk performance quite a bit. Yay. Anyone know the state of the nfs code, is it still god awful?
  • Sturgeon's Law: "Ninety per cent of everything is crud."
    Slashdot's Law: "Ninety per cent of everything is FUD."
  • I was just browsing good ol' slashdot. I clicked on a _read_the_rest_of_this_comment_ link and I got a different comment than the one I had been reading. I looked around a little, comparing the old version of this page in my browser with the newer version. The following posts seem to have been eaten. They do not even show up on the author's user info page. I apologize for not knowing html well enough to make this look a little better.

    1.) posted by Luyseyal

    http://slashdot.org/users.pl?op=userinfo&nick=Lu yseyal

    Re:Woo Hoo! (Score:1)
    by Luyseyal (swaters_AT_amicus_DOT_com) on Thursday February 17, @07:11PM EST (#74)
    (User Info)
    LWN has some really good information on 2.3.46. can we say devfs??!! can we say new RAID??!! http://lwn.net/2000/0217/kernel.phtml

    -l

    [ Reply to This | Parent ]

    2.)

    Nerds don't run Windows (Score:0)
    by Anonymous Coward on Thursday February 17, @07:13PM EST (#75)

    Only lusers run Windows; therefore, it is not "News for Nerds."

    [ Reply to This | Parent ]

    3.) posted by Zurk

    http://slashdot.org/users.pl?op=userinfo&nick=Zu rk

    Re:No bias here at /. (Score:1)
    by Zurk (zurk@SPAMSUCKSgeocities.com) on Thursday February 17, @07:14PM EST (#76)
    (User Info)

    oh bullshit. if you dont like it dont read here dimwit. yes, a kernel change is interesting for some of us if it has been a slow day. windows 2000 is the most anticipated OS for who ? i dont even use the current release.

    [ Reply to This | Parent ]

    4.) posted by mrsam

    http://slashdot.org/users.pl?op=userinfo&nick=mr sam

    Thank you for your support. (Score:1)
    by mrsam (sam@email-scan.webcircle.com) on Thursday February 17, @07:14PM EST (#77)
    (User Info) http://www.geocities.com/SiliconValley/Peaks/5799/ etrouble/

    Please post all the usual "Why Are You Announcing Kernel Revisions In Slashdot?" flames in this thread.

    Thank you very much.

  • No matter what the header says, in substance, Slashdot is a discussion forum. People send in topics they think will be of interest, and we discuss them. This isn't much of a news forum because (1) it's way too selective, there's a lot of nerdnews that doesn't get posted because it's not interesting enough, (2) a lot of the stuff is months old or isn't news for other reasons (interviews, opinion pieces, Katz, "I found this real neat Lego site," "I have something more to say about Star Wars," etc). (Compare Slashdot to the Register, which is a real news site.) I'm not saying it's a bad thing, I'm saying what matters is not what is newsworthy, but what is discussion worthy, which development kernel releases are (IMHO).
  • There isn't any Linux system in this list because there is no single Linux company member of TPC. (Membership cost's money!)
    NT is cheeper than ("real") UNIX, esp. for entry level applications/systems. (I believe this is generally agreed upon). This is wy we only see NT on the low performance list.

    The question remains whether choosing for a low performance (NT) solution will be cheaper in the end, but that's a different story...
  • My fellow Americans, this kernel release marks a glorious occasion in our history. It is with great joy that I recognize the accomplishments of Mr. Torvalds; he is a fine, young man with a purpose... a purpose, in life. Only in America could a poor, disheveled Finnish refugee, with nothing but the shirt on his back and a dream, an American-inspired dream, come to our great nation and create what is probably the fastest-growing, most important new technology for the new millennium, the most important thing to ever be created for what are known as "computers," possibly since my vice president created the Internet.

    My fellow Americans, please join me in congratulating Mr. Torvalds with his recent accomplishments. He is an inspiration to all.

    Thank you, and God bless.

  • No. Because x86 systems are notorius for lacking registers. Because of the lack of registers the memory bandwidth that they do have is severly limited. Used for storing things to temporary memory addresses that on risc systems you could just use a spare register. Aside risc workstations/servers happen to have better memory bandwidth/throughput. I could go on for quite a while but I won't. The original posting was fud. Ignore it. If it even has a glimmer of truth to it which I doubt it does i'd say that there's a reason behind it. Perhaps you should look at an actual real world test. e-commerce seems to be all the rage these days. Lets benchmark that and see who wins. Myself i've worked with hpux just a little (Working with the parisc-linux guys) it's nasty to work with from a user perspective. But I think solaris has a good chance of kicking nt2k's ass.
  • We had to move the DB to another machine for a moment then move it back. I'm afraid those comments were apprently posted while we were copying the DB back -- we caught most of them but it appears these few slipped in while the copy was happening. Thanks for reposting them. I would moderate you up but since I'm posting here I can't.

  • Re: At the day Win2000 launches... (Score:5)

    by coyote-san ( 38515 ) on Thursday February 17, 2000 @04:34PM (#1264049)
    <b>Have a sense of perspective.</b>

    We do. You're the one with things out of balance.

    Is this the first day we could acquire W2K? No, it's been released to OEMs for weeks.

    Is this the first day we could get a peek at W2K? No, release candidates have been out for many months. I'm tempted to say years.

    Does Win2K redefine the fundamental paradigms used by software? Nope, it's a incremental change from NT4, but it doesn't have true innovations like GUIs or a NOS, and it doesn't even have false invocations like doing everything through the fully integrated (but still available as a separate product in stores!) web browser.

    On the other hand, the inclusion of devfs *will* go a long way to heading off a critical problem. Users notice that devfs eliminates the need to have thousands of files in /dev. Kernel developers know that one of the *real* wins with devfs is that we can have more than 256 major devices and 256 minor devices. That will make it *much* easier to provide fine-grained support to SCSI, USB, and similar devices, to implement "volume" managers where you associate each removable media object with a unique "minor device number," etc.

    The fact that I can get Win2K "in a box" instead of "OEM'd" today does not really change my life. If I really needed Win2K, I would already have it.
    The fact that Linux now includes devfs *does* significantly change my life because some very cool kernel modules and applications are now much easier to write without requiring the end user know how to apply a kernel patch.
  • Funny that you mention it. Back when Slashdot got a Netfinity server from IBM and people were up in arms about sacrificing integrity for profit. Of course, that was before the whole Linux market hysteria.
  • Now if this wasnt Slashdot.... news for Linux nerds we might have seen this as a true story.

    The one reason I started using Linux 3 years ago was due to the Windows camp ignoring all other OS and technology out there.

    I dont like this "boys club" linux is turning in to.
  • If you actually had the ability to comprehend, then you would understand that I was stating that people probably read the same story on multiple sites, so what's the problem if they read something on Slashdot and Freshmeat?
  • Not really. All benchmarks should be taken with a hefty pinch of salt, especially TPC-C

    Linus would be the first to admitt that Linux is not all things to all people. However fine grain locking in 2.4 should make SMP suck slightly less.

    The ProLiant systems were built using 3 clustered boxes each with 32 550MHz(2MB L2 cache) processors. The Alpha tested is not representative of the current state of the art, they were only 21164's at 612MHz. The lastest is a 21264 at 700Mhz with 16MB L2 cache! This which would smoke the Intel based box. Sun is a generation behind in the processor stakes, Ultra Sparc II max out at 400MHz and do not have as much cache so it is not suprising that the system is slower. This should be corrected when the Ultra Sparc III is brought out. It is a testement to Solaris that they are competative at all.
  • Notice how they've been posting a lot of editorials and features lately?
    ---
    "'Is not a quine' is not a quine" is a quine [nmsu.edu].
  • My wife threw up today due to her being pregnant. I tried getting CNN.com to post this but they told me

    "go post that crap on slashdot. tell them your wife runs linux and they will mod your story up big time"
  • >Thanx ;) Ain't it nice to have the source code Ain't it the truth! Much better than some released systems with thousands of bugs (and this isnt' released yet), where your only option is to wait months for the first service. I'm convinced!

  • Does anyone know whether 2.4 will include support for > 2 GB files on 32 bit machines?

    It is surprisingly difficult to find a definitive answer on this. I'm aware of the LFS patches at ftp://mea.tmt.tele.fi/linux/LFS/ [tmt.tele.fi], but I need everything to 'just work', including iostream libraries etc...

    In redhat 6.2 or 6.3, what are the chances that a tar cvzf stuff.tgz /data/* will just work without truncating my tarball to 2GB?

  • >Why is it inconceivable to you? Because your blind linux zealtry doesnt allow you to believe anything else? Nope. but my 20+ years in this business tells me to be VERY suspicious and read the fine print, not the press release.

  • There was a chart with the results of a third-party stress test, which showed that the average uptime for Windows 95 was 2.5 days, NT 4 was 5.4 days, and Windows 2000 was 90 days (and counting...the test machine was still running)

    You mean all the windows 2000 servers were still running, don't you? I'm assuming they're not finding the "average uptime" by counting how long one machine stays up...

  • There was a chart with the results of a third-party stress test, which showed that the average uptime for Windows 95 was 2.5 days, NT 4 was 5.4 days, and Windows 2000 was 90 days (and counting...the test machine was still running)

    I wonder if there is more information about this or these particular stress tests. If they are not internal Microsoft tests, it would be interesting to run them on the Linux development kernels.

    How much dedicated stress testing do people really do for Linux development kernels? I know Microsoft "prides itself" in the stress tests it runs on its internal NT builds.

  • Kurt, Not all the posts made it back after the move. Can you guys refresh or try again??
  • Note that those two Compaq systems won't even be available until Aug this year.
  • Let's face it. In the dawn of the Windows 2000 release, this kernel update is a nerdy guy in a party going "hey look at me everybody! I just updated the kernel! let's party!@#"
  • Hmm, does anybody have a better understanding of this benchmark, I don't understand what the # of clients means. I ask this because if you look, the top one has 36 clients with 36 CPUs/client! The second one is 24 clietns with 24 CPUs/client! What the heck? All of the other systems have 1 or 2 processor clients. I don't get it, why use a huge ass server for a client? Do you get a higher result with bigger clients? I do notice that #2 has about 40000 more users than the #3 system, and #1 has about 100000 more users.

  • stuff that matters? (Score:2)

    by Anonymous Coward
    Seems its all a matter of opinion, I for one read a LOT of news sites, and i honestly can say i did not get a hint that win2k was released today. I thought it would get released sometime this week, but i never saw any indication it was today. If it was such a huge release, why report it? seems pretty redundant to me. Same goes for the kernel posting. I don't come to slashdot just so i can get a link back to news.com or cnn.com about some overhyped win2k release, i come to slashdot to get the tech goodies that other news agencies take 6 months to realize whats going on. its all a matter of opinion, you people are WAAY too sensitive. many post without thinking. and a flame war starts. *sigh* too bad... nate aphro@aphroland.org
  • Maybe I've just been lucky (sort of). What sort of things have been fixed? The changelog doesn't seem to be too specific. The only problems I've had (and they don't seem to be occurring anymore) are xawtv not working in overlay mode and, once, the bttv drivers took my system down hard. Insmod'ing the drivers in the proper way seems to have fixed that.

    - A.P.
    --


    "One World, one Web, one Program" - Microsoft promotional ad

  • Can you please produce a source for this benchmark?

    Also, we would be well advised to approach this comparison with a bit more composure than we did the last time around.

  • I am sick of seeing linux development kernel upgrades posted on Slashdot. I think if Slashdot is going to get in the business of announcing minor software updates, they should announce all software updates.

    Well Nate, I am sick of seeing NPWIII Posting your never ending trolls on Slashdot. I think if NPWIII is going to get in the business of trolling your advertising without adding anything but flame to the thead, You should buy a banner and put your money where your troll is.

    Yeah I know -1:Flamebait, mark me down.

  • Try the kernelnotes.org slashbox. It gives the current stable and development kernels as well as the old stable and development (i.e. 2.0.x and 2.1.x), with links to the changelogs and all at kernelnotes.
  • I believe that there never was a limit in the filesystem code. I don't know if the OS limitations have been lifted (I suspect they have). But a lot of applications won't "just work" and they won't "just work" on 64-bit machines either.

    The problem is that to fix the 2GB limit you need to change every program to understand that they cannot just seek to an integer (or long) location. This kind of type change simply cannot be easily done, there are too many problems that can arise. Therefore even on 64-bit machines many programs only use 32-bit constructs and will require workarounds.

    A sample problem? Perl programmers on 64-bit machines are advised not to open large files directly in Perl, but rather to open pipes to and from processes that can read and write the files. Fixing that requires compiling perl (the program is lower case) with still-experimental 64-bit support. There is nothing that could possibly be done at the OS level about this - the issues are internal to how perl was written.

    Cheers,
    Ben
  • I had the opportunity to see Mr. Gates give the keynote at the spring Comdex (it is put on by ZD and named "Windows World" after all) last year.

    He opened with a video showing the previous year's (Spring 1998) presentation where they demonstrated the wonderful ease Windows 98 installed drivers for USB devices. He showed a video of that presentation and how it failed. I wasn't there in '98, so I assume that video was legit. A scanner (I think) was plugged in, Windows 98 saw that and displayed the "Adding New Hardware" box. The next dialog that came up requested the driver disk. A little bit later the blue screen came up and they immediately shut off the display.

    So, to correct that "problem," Mr. Gates said they had worked on the problem the entire year since, and would give it another try, with the same computer and scanner. They plugged it in and it went beautifully, with one minor difference. The computer didn't prompt for the driver disk. My guess, with what I've known of Windows 9x over these years, is they already had the driver installed once before on that computer. So when it came around this time, it did not need the drivers because Windows already knew they were already there...

    I did see some write ups on those Windows type magazine web sites about this event, how it couldn't have gone smoother and Windows will save the universe from certain damnation. Not a word was mentioned of the difference in the presentation. I guess I was the only one there that paid attention to such details (the Windows tech support life did it to me ;)).

    Seeing him in person and watching this incident revealed to me much about Mr. Gates that is never covered in trade magazines/web sites. It can also be surmised by many accounts from people that knew him in the early days that he is a very competitive guy and hates to be embarrassed like that, PBS' Triumph of the Nerds (or whatever it was called), TNT's Pirates of Silicon Valley. I saw this side of him there at the keynote in Chicago. He will never let that happen to him again.

    Where's this going? Well, I wouldn't be surprised if what was shown for the audiece to "ooo" and "ahh," and give these lapdog magazines something to proclaim as a true renevation for the new millennium we just entered, and what was really going on behind the scenes were not exactly the same. A screen saver on these desktops that reloads the page every second, some graph app that shows the "load" on the servers? I wouldn't put it past the man. Things must run completely smoothely, no matter what...
  • grytpype wrote:

    "No matter what the header says, in substance, Slashdot is a discussion forum," and a few lines later, "I'm saying what matters is not what is newsworthy, but what is discussion worthy, which development kernel releases are (IMHO)."


    Agreed. There is a lot more going on in the world than could possibly be discussed on slashdot, even when you narrow your field with parameters like "must appeal to nerds" and "ought probably apply to technology and its effects." There are sites which better cater to this need. Slashdot tends to list 15-20 stories on the main page each day rather than, say, 115 to 120 stories. :)

    Not everyone would agree on what the most important, discussion-worthy, news-worthy events or ideas of the day are -- that much is made brutally clear by slashdot comments, eh? But let's say that Rob and the other authors consider their own interests as well as those of readers and the collaboration generally results in the posting of stories which interest a pretty broad swath of the readership. That, and check-boxes make it hard to complain too much about the news that doesn't fit any particular denomination of nerd-dom.

    But 2.4 promises to be a big step, and the steps that lead to dot-4 are interesting. Devfs alone seems to justify the news of the recent kernel changes.

    And Win2000? Well, did you see much coverage on the local TV news? Do you know a lot of IT guys who are anxious to switch a middle-size (say, 50-person) business to it prior to the first service pack? It's news, but only in a pretty pre-digested, press-release way. A new Linux kernel is more newsworthy (imho) than win2k in part because the date of release is an MS marketing tool more than it is "news" ... and yes, OEMs have had for a while, and beta testers both pro- and con- have had a while to play with it and reported their conclusions both here and elsewhere.

    just thoughts,

    timothy
  • Frankly I agree that posting kernel updates may be not so useful. Specially if they are Beta. Specially THIS one.

    I don't consider that Slashdot should not post such news. However I consider that they may have a use only in cases when we deal with stable releases. We don't have them every day and not everybody reads freshmeat every hour.

    In cases such as development kernels, I would highly recomend to restrict news to moments when important changes are made. And, in any case, to check up the stuff before publishing.

    This guy here, 2.3.46, has some serious bugs in it. For the hardware I use, it is not a runday kernel to be trusted. I had to make a few patches and sit until 5 in the morning to see it well and alive in my Ragnarok Linux Box. It is a beauty, it is powerful, it runs fast, it eats less memory, it already covers all hardware I have, it holds my dual-proc with a boost, it does not crash like old good 2.2. But damn, I took three hours checking up the source to make it run and, from time to time, it shows some weirdnesses. For a person with a middle knowledge of systems and programming this kernel may do and may go. But for the majority it may be a serious delusion.

    I was pretty amazed to see the damn thing this morning at slashdot. When I know that 80% of the people may not be able to even compile it normally... Frankly this starts to remind the hypes of Redmond's Mag00. People, we are not Mazdiers, and I think no one want to be such.
  • I was a little skeptical about the utility of devfs, but after reading through this discussion of it's virtues, I guess I'm sold on it: http://www.atnf.csiro.au/~rgo och/linux/docs/devfs.txt [csiro.au]

    In particular, I was impressed by this argument:

    Having your device nodes on the root filesystem means that you can't operate properly with a read-only root filesystem. This is because you want to change ownerships and protections of tty devices. Existing practice prevents you using a CD-ROM as your root filesystem for a *real* system. Sure, you can boot off a CD-ROM, but you can't change tty ownerships, so it's only good for installing.
    It strikes me that a completly read-only boot system would be a nice anti-cracker trick. But there are a lot of different points here... read it to see if something clicks with you.

    In general, using devfs sounds like it does some things a bit more cleanly than the traditional /dev, but it does a *lot* of things more cleanly. Sounds like a win.

  • I am sick of seeing linux development kernel upgrades posted on Slashdot.

    I'm not. Would you kindly not try to impose your opinion on me?

    Kernel releases are the most exciting thing about Linux, and if you don't understand that, you just don't get it. Don't read them if you don't want to. Just don't try to keep me from reading them.
  • the 2GB limit is a bit of a pain for advocates of Lotus Domino on Linux. Every other Domino platform can have databases with no practical size limit and many enterprise scale knowledge management applications with multimedia content make use of the large database sizes. Domino on Linux is as stable and powerful as the AS/400 and Unix platforms (no daily or weekly reboots as NT requires) but if it can't support substantial databases you can't always perform a trivial migration process from a legacy operating system.
  • And I think Microsoft's Windows 2000 beta News groups were run on a dual P100.
  • I disagree, Windows 2000 is a HUGE advancement over NT4. It is by no means an incremental change. Lets see, there's the Active Directory, there's Terminal Servers included, unlimited clustering, up to 32 processors per machine (well 8 in the current release), quotas(about time), QoS, Power Management, DirectX7 ...actually I could go on and on listing new features but I'd get flamed for being a Microsoft marketing guy. Well, anyway, I disagree :P Windows 2000 is a HUGE improvement and deserves to be recognised than just an incremental upgrade.
  • Um, I think that second demonstration was done on a Windows 2000 Beta machine. Windows 2000 'knows' of many devices thru the inf files, and it doesn't prompt you for drivers it knows i has. It just searches it's driver database and if it finds something, it installs it.
  • Weeeeellll... since VA and Andover are both publicly held, their administration can't do anything that doesn't benefit the stockholders financially. If they *do* do that, they'll get their pants sued off, and lose their jobs.

    Not true. Corporations do things which are adverse to profit all the time. Sometimes it's more accurately a case of putting off short-term profits for a long-term gain, or sacrificing long-term profits for a short-term gain. Or sometimes it's just because they think it's the right thing to do -- kind of like how some European automakers do not enforce patents on safety mechanisms, because they feel that safety is more important than profit.

    Microsoft has long been a significant contributor to the Free Software Foundation. I don't know about you, but I'd consider that adverse to Microsoft's self-interest. (The donations come through the United Way campaign. Microsoft has a pledge to match employee donations to United Way charities, and the FSF is a UW-approved charity.)

    Id Software has GPLed Quake; Bungie Software has GPLed Marathon 2: Durandal. While these software products are getting long in the tooth there was still a market for them. These two prestigious gaming companies intentionally forfeited profit, because ... well, you'll have to ask Id and Bungie.

    Corporations do things adverse to their own financial interest all the time. Claiming otherwise shows a lack of historical knowledge.

    I don't have any personal opinion on the merger, except that VA now owns both sourceforge and server51, so the only non-VA free development platform-type-site that I know of is openprojects.net. But that's a different point altogether.

    RHAD doesn't count as a free software development site? What about all the websites devoted to kernel hacking? Whatever happened to email?

    The Linux community survived just fine before Sourceforge or Server 51. I've made significant contributions to free software projects and I've never even visited either of those two sites.

    If Sourceforge and Server 51 were essential to the development of free software, then yes, I'd be irked about one company owning both. But they're not essential, so why worry?

    Oh, and themes.org and Slashdot are now owned by the same people...isn't that exciting? But, of course, the Andover/VA Linux staff has no say in what gets posted and what doesn't.

    The first rule of journalism is don't allege something unless you've got evidence to support the allegation.

    Gleam, it's been alleged that you're a monkey-eating child pornographer who had a homoerotic relationship with President Bill. But, of course, that's just speculation.

    Moral of the story: if you're going to allege that the Andover/VA staff has undue editorial influence, then for Pete's sake, show some evidence to back up your allegations.

    And...what about the guy who posted to the original merger (VA-Andover) thread, from valinux.com, who got an automatic +4(!), without any moderation. Hmm.

    I'm a certifiable Karma Whore; when I make posts they start out automagically at 2. This is kind of a cool thing. And y'know what? The other day, I saw one of my posts had a score of 1, with no moderation attached to it! My God! My evil nemesis must be out there, maliciously dropping my scores without moderation!

    ... or it could just be a bug.

    Never attribute to malice what can easily be explained by random chance.

    Am I concerned about VA/Andover and potential risks to Slashdot's editorial integrity? Yes, I am, and because of it I'm going to be watching the site closely. If I ever find real evidence of editorial malfeasance, then I'll take my marbles and play elsewhere.

    This is, incidentally, exactly what Taco, Hemos and everybody else on staff wants. They want the users to keep them honest. As long as we keep our eyes open, Slashdot will keep its editorial integrity. Then Slashdot gets what it wants (our viewership) and we get what we want (News for Nerds, Stuff that Matters).

    But there is a significant difference between keeping our eyes open for editorial abuse, and a paranoid belief that the few minor things we're seeing are just the tip of an iceberg of evil.

    Rant done.
  • Did I miss something? Did Microsoft cancel the W2K launch at the last moment? Did the spooks cover it up? No, everybody else is reporting the news except /.

    Huh? Microshaft? Who are they? W2K? Is that anything like Y2K? I'm confused!

    :)

    *ducking*

  • ..the news that another minor development release of Linux is out shouldn't really be on SlashDot, IMHO. However this particular release, I understand incorporates some major new functionality (devFS filing system), and *THAT* should have been reported.

    I think that SlashDot should confine its comments to releases on the main branch, or be a little more careful about how it describes the significance of development branch releases. Freshmeat is also an Andover site, and most of us have the Freshmeat window on our Slashdot displays too, so there is little need for this double reporting.

    On the other hand, SlashDot itself has been remarkably quiet about Windows 2000, and that should have qualified for a news story or two. Regardless of what you feel about Windows, this release is a MAJOR release of the current #1 operating system in the world and should have been covered as such. It is unfortunate that Slashdot did not feel it could put up a news article with some appropriate links describing new features, and even better if someone has a copy, a fast initial impression.

    Normally I`m on the side of Linux in most things [Linux leads 5:3 against Windows installs in my home], but I feel SlashDot should have at least tried to give us some objective reporting.

  • Open Source. Open Minds.

    Closed Source. Microsoft Astroturfers. Sour Grapes.

    This is an open source site, with most of us (astroturfers like you aside) far more interested in the most trivial and uninteresting patch to the ever changing development kernel of Linux, or patches to FreeBSD current, than we are in the overhyped release of a bloated, unstable, closed OS from an organization dedicated to denying all of us the freedom to chose our own platform on our own terms.

    Get over it and leave the rest of us in the open source community alone to continue building the future. If the content of this site offends you so, go back to microsoft.com and hang out with your buddies there. I'm sure they will be more than willing to wallow with you in your disillusion and self-pity while the rest of us, and the future itself, leaves you weeping in its wake.
  • Yes, several fixes have been posted in this very thread.

    Read here [slashdot.org]

    That said, I'm not sure why this kernel even made it to release. I know the kernel developers are busy, but they must be using some whacked settings if they could get that kernel to compile out of the box.

    Also, it seems 2.3.47pre3 is already up.
    http://www.kernel.org [kernel.org]
  • This totally screws up^H^H^H^H^H^H^H^H^Hchanges the way networking works on Linux. Now, networking will be cleanly multi-threaded, from start to finish. With any luck. :)

    This is a -major- change to Linux networking, and from the sounds of it, it's going to break a lot.

    As for devfs, I've played with it, and like it, but it does totally mess up my fstab file. Everything sat nicely on one console screen, and now I'll have line-wraps. Ugly!

  • Not like we would have already seen it on Freshmeat...
    And since when is another development kernel release NEWS FOR NERDS/STUFF THAT MATTERS?
    Anyone who gives a hoot about new linux kernels will be checking kernel.org for their new kernels..
    Give me a break.

    Unless I'm missing something.. is there something great about this new kernel? Tons of new long-awaited features? (I mean, compared to the one we had yesterday.... )

    No.. I didn't think so.

    Sheesh.

  • Those are DEVELOPMENT RELEASES for DEVELOPERS.
    That means PEOPLE WHO AREN'T AFRAID TO FIX BUGS.
    If there is a fsck up in the build scripts.. FIX IT!

    EVERYBODY KNOWS that 2.3 kernels are DEVELOPMENTAL, UNSABLE, UNRELIABLE kernels.
  • Here we are in a code freeze preparing for a 2.4 release and Linus includes a "small" change to the networking structure -- a structure widely deployed, tested, and proven to be functional -- that breaks every network driver in existance and introduces highly untested and possibly unstable code into the tree. I'm all for progress, but now is not the time to be introducing things that break everything. I trust I don't need to quote Linus' on words to this effect.

    A word to the one(s) breaking things... you deleted three elements from the network device structure; how hard is it to grep the other drivers and fix what the #^#!%#^ you just broke? Answer: TRIVIAL. I did it in an hour -- and I had to figure out what your were up to.

    And while I'm ranting... the last update to the raid code (arounf 2.3.43?) failed to include a file... 'xor.c' so now software raid is broken. Of course, the documentation is four (4) YEARS out-of-date so finding the development patches is out of the question -- in fact, the raidtools on kernel.org haven't been updated in months (Aug 24 1999) YES, programmers are the worst people ever to document anything, but it only takes a few seconds to update a URL in a header.
  • On the other hand, SlashDot itself has been remarkably quiet about Windows 2000, and that should have qualified for a news story or two. Regardless of what you feel about Windows, this release is a MAJOR release of the current #1 operating system in the world and should have been covered as such. It is unfortunate that Slashdot did not feel it could put up a news article with some appropriate links describing new features, and even better if someone has a copy, a fast initial impression.

    You know, I can't honestly decide if it's complete obliviousness, or a joke of Kaufmanesque stature... After all, it's only when you see that it's a joke that you can appreciate it.

    Think: nice juxtaposition: Linux Kernel release 2.3.46 (minor rev increase) vs. Win2K. Interesting anti-hype statement. Nice contrast.

    Though I still reckon it's somewhat oblivious and ostrich-like not to cover it in at least some way. Maybe they'll do so today.

    Si
  • It seems to me that most posts are moderated up that go against the sort of slashdot mindset or go against what most commenters are saying.

    Chris Hagar
  • It seems to me that announcing Windows 2000 on Slashdot would create no new informed or insightful discussion. I think that discussion is what Slashdot is all about. However, it does not really matter. Even if slashdot were merely a news site, the release date of Windows 2000 has already been posted here. I would wager that most people already know that Windows 2000 is released.

    Also, it might only serve to be a story with a bunch of flames wasting moderators' time, rather than a useful and insightful discussion.

    Chris Hagar

  • My guess, with what I've known of Windows 9x over these years, is they already had the driver installed once before on that computer. So when it came around this time, it did not need the drivers because Windows already knew they were already there...
    In fairness, that scanner was by then well over a year old, and Windows98SE was released in the interim - it is possible it has been added to the standard list of supported devices (given it embarrassed MS before, I suspect they would make DAMNED SURE that there was a working, MS-written driver for it) and if the CABs for those were on the HD, it wouldn't prompt for a floppy.
    But I agree - it is probable that the exact hardware setup was tested by adding and removing the scanner several times, and then BG refused to have Windoze reinstalled on the offchance it still didn't work on the night :+)
    --

Slashdot Top Deals

Somebody ought to cross ball point pens with coat hangers so that the pens will multiply instead of disappear.

Close