www.quake3arena.com is running WebSitePro/2.3.15 on NT4 or Windows 98
I'll have to take their word for it. Can't actually connect, myself... Fortunately I have other Quake geek friends who were up in the middle of the night to download local copies.
Simple: you didn't have a route set up to your DNS server (or possibly had a route set up that was broken somehow; I forget which causes DNS lookups to hang instead of just break), so "netstat -r" blocked waiting to do a reverse dns lookup on something like your gateway. On Linux use "netstat -rn" to avoid reverse lookups; it's probably the same for solaris.
Their claim that it's a uClinux-based kernel, and especially their claim that it's binary compatible with PalmOS apps, seems to indicate that they've made kernel modifications. They've got a link (currently quite slashdotted) to a ROM image including said kernel. So they're distributing a modified GPL'ed program: where's the source code?
Ok, I haven't read most of the comments yet, but I am worried that I'll see a number of the "Communism is fundamentally evil!" or "Linux isn't Communist! Don't say that!" comments that might be expected from the heavily libertarian Slashdotter demographic.
I'd just like to point out that one of the ways I get a kick out of Linux is considering this little paradox: Linux development is communist, libertarian, and successful. It's rare enough that you see two of adjectives applied to the same concept, much less all three.
Think about it:
Linux *is* a communist-developed OS, in the Marxist sense of the word, "from each according to his ability, to each according to his means". Every Linux developer who can improve the operating system in some way does so, not in proportion to how much he's getting paid to do it or because he's being ordered by the government to do it, but because he has the ability to do it. Every Linux user who needs features that the operating system and related software can provide gets those features, not in proportion to how much they've paid or because they've been doled out some limited feature set by a bureaucracy, but because they can freely download whatever they need.
Linux is a libertarian OS, too. The development may be communist, but not Stalinist communist - the top developers like Linus and Alan are followed not because they wield any political or economic power to enforce what they say, but because they've proved themselves extraordinarily capable in the past, and so people voluntarily listen to them. You have the freedom to choose your software from a number of competing vendors, to extend and modify it yourself, or to apply other people's modifications whether or not they have official approval. What few restrictions there are come from voluntary software licenses decided by the software authors.
It's kind of cool, when you think about it. In a system where the economics of scarcity are non-existant (the marginal cost of copying software is trivial), communism actually seems to work, and works without using force or coercion on anyone who takes part in it. At a time when most totalitarian communist countries are spectacular failures, it's kind of cool to see a voluntarily communistic system work.
Who knows, maybe when nanotech is cheap and the production of a material item is a matter of feeding enough matter and electricity into your properly programmed Seed, open source economics might play a big factor in the physical economy too.
People here are saying that yes, even NT has the ability to dump kernel core when it BSODs, but:
What exactly are you supposed to do with a kernel core dump under a closed source OS? Throw a printout of it into a bonfire to propitiate the Windows Demons? Send it to Microsoft and wait for their rigorous QA process to leap into action and send you a fixed kernel? I can't imagine trying to debug it yourself without being able to get a backtrace and look at the problem source code. Does Microsoft even leave a symbol table of internal function names in the NT kernel? What exactly do you do with a Kernel Debugger in Solaris if you can't see anything more than what a disassembler will tell you about the kernel being debugged?
Yeah, yeah, I know: you're not a real old school hacker unless you were working on Unix before Unix bacame popular. Good for you.
I've only used Solaris, AIX, HP-UX, and Linux (and only written code for the last two), but frankly I'd much rather be using Linux than any of the above.
I've heard Unix admins gripe about non-standard Linux-isms, but is a BSD/SysV syntax mix really that much of a problem compared to compilers that default to K&R instead of ANSI C, and system calls that were created solely to make your Unix source code incompatible and lock in software developers?
And as for the quality of said systems... well, Solaris is pretty sweet, but AIX and HP-UX are dead as soon as someone finds a hole deep enough to bury the remains. I remember telling people how aggravating it was that you couldn't upgrade a shared library or program in Windows while it was in use... only to discover that HP-UX 10.20 had the same limitation, not to mention further bass-ackwards problems with dynamic linking.
And don't get me started on CDE. The idea that multiple massive software vendors collaborated to produce the best desktop environment they could, and came up with *CDE* anyway, boggles the mind.
Don't get me wrong, there's lots of things that commercial Unices have (64-way SMP, high availability clustering, more optimized compilers, etc.) that Linux has yet to catch up to. But just because they're too expensive for the unwashed masses doesn't mean they're always worth it.
Oh, no, they used a weak, 40-bit encryption scheme with 200 different keys lying around, and it bit them in the ass, you say? I'm sorry, but if out of 200 different companies there wasn't one who would say "Hey, look, this encryption system is as solid as swiss cheese!" before creating the standard, then they're responsible for what's coming to them.
It's as if someone discovered that every door lock and ignition lock on General Motors' cars could be disabled with a refrigerator magnet. Too bad for GM.
Do you have any idea what the staggering support cost will be of being on the upgrade treadmill from Linux?
Yes, but "staggering" isn't the right word.
Can you really afford to hire staffers whose whole job will be to scan Usenet posts and mailing lists to apply the latest security patches?
No, but having one sysadmin whose job includes checking his email every so often for Red Hat security updates, that's OK. And since he can remotely apply security patches to thousands of machines at once with one command, and can do so without rebooting any of them, the costs would be vastly lower than when a new MS Hotfix or Service Pack comes out.
Can you afford the downtime to apply the weekly kernel upgrade?
No, which is why it's good that upgrading the kernel weekly isn't necessary. Every six months or so should do it. Oh, yeah, and that "downtime" will be less than five minutes. Not a problem for your workstations, and you're already doing loadbalancing/failover on your critical servers in case of hardware failure, right? I've had Linux crash due to one kernel bug in the past three years, but I've had network & SCSI cards (and a hard drive) die on me and need replacement at about a part per year.
Do you still remember when they discovered that there was a millisecond timer in Windows 9x that wrapped after 49.5 days, crashing the machine hard? Do you remember how amusing it was that it took them 4 years to discover this, because nobody expected a Windows machine to stay up for a month anyways?
Can you afford the support costs of handholding every user who needs to change something?
Needs to change *what*? Some specific details, rather than cloudy fearmongering, would be useful.
Something root-level? They shouldn't be changing it anyway. The support costs of ssh'ing in to do something as root are far outweighed by the maintenance costs of fixing the whole damn Mac or Win9x box when some luser deletes the wrong file, installs broken software overwriting system DLLs, or just does something stupid that the OS shouldn't have given him priviledges to do.
Something user-level? How much time do your employees spend playing with their window manger anyway, and why are they incapable of figuring out how to do it themselves?
When a power outage hits, can you afford the cost of recovering each desktop machine's fragile ext2fs, a decent percentage of which WILL be permanently corrupted by the sudden power outage?
Options: 1. Use a UPS. Duh. 2. Use a journaling filesystem. Journaling ext2fs is in beta now, with no killer bugs I've seen. In 6 months it'll be in the default kernel. 3. Trust ext2fs. I've seen a dozen machines survive dozens of power outages, inadvertantly hit reset buttons, pulled plugs, and similar gaffes. I lost files that hadn't been written to disk once (but even a journaling filesystem won't save buffered data), but I've only seen one partition that wasn't recovered by running fsck. Is NTFS really that much better? Nope. 4. Don't store important files on the local machine. You can seamlessly mount all your home directories over NFS or CODA, so why have anything in need of saving on your workstations at all? If a workstation dies, throw on a new disk image copy, change the hostname, and you're set.
These are meant as constructive questions for any IS organization seriously contemplating a major Linux rollout to consider.
These are questions with simple answers, obvious for anyone who has administered Linux (or any Unix, really) for more than a couple months. I should hope anyone contemplating a major Linux rollout has given it more in depth thought than you have.
Clue: they've already started thinking about this stuff.
Clue: They're not omniscient. People go with what they're comfortable with, not what makes the most sense. We've got PC kiosks in the Rice University library with a $200 NT license a piece, to *run a continuous telnet session in a window*. Why not use Win95? Because fixing it when someone maliciously or inadvertently broke it is too messy. Why not use Linux? Because "nobody gets fired for buying Microsoft". People are afraid of change, and will waste money on ludicrous decisions to avoid it.
Translation: "I don't have a lot of money, don't expect to ever have a lot of money, and feel a deep resentment at people who do, and who thus expose my deep feelings of avarice and inadequacy."
How do you suppose that?
Because I defined "a lot of money" as "the amount of money past which debrain thinks money is immoral". Should have made that more clear, but I figured you'd get it from context. So do you have (or ever expect to have) "a lot of money" by that definition? If so, it would be hypocrisy, and if not, then the first two clauses above were right.
The "deep resentment" was my interpretation of your generalizing statements such as "perpetual greed", "exploited", and "money eater". Maybe "deep" was a stretched guess, but you do seem to have a bone to pick with people who do nothing but make movies and offer (not force, mind you, just offer) to let you see them for a small price.
"Avarice" was a reference to the penny-pinching, pathetic greed it takes to publicly, proudly claim apathy toward producers risking million dollar losses, while you whine about losing a couple dollars because you didn't read the right reviews before going to see a bad movie.
My reasons for defending the rights of individuals is that I've worked with the Red Cross and Amnesty International, refugees of Yugoslavia and Kosovo and Ukraine and Nigeria
You're comparing Hollywood producers to genocidal tyrants and armed warlords, and you're calling me an asshole? Do you really not see the difference between putting a gun to someones head so you can take their livelihood vs. putting a commercial on the air to promote your movie? One action shreds the rights of individuals, one action doesn't impinge upon anyone's rights.
leading me to believe that you are a genuine asshole
Well, it's not my day job, but whatever, it's no worse than a couple of the assumptions I made about you. I take back "inadequacy", to start. I'll apologize for "loser" too; the Red Cross & Amnesty International service more than makes up for a couple muddy socialist inclinations.
But honestly, look at yourself - you really started out defending individuals' rights to freedom from totalitarian governments, and ended up defending individuals' rights to free copies of blockbuster movies? That's kind of a sad progression. At least most socialist activists talk about collectivist farming or factories first, and worry about how to entertain couch potatoes later.
I mean that, literally. I notice someone already mentioned Atlas Shrugged below; I thought a lot of things in that book stretched plausibility, but the characterization of people like you is vicious and pretty stabbingly accurate.
At some point of earning money, I conclude, the possession of more money is fundamentally wrong.
Translation: "I don't have a lot of money, don't expect to ever have a lot of money, and feel a deep resentment at people who do, and who thus expose my deep feelings of avarice and inadequacy."
Do you realize just how pathetic it is, whining that you have to pay a few bucks to rent a hundred million dollar movie? It's like listening to my cat whine that the can opener is opening his food too slowly, when both the can and the opener are completely beyond him. No, it's worse that: my cat can catch birds and could feed himself if necessary, whereas your plaintive message makes it clear that you lack the intelligence to keep yourself intertained without Hollywood-manufactured stimuli for as much as a year.
Guess what? If you don't think it's "fair" to pay $7 to sit in a million dollar theater and watch a ten million dollar plane crash, then go to the library and check out a freaking book. There are 6 billion people in the world, and if you think the thousands who are making movies for you are being "unfair" then you don't have to buy their movies, you don't have to see their movies, you don't even have to acknowledge they exist.
And lest you fail to grasp the broader applications of said theory:
If you don't like the salaries that overpaid, steroid-enhanced athletes get (as I don't), then don't go to a stadium, and they won't get a penny of your money. Hell, they might even be *subsidizing* TV shows you get to watch on your couch for free. Lucky us.
If you think Bill Gates got rich by taking money from a bunch of suckers (yeah, me too), then don't warez Microsoft Office (that would be admitting you were wrong, if you ever thought about it honestly), simply don't buy MS software and don't let him take your money.
From what I've gathered, the original questionable "for Dummies" use wasn't a "chat room subtitle" at all, but the subject line on an archived mailing list. How stupid does a lawyer have to be not to tell the difference? Or is the phrase "for Dummies" really illegal to use in private email?
What exactly does IDG expect people to do? Not use the phrase "for dummies" without footnoting it in private email? Hunt down every recipient of that email to make sure none are making a public archive of it? Or since they're going after the archiver, do they expect him to run every mail message through a million-entry trademark pattern search before converting it to HTML?
Ironically, though the title of this Slashdot article could be considered an infringement, and isn't an automatic archive of a random email, but was a deliberate word choice by an employee of Andover, a for-profit corporation. If IDG's lawyers don't drop out of hypersensitive mode soon, I know whom they're writing to next...
It's really just a plot to find out which of us actually use linux. You don't know what they'll do with this information. When a fascist totalitarian regime meant to oppress our resistance (read: Microsoft Fascism v2.0) comes about, will those who have registered be forced to wear penguin armbands and relocate to getto camps with only 9600 baud access?
You know, the funniest thing about that post is the implicit assumption that Microsoft Fascism 1.0 will be too buggy to work...
First you have the Palm. Then you add a modem. Then a bunch more RAM. Then arbitrary peripherals. Then color. Now a hard drive. Each of these adds complexity, size and cost--all of which are anti-thetical to the purpose of the original Palm.
Well, if by "the original Palm" you mean "those little palm-sized yellow post-it notes", then sure.
If you actually think that the original palm computers wouldn't have had color screens, hard drives, peripheral expansion, and modems had that been practical at the time, on the other hand, you're nuts. How is a tiny hard drive more complicated to the user than battery backed memory? It isn't, it's just more spacious. How does a color screen make things more complicated? Again, it doesn't. I guess you have to figure out that "complicated Internet thingie" to use a modem with your Palm, but that's a tradeoff I'm willing to make. And arbitrary peripherals, well, that involves the complication of buying the peripheral, plugging in the peripheral, installing software for the peripheral... but if you want to listen to your MP3 playlists on the road, it sure beats whistling.
I heard the title alone (mentioned very favorably) years ago, got a mental picture of some antarctic rescue story, was completely disinterested, and forgot about it.
Last summer, one of my most hackerish friends at work told me it was one of the best books he'd ever read. I picked it up from there, as did my girlfriend, three or four friends at school, my dad... It's the Stephenson nam-shub. I'm through Diamond Age & Zodiac too now, and waiting for Cryptonomicon to hit paperback anxiously...
Sure, it's still in alpha testing with Linux, but HP-UX, just about any commercial Unix out there has ACLs and has had them for some time. If you have 35,000 users, you probably aren't using Linux. If you have 35,000 users and you're smart, you aren't using NT either.
My good old Viper550 (TNT) card seems to work with Mesa-3.0 and NVidia's "glx 1.0" driver, but I'm under the impression this is only a "partial" set of drivers.
It's a complete OpenGL driver AFAIK, but it doesn't do direct rendering (it goes over the X pipe), and it's not nearly as optimized as the Windows drivers yet.
(Myth II supposedly only works with 3dfx brand boards because they're the only ones that have a 'complete' set of OpenGL drivers - though I've been trying to find out if that's really true or not...)
It's not. Myth II uses Glide. 3Dfx is actually unique in being one of the last people to *not* have a complete set of OpenGL drivers; that's why they have their "MiniGL" to run Quake* games.
Has Nvidia said or done anything on the Linux front since the initial release of their drivers?
There was an interview where an Nvidia rep said they'd have GeForce Linux drivers (but X server? Mesa drivers? Who knows?) when the card shipped, but I haven't heard anything since.
See Larry Snyder's work on ZPL at the University of Washington.
Thank you; I'll take a look at it. I'd prefer an OO language, but anything that automatically parallelizes code, that links to C code, and that isn't Fortran, is good to hear about.
The idea of embedding parallelism into the language is nothing new.
When did I say it was? I mentioned HPF. Well, not by name, but I at least said Fortran had parallizing keywords already, didn't I?
I don't get it. The ONLY difference between the two code fragments above was a keyword change.
Right. However, the two keywords would have significantly different effects, and would be used in different situations by the programmer.
Why would we have to have a new keyword when we can simply detect a parallel operation?
Because we *can't* simply detect a parallel operation, 90% of the time. If your code calls a function that isn't in the same object file (or is in a shared library), then there's no way to know whether the for loop is parallelizable. Unless, of course, the programmer tells you that the loop is parallelizable. Hence, the new keyword. If you tried to parallelize every for() loop in existing code, you would break most of it horribly.
There's also the fact that automatically parallelizing foreach() type code will have dramatically different performance effects on different systems; the threshhold size above which one would want incur the overhead of contacting separate threads to run code will vary depending on the CPU interconnections; a cluster would only be helpful for much larger loops than an SMP system. You could write every parallelizable loop with foreach() and hope the compiler will sort it out, but how is a compiler supposed to figure out that foreach(i=0;i<3;i++) { tinyfunction(i); }
shouldn't be split among different threads, while foreach(i=0;i<3;i++) { hugefunction(i); }
should be?
Why generate values of i if they aren't used? If the compiler can prove array[] is never used in your function, it'll just drop the loop entirely Can't get much more efficient than that.
Um, that wasn't at all what I was talking about. I was discussing cases where every value of array[] will be used, but where they could be calculated independently.
Functional languages, which can evaluate arguments to functions in any order (thus parallel), will often not bother to even run the loop until array[i] is needed in a fashion that can't be delayed. The best way to optimize code is to find ways to not run it at all.
Agreed.
I suggest thinking in languages other than C.
Disagreed. The point of SMP isn't to demonstrate some academic feature of Lisp, it is to make programs run faster. If you have a program that requires more than a single fast CPU to run, then you *definitely* don't want to run it in a non-compiled language. And if you have a Scheme compiler whose output will run as fast as C, C++, or Fortran code, I'd like to know about it.
That sounds antagonistic, but I'm absolutely serious; I'll probably be working next summer with fortran code on a 500-node system, and if you've got some means to let me write anything but Fortran, please take pity and let me know about it.
That all means unless you get a sadistic pleasure out of watching engineers shackled by fortran, it would be nice to see more parallel programming features available in C/C++.
It's happened to a great extent, anyway. Go hunt down the Toms Hardware review on the Athlon for a good understandable discussion on their architecture - the main reason they're getting up to 50% better FPU performance than Pentium IIIs is the fact that they've got 3 concurrent execution units there instead of 2.
But there's only so much of that you can do in hardware. What I'd like to see is multithreaded software produced *automatically* by C/C++ compilers when possible, the way high end Fortran compilers do for multinode supercomputers today. So instead of writing for(i=0;i<1000;i++) { array[i] = function(i); }
which generates i, executes the function of i, generates the next i, executes the function with that next i, etc; you might write foreach(i=0;i<1000;i++) { array[i] = function(i); }
which will generate as many values of i as you have CPUs, execute the function for each value of i on a different CPU, then generate the next set of i's, sending them to threads as necessary to keep every CPU busy.
We'd have to have a new keyword (I like foreach) for this, since the overhead involved would make it counterproductive in many circumstances and would break code (anything where function() isn't reentrant, or where the i's are assumed to be evaluated in order) in others.
Granted, I've said a lot of stupid things in the past, said a lot of them in public forums, and plan to continue doing so in the future... but anything I post publicly I intend to be read publically, damn it, and I hope to be able to either stand behind or apologize for any of it if it's ever brought up in the future.
Ironically, my problem is that there aren't *enough* archives to suit me. I wrote a lot in #politics, #religion, and a few other MajorNet/FidoNet forums back when I was 12/13, and I'd love to go rooting through it all now and see how my views/writing have changed. Unfortunately there's only a couple weeks of posts that didn't get deleted from my HDD, and I've never found any public archives. Anyone know of anything?
First, it's a real printing system, not something layered on top of LPR that doesn't let you set options.
Whew, glad I can finally stop printing everything with a fake printing system now.
What options does CUPS offer that you can't put in a postscript file or pass to lpr? Granted the feedback could be better than "the printer isn't working" when something goes wrong, but that's about all I'd like to improve.
Second, the RH filters limit the printer support to what is compiled into GhostScript; CUPS allows you to add new drivers without recompiling.
This is a technical plus, but not a huge practical plus. GS 5.10 takes up a little over a meg on my hard drive, and is updateable with one command.
Finally, file filters are a lot easier to deal with in CUPS - it will run multiple filters as needed to get to the "destination format", while the LPR filtering mechanism only runs a single filter.
This is factually incorrect. Take a look at the print filters system in Red Hat (or Debian, or anyone else) sometime. If there's a jpgtopnm filter and a pnmtops filter installed, you don't need a jpgtops filter to be able to print through ghostscript.
The only serious problem with Linux printing is the lack of ghostscript drivers for some printers... but since this CUPS only seems to have ps and pcl drivers, we'll need to use ghostscript for 90% of it's printer support for some time.
Granted, CUPS looks like an improvement, but not a huge one.
If you have pages on your site that you only want to be linked from other pages on your site, then you use CGI/mod_perl/apache modules/whatever to check the referer for a valid link source, and you return an error if you didn't get a good referrer.
If people make bots fake their referrer fields to get in anyway, then you've got a case for wire fraud. If not, then you can shut up and be happy.
Listening to companies whine that "Those people are accessing the data we made publically accessible! Make them stop!" is getting annoying.
Linux would be a big deal without Microsoft - because without Microsoft it would be competing with multi kilobuck Unix operating systems on the server side, it would be competing with MacOS on the desktop, and in all but high-end and specialized desktop app systems it would eat them both for lunch.
Until MacOS 8 I hadn't seen an OS from apple that wasn't as unstable as Win3.x, and they're still barely even with Win95 systems in my experience, UI and stabilitywise. Do they have full memory protection and preemptive multitasking in anything but OS X yet? The last amusing anecdote I heard claimed they were using "guard pages" in VM in OS 8 between kernel memory and app memory so that at least "decrement and write to pointer" infinite loops wouldn't trash the system, but they couldn't actually use the MMU to protect kernel-used memory without breaking things.
On the other end, I've watched $3000 Linux systems race past $15000 HPUX systems running the same software. This is after the HPUX systems have fallen in price due to NT Server influence. If the back office today was 100% Unix based (thus easy to port) software the middle end of it would be completely Linux today.
Slashdot Mirror
www.quake3arena.com is running WebSitePro/2.3.15 on NT4 or Windows 98
I'll have to take their word for it. Can't actually connect, myself... Fortunately I have other Quake geek friends who were up in the middle of the night to download local copies.
Simple: you didn't have a route set up to your DNS server (or possibly had a route set up that was broken somehow; I forget which causes DNS lookups to hang instead of just break), so "netstat -r" blocked waiting to do a reverse dns lookup on something like your gateway. On Linux use "netstat -rn" to avoid reverse lookups; it's probably the same for solaris.
Their claim that it's a uClinux-based kernel, and especially their claim that it's binary compatible with PalmOS apps, seems to indicate that they've made kernel modifications. They've got a link (currently quite slashdotted) to a ROM image including said kernel. So they're distributing a modified GPL'ed program: where's the source code?
from each according to his ability, to each according to his means
s/means/need/
I even used the preview button on that one, damn it...
Ok, I haven't read most of the comments yet, but I am worried that I'll see a number of the "Communism is fundamentally evil!" or "Linux isn't Communist! Don't say that!" comments that might be expected from the heavily libertarian Slashdotter demographic.
I'd just like to point out that one of the ways I get a kick out of Linux is considering this little paradox: Linux development is communist, libertarian, and successful. It's rare enough that you see two of adjectives applied to the same concept, much less all three.
Think about it:
Linux *is* a communist-developed OS, in the Marxist sense of the word, "from each according to his ability, to each according to his means". Every Linux developer who can improve the operating system in some way does so, not in proportion to how much he's getting paid to do it or because he's being ordered by the government to do it, but because he has the ability to do it. Every Linux user who needs features that the operating system and related software can provide gets those features, not in proportion to how much they've paid or because they've been doled out some limited feature set by a bureaucracy, but because they can freely download whatever they need.
Linux is a libertarian OS, too. The development may be communist, but not Stalinist communist - the top developers like Linus and Alan are followed not because they wield any political or economic power to enforce what they say, but because they've proved themselves extraordinarily capable in the past, and so people voluntarily listen to them. You have the freedom to choose your software from a number of competing vendors, to extend and modify it yourself, or to apply other people's modifications whether or not they have official approval. What few restrictions there are come from voluntary software licenses decided by the software authors.
It's kind of cool, when you think about it. In a system where the economics of scarcity are non-existant (the marginal cost of copying software is trivial), communism actually seems to work, and works without using force or coercion on anyone who takes part in it. At a time when most totalitarian communist countries are spectacular failures, it's kind of cool to see a voluntarily communistic system work.
Who knows, maybe when nanotech is cheap and the production of a material item is a matter of feeding enough matter and electricity into your properly programmed Seed, open source economics might play a big factor in the physical economy too.
People here are saying that yes, even NT has the ability to dump kernel core when it BSODs, but:
What exactly are you supposed to do with a kernel core dump under a closed source OS? Throw a printout of it into a bonfire to propitiate the Windows Demons? Send it to Microsoft and wait for their rigorous QA process to leap into action and send you a fixed kernel? I can't imagine trying to debug it yourself without being able to get a backtrace and look at the problem source code. Does Microsoft even leave a symbol table of internal function names in the NT kernel? What exactly do you do with a Kernel Debugger in Solaris if you can't see anything more than what a disassembler will tell you about the kernel being debugged?
Yeah, yeah, I know: you're not a real old school hacker unless you were working on Unix before Unix bacame popular. Good for you.
I've only used Solaris, AIX, HP-UX, and Linux (and only written code for the last two), but frankly I'd much rather be using Linux than any of the above.
I've heard Unix admins gripe about non-standard Linux-isms, but is a BSD/SysV syntax mix really that much of a problem compared to compilers that default to K&R instead of ANSI C, and system calls that were created solely to make your Unix source code incompatible and lock in software developers?
And as for the quality of said systems... well, Solaris is pretty sweet, but AIX and HP-UX are dead as soon as someone finds a hole deep enough to bury the remains. I remember telling people how aggravating it was that you couldn't upgrade a shared library or program in Windows while it was in use... only to discover that HP-UX 10.20 had the same limitation, not to mention further bass-ackwards problems with dynamic linking.
And don't get me started on CDE. The idea that multiple massive software vendors collaborated to produce the best desktop environment they could, and came up with *CDE* anyway, boggles the mind.
Don't get me wrong, there's lots of things that commercial Unices have (64-way SMP, high availability clustering, more optimized compilers, etc.) that Linux has yet to catch up to. But just because they're too expensive for the unwashed masses doesn't mean they're always worth it.
On Slashdot it's very common to see a hopeless, drooling idiot pretending to be an expert on something
The only thing more common is seeing an AC read an objective argument and respond with a flaming rant full of ad hominem attacks.
Oh, no, they used a weak, 40-bit encryption scheme with 200 different keys lying around, and it bit them in the ass, you say? I'm sorry, but if out of 200 different companies there wasn't one who would say "Hey, look, this encryption system is as solid as swiss cheese!" before creating the standard, then they're responsible for what's coming to them.
It's as if someone discovered that every door lock and ignition lock on General Motors' cars could be disabled with a refrigerator magnet. Too bad for GM.
Do you have any idea what the staggering support cost will be of being on the upgrade treadmill from Linux?
Yes, but "staggering" isn't the right word.
Can you really afford to hire staffers whose whole job will be to scan Usenet posts and mailing lists to apply the latest security patches?
No, but having one sysadmin whose job includes checking his email every so often for Red Hat security updates, that's OK. And since he can remotely apply security patches to thousands of machines at once with one command, and can do so without rebooting any of them, the costs would be vastly lower than when a new MS Hotfix or Service Pack comes out.
Can you afford the downtime to apply the weekly kernel upgrade?
No, which is why it's good that upgrading the kernel weekly isn't necessary. Every six months or so should do it. Oh, yeah, and that "downtime" will be less than five minutes. Not a problem for your workstations, and you're already doing loadbalancing/failover on your critical servers in case of hardware failure, right? I've had Linux crash due to one kernel bug in the past three years, but I've had network & SCSI cards (and a hard drive) die on me and need replacement at about a part per year.
Do you still remember when they discovered that there was a millisecond timer in Windows 9x that wrapped after 49.5 days, crashing the machine hard? Do you remember how amusing it was that it took them 4 years to discover this, because nobody expected a Windows machine to stay up for a month anyways?
Can you afford the support costs of handholding every user who needs to change something?
Needs to change *what*? Some specific details, rather than cloudy fearmongering, would be useful.
Something root-level? They shouldn't be changing it anyway. The support costs of ssh'ing in to do something as root are far outweighed by the maintenance costs of fixing the whole damn Mac or Win9x box when some luser deletes the wrong file, installs broken software overwriting system DLLs, or just does something stupid that the OS shouldn't have given him priviledges to do.
Something user-level? How much time do your employees spend playing with their window manger anyway, and why are they incapable of figuring out how to do it themselves?
When a power outage hits, can you afford the cost of recovering each desktop machine's fragile ext2fs, a decent percentage of which WILL be permanently corrupted by the sudden power outage?
Options:
1. Use a UPS. Duh.
2. Use a journaling filesystem. Journaling ext2fs is in beta now, with no killer bugs I've seen. In 6 months it'll be in the default kernel.
3. Trust ext2fs. I've seen a dozen machines survive dozens of power outages, inadvertantly hit reset buttons, pulled plugs, and similar gaffes. I lost files that hadn't been written to disk once (but even a journaling filesystem won't save buffered data), but I've only seen one partition that wasn't recovered by running fsck. Is NTFS really that much better? Nope.
4. Don't store important files on the local machine. You can seamlessly mount all your home directories over NFS or CODA, so why have anything in need of saving on your workstations at all? If a workstation dies, throw on a new disk image copy, change the hostname, and you're set.
These are meant as constructive questions for any IS organization seriously contemplating a major Linux rollout to consider.
These are questions with simple answers, obvious for anyone who has administered Linux (or any Unix, really) for more than a couple months. I should hope anyone contemplating a major Linux rollout has given it more in depth thought than you have.
Clue: they've already started thinking about this stuff.
Clue: They're not omniscient. People go with what they're comfortable with, not what makes the most sense. We've got PC kiosks in the Rice University library with a $200 NT license a piece, to *run a continuous telnet session in a window*. Why not use Win95? Because fixing it when someone maliciously or inadvertently broke it is too messy. Why not use Linux? Because "nobody gets fired for buying Microsoft". People are afraid of change, and will waste money on ludicrous decisions to avoid it.
How do you suppose that?
Because I defined "a lot of money" as "the amount of money past which debrain thinks money is immoral". Should have made that more clear, but I figured you'd get it from context. So do you have (or ever expect to have) "a lot of money" by that definition? If so, it would be hypocrisy, and if not, then the first two clauses above were right.
The "deep resentment" was my interpretation of your generalizing statements such as "perpetual greed", "exploited", and "money eater". Maybe "deep" was a stretched guess, but you do seem to have a bone to pick with people who do nothing but make movies and offer (not force, mind you, just offer) to let you see them for a small price.
"Avarice" was a reference to the penny-pinching, pathetic greed it takes to publicly, proudly claim apathy toward producers risking million dollar losses, while you whine about losing a couple dollars because you didn't read the right reviews before going to see a bad movie.
My reasons for defending the rights of individuals is that I've worked with the Red Cross and Amnesty International, refugees of Yugoslavia and Kosovo and Ukraine and Nigeria
You're comparing Hollywood producers to genocidal tyrants and armed warlords, and you're calling me an asshole? Do you really not see the difference between putting a gun to someones head so you can take their livelihood vs. putting a commercial on the air to promote your movie? One action shreds the rights of individuals, one action doesn't impinge upon anyone's rights.
leading me to believe that you are a genuine asshole
Well, it's not my day job, but whatever, it's no worse than a couple of the assumptions I made about you. I take back "inadequacy", to start. I'll apologize for "loser" too; the Red Cross & Amnesty International service more than makes up for a couple muddy socialist inclinations.
But honestly, look at yourself - you really started out defending individuals' rights to freedom from totalitarian governments, and ended up defending individuals' rights to free copies of blockbuster movies? That's kind of a sad progression. At least most socialist activists talk about collectivist farming or factories first, and worry about how to entertain couch potatoes later.
I mean that, literally. I notice someone already mentioned Atlas Shrugged below; I thought a lot of things in that book stretched plausibility, but the characterization of people like you is vicious and pretty stabbingly accurate.
At some point of earning money, I conclude, the possession of more money is fundamentally wrong.
Translation: "I don't have a lot of money, don't expect to ever have a lot of money, and feel a deep resentment at people who do, and who thus expose my deep feelings of avarice and inadequacy."
Do you realize just how pathetic it is, whining that you have to pay a few bucks to rent a hundred million dollar movie? It's like listening to my cat whine that the can opener is opening his food too slowly, when both the can and the opener are completely beyond him. No, it's worse that: my cat can catch birds and could feed himself if necessary, whereas your plaintive message makes it clear that you lack the intelligence to keep yourself intertained without Hollywood-manufactured stimuli for as much as a year.
Guess what? If you don't think it's "fair" to pay $7 to sit in a million dollar theater and watch a ten million dollar plane crash, then go to the library and check out a freaking book. There are 6 billion people in the world, and if you think the thousands who are making movies for you are being "unfair" then you don't have to buy their movies, you don't have to see their movies, you don't even have to acknowledge they exist.
And lest you fail to grasp the broader applications of said theory:
If you don't like the salaries that overpaid, steroid-enhanced athletes get (as I don't), then don't go to a stadium, and they won't get a penny of your money. Hell, they might even be *subsidizing* TV shows you get to watch on your couch for free. Lucky us.
If you think Bill Gates got rich by taking money from a bunch of suckers (yeah, me too), then don't warez Microsoft Office (that would be admitting you were wrong, if you ever thought about it honestly), simply don't buy MS software and don't let him take your money.
From what I've gathered, the original questionable "for Dummies" use wasn't a "chat room subtitle" at all, but the subject line on an archived mailing list. How stupid does a lawyer have to be not to tell the difference? Or is the phrase "for Dummies" really illegal to use in private email?
What exactly does IDG expect people to do? Not use the phrase "for dummies" without footnoting it in private email? Hunt down every recipient of that email to make sure none are making a public archive of it? Or since they're going after the archiver, do they expect him to run every mail message through a million-entry trademark pattern search before converting it to HTML?
Ironically, though the title of this Slashdot article could be considered an infringement, and isn't an automatic archive of a random email, but was a deliberate word choice by an employee of Andover, a for-profit corporation. If IDG's lawyers don't drop out of hypersensitive mode soon, I know whom they're writing to next...
It's really just a plot to find out which of us actually use linux. You don't know what they'll do with this information. When a fascist totalitarian regime meant to oppress our resistance (read: Microsoft Fascism v2.0) comes about, will those who have registered be forced to wear penguin armbands and relocate to getto camps with only 9600 baud access?
You know, the funniest thing about that post is the implicit assumption that Microsoft Fascism 1.0 will be too buggy to work...
First you have the Palm. Then you add a modem. Then a bunch more RAM. Then arbitrary peripherals. Then color. Now a hard drive. Each of these adds complexity, size and cost--all of which are anti-thetical to the purpose of the original Palm.
Well, if by "the original Palm" you mean "those little palm-sized yellow post-it notes", then sure.
If you actually think that the original palm computers wouldn't have had color screens, hard drives, peripheral expansion, and modems had that been practical at the time, on the other hand, you're nuts. How is a tiny hard drive more complicated to the user than battery backed memory? It isn't, it's just more spacious. How does a color screen make things more complicated? Again, it doesn't. I guess you have to figure out that "complicated Internet thingie" to use a modem with your Palm, but that's a tradeoff I'm willing to make. And arbitrary peripherals, well, that involves the complication of buying the peripheral, plugging in the peripheral, installing software for the peripheral... but if you want to listen to your MP3 playlists on the road, it sure beats whistling.
I heard the title alone (mentioned very favorably) years ago, got a mental picture of some antarctic rescue story, was completely disinterested, and forgot about it.
Last summer, one of my most hackerish friends at work told me it was one of the best books he'd ever read. I picked it up from there, as did my girlfriend, three or four friends at school, my dad... It's the Stephenson nam-shub. I'm through Diamond Age & Zodiac too now, and waiting for Cryptonomicon to hit paperback anxiously...
Sure, it's still in alpha testing with Linux, but HP-UX, just about any commercial Unix out there has ACLs and has had them for some time. If you have 35,000 users, you probably aren't using Linux. If you have 35,000 users and you're smart, you aren't using NT either.
I was wondering the same thing.
Me too.
My good old Viper550 (TNT) card seems to work with Mesa-3.0 and NVidia's "glx 1.0" driver, but I'm under the impression this is only a "partial" set of drivers.
It's a complete OpenGL driver AFAIK, but it doesn't do direct rendering (it goes over the X pipe), and it's not nearly as optimized as the Windows drivers yet.
(Myth II supposedly only works with 3dfx brand boards because they're the only ones that have a 'complete' set of OpenGL drivers - though I've been trying to find out if that's really true or not...)
It's not. Myth II uses Glide. 3Dfx is actually unique in being one of the last people to *not* have a complete set of OpenGL drivers; that's why they have their "MiniGL" to run Quake* games.
Has Nvidia said or done anything on the Linux front since the initial release of their drivers?
There was an interview where an Nvidia rep said they'd have GeForce Linux drivers (but X server? Mesa drivers? Who knows?) when the card shipped, but I haven't heard anything since.
See Larry Snyder's work on ZPL at the University of Washington.
Thank you; I'll take a look at it. I'd prefer an OO language, but anything that automatically parallelizes code, that links to C code, and that isn't Fortran, is good to hear about.
The idea of embedding parallelism into the language is nothing new.
When did I say it was? I mentioned HPF. Well, not by name, but I at least said Fortran had parallizing keywords already, didn't I?
I don't get it. The ONLY difference between the two code fragments above was a keyword change.
Right. However, the two keywords would have significantly different effects, and would be used in different situations by the programmer.
Why would we have to have a new keyword when we can simply detect a parallel operation?
Because we *can't* simply detect a parallel operation, 90% of the time. If your code calls a function that isn't in the same object file (or is in a shared library), then there's no way to know whether the for loop is parallelizable. Unless, of course, the programmer tells you that the loop is parallelizable. Hence, the new keyword. If you tried to parallelize every for() loop in existing code, you would break most of it horribly.
There's also the fact that automatically parallelizing foreach() type code will have dramatically different performance effects on different systems; the threshhold size above which one would want incur the overhead of contacting separate threads to run code will vary depending on the CPU interconnections; a cluster would only be helpful for much larger loops than an SMP system. You could write every parallelizable loop with foreach() and hope the compiler will sort it out, but how is a compiler supposed to figure out that
foreach(i=0;i<3;i++) {
tinyfunction(i);
}
shouldn't be split among different threads, while
foreach(i=0;i<3;i++) {
hugefunction(i);
}
should be?
Why generate values of i if they aren't used? If the compiler can prove array[] is never used in your function, it'll just drop the loop entirely Can't get much more efficient than that.
Um, that wasn't at all what I was talking about. I was discussing cases where every value of array[] will be used, but where they could be calculated independently.
Functional languages, which can evaluate arguments to functions in any order (thus parallel), will often not bother to even run the loop until array[i] is needed in a fashion that can't be delayed. The best way to optimize code is to find ways to not run it at all.
Agreed.
I suggest thinking in languages other than C.
Disagreed. The point of SMP isn't to demonstrate some academic feature of Lisp, it is to make programs run faster. If you have a program that requires more than a single fast CPU to run, then you *definitely* don't want to run it in a non-compiled language. And if you have a Scheme compiler whose output will run as fast as C, C++, or Fortran code, I'd like to know about it.
That sounds antagonistic, but I'm absolutely serious; I'll probably be working next summer with fortran code on a 500-node system, and if you've got some means to let me write anything but Fortran, please take pity and let me know about it.
That all means unless you get a sadistic pleasure out of watching engineers shackled by fortran, it would be nice to see more parallel programming features available in C/C++.
It's happened to a great extent, anyway. Go hunt down the Toms Hardware review on the Athlon for a good understandable discussion on their architecture - the main reason they're getting up to 50% better FPU performance than Pentium IIIs is the fact that they've got 3 concurrent execution units there instead of 2.
But there's only so much of that you can do in hardware. What I'd like to see is multithreaded software produced *automatically* by C/C++ compilers when possible, the way high end Fortran compilers do for multinode supercomputers today. So instead of writing
for(i=0;i<1000;i++) {
array[i] = function(i);
}
which generates i, executes the function of i, generates the next i, executes the function with that next i, etc; you might write
foreach(i=0;i<1000;i++) {
array[i] = function(i);
}
which will generate as many values of i as you have CPUs, execute the function for each value of i on a different CPU, then generate the next set of i's, sending them to threads as necessary to keep every CPU busy.
We'd have to have a new keyword (I like foreach) for this, since the overhead involved would make it counterproductive in many circumstances and would break code (anything where function() isn't reentrant, or where the i's are assumed to be evaluated in order) in others.
Granted, I've said a lot of stupid things in the past, said a lot of them in public forums, and plan to continue doing so in the future... but anything I post publicly I intend to be read publically, damn it, and I hope to be able to either stand behind or apologize for any of it if it's ever brought up in the future.
Ironically, my problem is that there aren't *enough* archives to suit me. I wrote a lot in #politics, #religion, and a few other MajorNet/FidoNet forums back when I was 12/13, and I'd love to go rooting through it all now and see how my views/writing have changed. Unfortunately there's only a couple weeks of posts that didn't get deleted from my HDD, and I've never found any public archives. Anyone know of anything?
First, it's a real printing system, not something layered on top of LPR that doesn't let you set options.
Whew, glad I can finally stop printing everything with a fake printing system now.
What options does CUPS offer that you can't put in a postscript file or pass to lpr? Granted the feedback could be better than "the printer isn't working" when something goes wrong, but that's about all I'd like to improve.
Second, the RH filters limit the printer support to what is compiled into GhostScript; CUPS allows you to add new drivers without recompiling.
This is a technical plus, but not a huge practical plus. GS 5.10 takes up a little over a meg on my hard drive, and is updateable with one command.
Finally, file filters are a lot easier to deal with in CUPS - it will run multiple filters as needed to get to the "destination format", while the LPR filtering mechanism only runs a single filter.
This is factually incorrect. Take a look at the print filters system in Red Hat (or Debian, or anyone else) sometime. If there's a jpgtopnm filter and a pnmtops filter installed, you don't need a jpgtops filter to be able to print through ghostscript.
The only serious problem with Linux printing is the lack of ghostscript drivers for some printers... but since this CUPS only seems to have ps and pcl drivers, we'll need to use ghostscript for 90% of it's printer support for some time.
Granted, CUPS looks like an improvement, but not a huge one.
If you have pages on your site that you only want to be linked from other pages on your site, then you use CGI/mod_perl/apache modules/whatever to check the referer for a valid link source, and you return an error if you didn't get a good referrer.
If people make bots fake their referrer fields to get in anyway, then you've got a case for wire fraud. If not, then you can shut up and be happy.
Listening to companies whine that "Those people are accessing the data we made publically accessible! Make them stop!" is getting annoying.
Linux wouldn't be such a big deal with out MSFT
Linux would be a big deal without Microsoft - because without Microsoft it would be competing with multi kilobuck Unix operating systems on the server side, it would be competing with MacOS on the desktop, and in all but high-end and specialized desktop app systems it would eat them both for lunch.
Until MacOS 8 I hadn't seen an OS from apple that wasn't as unstable as Win3.x, and they're still barely even with Win95 systems in my experience, UI and stabilitywise. Do they have full memory protection and preemptive multitasking in anything but OS X yet? The last amusing anecdote I heard claimed they were using "guard pages" in VM in OS 8 between kernel memory and app memory so that at least "decrement and write to pointer" infinite loops wouldn't trash the system, but they couldn't actually use the MMU to protect kernel-used memory without breaking things.
On the other end, I've watched $3000 Linux systems race past $15000 HPUX systems running the same software. This is after the HPUX systems have fallen in price due to NT Server influence. If the back office today was 100% Unix based (thus easy to port) software the middle end of it would be completely Linux today.