Not feasible. The kernel contains a huge quantity of functionality, and including an explanation on each thing (besides wasting the time of developers) would grow the changelog to an insane size. I don't know about you, but I already don't read through every bit of the changelog. I only worry about a new kernel when I need new functionality or a bugfix.
For the guy who is irritated that his netconsole has been malfunctioning...he's a happy guy.
For the typical user-that-rebuilds-his-kernel-for-the-sake-of-havi ng-the-latest, just use the kernel if it has new functionality that you want. Otherwise, just search for "corruption" in the changelog (the devs seem to be good about marking corruption-causing problems as corruption-causing) and if you have any of the affected hardware, upgrade.
It would take a *huge* amount of explanation to describe every modification and why it happens to the end user, on the level of having an ongoing OS class. Microsoft just doesn't talk about most of their bugfixes. With Linux you can get the detailed changelog, various summaries people have made, whatever advisories/errata the distro maintainers have put out, etc.
Err...you're quoting an eBay price. That's hardly sustainable or standard. I can occasionally obtain a nice piece of hardware for free, too, but it isn't reliable.
As for the entire concept of using tablet PCs -- tablets are generally *more* expensive than the equivalent laptop, which would make a much better base system.
This sort of thing occurs in steps - gradually. However, eventually, a lot of people may well wake up one morning and say "Holy shit! It's 1984!"
We have a "Department of Defense" used to invade other countries, an "Office of Homeland Security" that can detain and monitor people without being controlled by the judicial branch and ignores any rights that they once had, a figurehead leader that keeps exhorting us to greater national security efforts against an ephemeral, unbeatable foe, and monitoring of your movements (you carry a cell phone, right?) and your purchases.
And yet the amazing thing is that Canadians don't get attacked by terrorists, because they don't actively play dirty tricks with the people and governments of other nations.
TV subscriptions are one thing because most shows are transient, and you can record forever the ones you like. But a music subscription offers no similar benefits, only an ongoing cost and limitations on use (can you burn real CD's with a subscription service?)
It's a reasonable bet that Microsoft is not providing any guarantees that their rates will stay the same. In the past, they have *consistently* followed the forumula of doing whatever it takes to get into and dominating a market, then leveraging that market for all it's worth. If they get 80% of the US population using Janus, as people do TV, then they need only raise rates, as the ISPs did. Sure, you could cancel -- but then they have all the money you've given them up to that point, and you have nothing to show for your money.
You know, I get frusterated when people won't admit a problem, but really, that has nothing on Foale and his counterpart.
FOALE: "Uh, Houston? We've got something, uh, drumming on our outer hull. Sounds kind of like a sheet of metal bending on itself or something."
HOUSTON: "Okay, let's take a look...well...nope, everything's working fine."
FOALE: "Really, it's kind of loud. And this is the second time -- last time we didn't get to try and find out what was going on."
HOUSTON: (thinks "Darn astronauts, it's all in their heads") "Well, we're all green here. You could wait a couple months to take a look at it last time. You can probably wait to do anything on this until the new guys arrive.
FOALE: (thinks "Dammit, I'm sitting here in a little pocket of air a foot away from hard vacuum and something is *breaking*.") "Yeah, how about we do a spacewalk sooner."
HOUSTON: "I'll 'escalate' the problem and we'll see what we can do. Just sit tight up there."
Genocide was certainly not a goal of Hitler that was clear to the public four years into his rule (it was not until well into World War II that genocide became a goal), though he exhibited stronger anti-racial standpoints at that time than Bush currently is.
Exile was the proposed answer instead.
I have, on my refrigerator, a picture of a man in the Netherlands. The picture dates back about a month ago. He has sewn his own eyes and mouth shut with heavy thread. He is being held his tearful Dutch wife. He is one of thousands of Iraqis that is now being forced out of the country, to go back to Iraq, thanks to pressure from Bush. I would like to know what more people have to do to get their message across -- that they they are desperate and being treated horribly badly.
The really awful thing is that a major thing we used to think despicable about Soviet Russia and Nazi Germany was the identification papers and the restrictions on travel.
With computer databases, your image and your fingerprints *are* identification papers, and now you are being forced to hand them over at checkpoints.
Seriously, it was all very funny when we *started* to point out the amazing number of similarities between Hitler and Bush's rise. There was a terrorist act on a national monument (and even, in the 9/11 case, *attempted* on the national legislature, same as Germany) that produced national fear, whipped up by leader, used to convince legislature to pass through critical bills granting extensive police powers. Political opponents were accused of being soft on terrorism. Fear and xenophobia against religious (Islamic/Christian) and racial (Arabic/Jewish) groups was used to greatly infringe those people's rights and persecute them. A number of undesireable people, in violation of national law, were locked up in a camp to isolate them from the rest of society (Guantanamo Bay/Nazi concentration camps). Nationalistic fervor was whipped up and whipped up again to build up a popular base. Personal vendettas were made good upon with the new power (Bush-Hussein/Hitler-a number of enemies). Other countries were invaded and occupied on poor pretexts, banking on the fact that other, less powerful, countries would not be willing to organize or do more than protest (Iraq/several countries). A primary motivation for the invasion was resources (and later Nazi invasion into the USSR was significantly for oil). Business and government had close ties, and war profiteer corporations did a number of nasty things to take advantage of cronyism with major political figures (Schindler's List is a nice example). Right now the third largest employer of armed forces in Iraq (after the US and Britain)are private corporations -- big companies that are answerable only to an extremely friendly occupational government that grants Iraqis almost no rights and consists mostly of people trying to curry favor with their US occupiers to try to get a more advantageous political position in the future. Neither leader is brilliant, but both are prone to violence and grudge-holding. Both managed to seize control of the legislature at about the time they gained office. Neither has much regard for the lives of the people they have conquered -- we have been using unarmed Iraqi guards as inspectors of cars into restricted areas before US personnel come close, making human shields out of them. Neither feels that international opinion is of much import. Both quickly established powerful police organizations with far stronger powers than their predecessors, little oversight, and the ability to bypass much of the judicial system (OHS/Gestapo). Both started their invasions based on punishing the terrorists that attacked their nation, and immediately spread out once they had the power they needed. Both had rising unemployment in their countries, and a growing degree of xenophobia towards foreign laborers.
There are some differences. Hitler respected and even idolized what Britain had done -- Bush treats Britain as a lapdog. Hitler actively physically intimidated his physical opponents -- Bush does not. Hitler invaded, occupied, and eliminated the governments of no countries within his first four years as ruler, whereas Bush invaded, occupied, and eliminated the governments of two countries within his first four years as ruler. Hitler wound up eventually killing many more people than Bush has thus far, though Bush is currently ahead for the first four years of rule. Hitler did not actively attempt to control other countries through diplomatic means -- Bush has a team that works hard to control other contries without needing to overthrow their government. Bush has computers and telecommunication monitoring systems, but Hitler did not.
Screw Goodwin's Law. The man didn't write it in 2004.
The first is the Novell copyright situation. To me, it's not clear who's in the right here.
McBride: Would you buy an operating system without the source-code copyright? If you don't have copyright, they can turn around the next day and screw you.
Sontag: Instead, they waited nine years.
McBride: We have no doubts that our Unix copyright claims are valid.
One must, of course, ask why SCO felt that they had to wait years before notifying Linux folks of their alleged horrific infringements, and then felt that it was necessary to avoid actually *telling* Linux folks what the alleged infringements year until months and multiple court orders forced them to do so.
Sontag: We don't have to knock out the GPL for us to succeed on the copyright issue. The GPL itself supports, in a lot of ways, our positions. Section 0 of the GPL states that the legit copyright holder has to place a notice assigning the copyright over to the GPL.
All these contributions of our IP did not have an assignment by SCO saying here, 'We assign these copyrights to the GPL.' The fact that we participated with Linux does not mean that we inadvertently contributed our code to the GPL. You can't contribute inadvertently to Linux. We feel we have a very strong position based on the GPL.
First, this tidbit:
'We assign these copyrights to the GPL.'
Okay, enough fun has been made of Sontag and McBride's lack of competence when it comes to IP, so I'll avoid the jokes. You don't "assign a copyright to a license" (though GNU contributors are required to assign their copyright to the FSF for a number of reasons, in addition to licensing it under the GPL -- Linux is not a GNU project.)
Uh, huh. The fact that you added them to a file containing a GPL header doesn't count, eh? It's been well understood for many years that one header works for multiple contributions. When it comes to licensing, intent matters, and there was very clearly intent to GPL this code. I can't understand how you could make any kind of a counterargument.
The fact that we participated with Linux does not mean that we inadvertently contributed our code to the GPL.
Well, the alternative you have is that you committed massive infringement of thousands of IP holders that licensed their Linux code under the GPL. It's one or the other, SCO. If you want to go after Linux (and it's a damned weak argument -- I can't see how you'd manage to win it), you're also admitting that you deliberately committed a far worse crime. The potential costs of years of theft of perhaps millions of copies of Linux would easily bankrupt your company. I would expect that a shrewd mediator would find that donation of your code's copyright to the IP holders as a group would be the most acceptable form of restitution (trying to work out monentary damages from a class action lawsuit by a mass of coders with no interest in your money would be hard to resolve), which would put you back at square one, except without your money.
McBride: We will admit the things we've contributed and that we can't claw them back.
Darl, your second-in-command just said otherwise five seconds ago. C'mon, guys. At least maintain a cohesive position.
We think we have protection under both the GPL and copyright law.
This makes no sense. Name one right granted you by the GPL to either your IP or anyone else's IP that would entitle you to "protection" from other people using this code. If your code or other people's code is GPLed, everyone is clearly in the right to use it.
the copyright holder must make an explicit assignment, typically in writing, in a contract.
No. Team-written software is a form of joint authorship, which does not require explicit copyright assignment. While SCO might be able to argue that perhaps they have sole copyright ownership of the patch itself, the patched work is also owned by all the other authors of Linux, who
So UNIX is better because you have to go out of your way to do something?
Because you have to go out of your way to do something less desireable, the "non-common case". If it was really easy to, say, blow away all your files instead of a user having to go "out of their way", I'd call that a misfeature.
Seriously, I think the no-sharing default is to help prevent anamalous behavior. Sharing a file requries extra planning; you wouldn't share a block of memory between two threads, read/write, without a lock; exclusive access provides a primitive kind of locking. If you don't want the file to be locked, you only have to pass one extra flag when opening the file. But you are right that it is silly to deny read access when it's only open for read anyways.
Tell me you've never gotten a sharing violation when using Windows. Describe to me under what circumstances you would want to avoid reading from a file by two processes at once. Tell me you haven't rebooted when installing software. I've used both systems, and haven't ever hit problems with the UNIX approach, and the Windows approach has caused me countless grief. I've worked with a Windows fileserver and a cluster of machines that were running MS Visual Studio and Explorer. Inevitably, MS Visual Studio on some machine would have a file locked or Explorer a directory locked, and to delete a directory on one machine I'd have to go to every machine killing off all the processes that might be using the file/directory. Incredibly stupid. On *IX, you blow away a file, and the OS refcounts the thing. It doesn't break any applications currently using the file -- the file just doesn't have a directory entry any more, and when the last application using a file goes away, so does the file.
If you don't want the file to be locked, you only have to pass one extra flag when opening the file
That's not the point. The problem is that *developers don't*. They plop a zero in that field and don't worry about it. The net effect is that a user can't delete a file because something has it open without passing the shared flag. He didn't write the program, he doesn't have the source, and there are roughly zero instances where not locking the file is going to cause problems. You don't see folks in the UNIX world walking around with corrupted files, you know?
Yeah, sharing violation errors. However, sharing violation errors are obvious and direct. Insufficent locking can result in corruption and intermittent behavior that is hard to diagnose.
Yes, in theory. And in theory memory overcommits can cause massive unpredictable system failure, but you don't see your typical Linux system dropping on its feet. In theory languages that use a calling stack (C, C++, etc) can have completely unavoidable and unpredictable deaths due to exhausting memory by growing their stack when there aren't any pages left, but it's not a real-world problem either. People don't write apps that try dumping data to the same file at once (if they did, say by writing to a log file, Linux users would probably see mysteriously corrupt entries and Windows users would probably see mysteriously missing entries). Hell, every 1/(very large number) times key generation fails uncatchably for any prime-factor based cryptosystem, but we don't worry about it. They aren't real-world issues.
I created a new directory, changed to it in a command prompt, and tried to delete it. Explorer told me that it couldn't be deleted because it was in use. This is on XP (not that Explorer is anywhere near perfect). If I didn't know what has something open, there is always proces explorer, where I can search for handles and force them to close.
Not in my testing, which was over a file server. XP failed silently, whereas 2000 failed with an error. I could be wrong, but I doubt that Process Explorer will let me kill off said handles from a remote system (and certainly not if the access is from a different account...I might even have to go sit in front of the file server to run Process Explorer...I'll admit that it could have been handy other times that I've run into issues though, and didn't know about it).
My advice is to stop making Linux the "elite" operating system. If an "instruction-manual-following monkey" can get the system up and doing what it's supposed to, mission accomplished.
Look, I'm not trying to criticize people that aren't using Linux. I'd say that it takes an equally long period of time to *really know* Windows. I just think that fewer people *really know* Windows. I know probably one person that *really knows* Windows, but I've run into many *IX admins that know an *IX system inside out.
And I've seen clueless *IX admins as well, so I'm certainly not trying to make a statement about all admins on either platforms.
How many Windows admins know (and I'm not a serious Windows guy, so I'm sure I'm missing crucial tools) Dependency Walker, Regmon, and Filemon around, for those times when *something* has changed and things aren't working? How many people could fix a system where Explorer has started refusing to boot, or would know what to start doing? What if a file association mysteriously didn't show up in Explorer and applications couldn't register that association all of a sudden? What if an admin password goes missing? What if some user tries installing Linux and blows away the Windows boot loader? I'm sure there are tons of Active Directory weirdnesses that a Real Windows Guru will know how to deal with that I can't even begin to describe.
The thing is, I really think that the only way to learn a system is to use it. A lot. And fix it when something goes wrong. And it's really hard to do that if your primary system is Windows and most of the problems someone has you fix are Windows-based. The same would go for a Mac OS X fan who maintains a Windows server or two or (me, a Linux guy) for maintaining a major Windows installation at a Fortune 500 company. It'd be silly. It'd not because I'm stupid, it's just because I can't learn everything about admining a Windows system in a week or probably even six months. And because Linux is new, a lot of admins are being handed a training course, and expected to be able to fix all problems. Their bosses are going to expect that if something goes down, the system will be back up again shortly. So suddenly there is this huge mass of newbie Linux admins expected to handle critical Linux systems. A lot of them have no interest in ever learning more, and are going to stay right at that level.
The fact that you're running Linux at home makes a pretty strong statement that you're interested in doing more than just yanking out the manual and never learning more than the ten things you had to do the last time something went wrong. Same would go for someone running MS Exchange or Solaris or Photoshop at home to learn it -- they aren't just doing the bare minimum to get by.
Could have easily eaten the entire thing in one bite. I took two, and they were possibly the best two bites i'd had in my life until then. Standard American food is... crap.
I'd just like to point out that this food was:
* freshly prepared (not under constraints that it exist in a freezer or on a shelf for weeks or months)
* You were paying attention to it and savoring it.
That doesn't mean that you're wrong that the food was good. However, the difference between fresh food (like fish just out of a lake or bread straight from the oven) made by an amateur and a master is much less (IMHO) than the difference between new and old food produced by the same person.
And the environment and your status matters a lot too. Dinty Moore stew is not what you'd call fine cuisine. I have a pretty low opinion of the stuff in an ordinary environment. However, I remember camping -- a long hike all day with no food, and I was starving, and waiting slowly as a can of the stuff heated up in a beat-up old pot in the cold. When I finally got the stew, I ate it slowly, in bliss. No spices, nothing fancy, but that stew was one of the most amazing experiences ever. I was just terribly hungry and had been anticipating and smelling the stew for some time. It was *astounding*, the difference being really hungry made. Normally, I just snack whenever I'm hungry, so I don't get really, really hungry, and I don't frequently exercise and work up an apetite. Try skipping lunch and then going for a jog before supper. It ratches up the quality of your dinner *immensely*.
How to make your Linux box boot like a fiend
on
Why PHBs Fear Linux
·
· Score: 1
My XP laptop boots into a useable GUI in about 8 seconds. SuSE takes 30 seconds.
How many servers and whatnot is it running?
Seriously, I've never figured out why Linux distributors don't have their Linux systems have the option to just start X11 early in the init process. Most users don't care about a failing service at startup (that's what syslog is for, and if something fails, you can always print an error onscreen once xdm/gdm/kdm is up and going).
XFree86 takes maybe a second or two to get to a login screen on my system. If grub/lilo is set to no delay, and XFree86 is started *first thing* in the init process (as opposed to last, as it is now), it should be possible to get to an X login prompt in something like seven or eight seconds after the grub screen appears. XP just lets all the other services start up in the background while the user is smacking away at the login screen. Linux can pretty easily be told to do the same thing.
Maybe I should just start up a business that does nothing but set up and train users with free software for a small fee. *grin* It'd still be cheaper than any of the solutions from Redmond...
If you do so, do it soon. For a couple of years, there's going to be an increasing flood of demand for this. It's easier to get into an expanding market than a static one...
A lot of people would be appalled to know that the crypto algorithms that run banks, military hardware, and the computer systems at Microsoft, Sun, Oracle, IBM, and Verisign (eeeeevil Verisign) are all open.
Re:textbooks in general are horrible
on
Why PHBs Fear Linux
·
· Score: 2, Interesting
from many of text books, you would think that the US is the greatest perpetrator of evil in the world. or at least, no better than most other nations. (okay slash trolls, flame on)
Really? I've found that US textbooks (assuming you teach in the US) are reasonably pro-US, though not to the point of USSR textbooks. Mine included things that reflected negatively on the US, like the origins of the Panama Canal, the origins of the Spanish-American and Mexican-American wars, the fact that the US didn't care too much about helping Jews (and had its own anti-Semetic problems at the beginning of World War II), the fact that the Emancipation Proclamation was less about liberty than a political tool, US finagling in Korea and Iran, points out the fact that the US was the a major bastion of slavery and so on. It gets much more rah-rah-rah, Red, White and Blue during the Cold War, though, and doesn't poke very hard at US-induced nastiness during that period.
/dev/cdrom/mnt/cdrom auto user,ro,noauto,nosuid 0 0
(assuming your administrator has a symlink pointing from/dev/cdrom to/dev/scd0 or/dev/hdc or whatever your cdrom drive is, as he should) in fstab will let you simply type the command "mount/mnt/cdrom" or "umount/mnt/cdrom". The distro should have set this up automatically. I'm not familiar with the automounting software that some people use, where the CD is mounted when accessed. The only major usability difference between Linux and Windows here is that you have to wait a bit (with automount) before hitting the eject button on Linux, or manually unmount or software eject a CD-ROM.
The admins probably failed to put the "user" bit in the file, so only root has access to mount a CD-ROM.
I dunno about printing -- I agree that I've run into some seriously stupid print filters in the past (lpr [pdf file] should *damn* well run the pdf file through a filter so that it prints nicely, not require you to manually convert the pdf to a ps).
His arguement was that it is not easy to use, it is not guarenteed to continue into the future, and there is no one to be held accountable for failures or for fixes.
Wow.
not easy to use
I'd give it currently, from an end-user standpoint, about roughly equal to Windows. It is different, though, which means that for a user skilled in Windows, it is more difficult to use at first, until they become familiar with the differences.
it is not guarenteed to continue into the future
I will bet a million bucks that the Linux kernel will be around longer than the Windows NT kernel. There is one company working on the NT kernel -- there are many people working on Linux. Many companies have an investment and the ability and desire to continue using it, and nobody has the ability to "discontinue" Linux.
Or did he mean the APIs? UNIX system and library APIs have been more or less constant since the *'70*s. On Windows, a programmer has had to learn (get ready for it) DOS goodies, Win16, Win32, potentially the missing functionality in Windows CE and the added functionality in WinNT (which, frankly, is vastly more of a pain in the ass than the differences between even "different operating systems" like FreeBSD and Linux). Toss MFC into the mix. Now Microsoft's moving their developers to.NET. This is all covering a span of under twenty years.
Or maybe he was talking about the applications? Sysadmins might learn an application and then it's yanked out from under their feet...but sendmail (then called delivermail) shipped in the *'70*s. How about Apache? It started out as NCSA httpd, and was the second web server ever written.
there is no one to be held accountable for failures or for fixes
Absurd. Unless you are Dell or the US Government (and then only *maybe*), Microsoft does not *care* whether there's a bug in Windows. Name one instance where someone successfully sued Microsoft for a flaw in, say, Windows, and recieved damages for the problems caused by it. You can call Microsoft "accountable" all you want -- they are simply not.
In the Open Source world, I can sit down right now and email the main author, the development team, the maintainer, or the author of a particular feature (and usually *exact* line of code that I care about). I can generally enter bugs into the same bug-tracking system that the developers themselves use. If I'm in a hurry and need a contract for a fix within a certain time bound, I can hire a contractor to fix a bug or add a feature and send that fix to them, even if my company does not have any in-house developers capable of fixing the problem. I can discuss the problem at a technical level and point out the exact lines of code causing the problem publically, with every interested eye in the world trained on the bug. Linux has seen bug fix times for crucial bugs on the order of less than an hour ("there's a TCP bug that needs to be fixed *NOW*) "we need a fix out ASAP". Let's say you use Photoshop and report a bug to Adobe. Maybe, if you're lucky, they'll fix a bug. WilberWorks (a company formed by some GIMP developers) sells service contracts with guarantees that bugs you run into and require fixes for will be fixed within ten *days*. Try getting Adobe interested in doing something like that. Plus, if I don't like WilberWorks, I can hire anyone else to deal with my problem -- there are consultants and programmers-for-hire all over, and I can pay them whatever it takes or have them sign whatever contract I want to get them to fix my problem. Getting someone to be accountable to ensure that Open Source works is much easier than closed source products, where you have only one option -- the original vendor, which generally does not provide support on par with open source developers that provide support contracts (at least of the ones I've noticed). Most closed-source companies have churn, and do not keep developers on a single project. Microsoft, for exampl
Look at the CreateFile() documentation. While CreateFile() technically supports sharing, the default is off, which means that if you're doing the obvious thing, you'll get sharing violations if anyone tries to, say, open a file for reading that you have open for reading.
On UNIX, you have to go out of your way to lock the file if you want exclusive access (and even if you do so, the lock is merely advisory -- forced locks are done via the permission system). The default is to *not* have a file locked.
Guess which is a more common problem -- issues caused by files not being locked when they should, or the dreaded "sharing violation"?
Plus, the aforementioned "sharing violation" is why Explorer frequently can't delete directories (in XP this is "fixed" to become a silent failure rather than popping up a messagebox alerting the user that the directory could not be deleted) and Windows generally has to be rebooted after major updates, whereas Linux doesn't.
I agree that the reasoning in the textbook is that of a doped-up orangutan, but there's a kernel of truth. It isn't a real-world issue, but it's probably something that the article author heard from some Windows guy who had once run into *IX and been surprised by the way things work.
Re:From sSomeone who pitches those PHB's...
on
Why PHBs Fear Linux
·
· Score: 1
No, a PHB is not going to call Red Hat, he's going to be pitched on FOSS by a consultant like myself. My experience has been that these consultants are, by and large, pretty slanted in the first place.
Okay, fair enough -- but it's not as if slanted consulting is uncommon for other software products, either. Try getting a consultant (who primarily deals with one package) to recommend a CAD package, say. Ugh.
Re:If you've ever wondered why your PHB...
on
Why PHBs Fear Linux
·
· Score: 1
And you read Slashdot. That's one *hell* of a bizarre personality you have going there.
Slashdot Mirror
Not feasible. The kernel contains a huge quantity of functionality, and including an explanation on each thing (besides wasting the time of developers) would grow the changelog to an insane size. I don't know about you, but I already don't read through every bit of the changelog. I only worry about a new kernel when I need new functionality or a bugfix.
i ng-the-latest, just use the kernel if it has new functionality that you want. Otherwise, just search for "corruption" in the changelog (the devs seem to be good about marking corruption-causing problems as corruption-causing) and if you have any of the affected hardware, upgrade.
For the guy who is irritated that his netconsole has been malfunctioning...he's a happy guy.
For the typical user-that-rebuilds-his-kernel-for-the-sake-of-hav
It would take a *huge* amount of explanation to describe every modification and why it happens to the end user, on the level of having an ongoing OS class. Microsoft just doesn't talk about most of their bugfixes. With Linux you can get the detailed changelog, various summaries people have made, whatever advisories/errata the distro maintainers have put out, etc.
I believe the quote in your .sig should be "...little or no influence...".
Personally I like the idea. I've always been confused trying to locate various files which belong to a single application in *nix.
Use rpm -ql [packagename]
Err...you're quoting an eBay price. That's hardly sustainable or standard. I can occasionally obtain a nice piece of hardware for free, too, but it isn't reliable.
As for the entire concept of using tablet PCs -- tablets are generally *more* expensive than the equivalent laptop, which would make a much better base system.
Sorry folks, we have the technology right now to support multiple version of libraries at the same time
Why would you want to do that?
On my Fedora box, if I upgrade glibc to fix a bug, I want *all* my applications to benefit.
Oh, and disk space is not the reason for having shared libraries -- memory usage is.
This sort of thing occurs in steps - gradually. However, eventually, a lot of people may well wake up one morning and say "Holy shit! It's 1984!"
We have a "Department of Defense" used to invade other countries, an "Office of Homeland Security" that can detain and monitor people without being controlled by the judicial branch and ignores any rights that they once had, a figurehead leader that keeps exhorting us to greater national security efforts against an ephemeral, unbeatable foe, and monitoring of your movements (you carry a cell phone, right?) and your purchases.
What do you *need*? Wake UP!
And yet the amazing thing is that Canadians don't get attacked by terrorists, because they don't actively play dirty tricks with the people and governments of other nations.
TV subscriptions are one thing because most shows are transient, and you can record forever the ones you like. But a music subscription offers no similar benefits, only an ongoing cost and limitations on use (can you burn real CD's with a subscription service?)
It's a reasonable bet that Microsoft is not providing any guarantees that their rates will stay the same. In the past, they have *consistently* followed the forumula of doing whatever it takes to get into and dominating a market, then leveraging that market for all it's worth. If they get 80% of the US population using Janus, as people do TV, then they need only raise rates, as the ISPs did. Sure, you could cancel -- but then they have all the money you've given them up to that point, and you have nothing to show for your money.
You know, I get frusterated when people won't admit a problem, but really, that has nothing on Foale and his counterpart.
FOALE: "Uh, Houston? We've got something, uh, drumming on our outer hull. Sounds kind of like a sheet of metal bending on itself or something."
HOUSTON: "Okay, let's take a look...well...nope, everything's working fine."
FOALE: "Really, it's kind of loud. And this is the second time -- last time we didn't get to try and find out what was going on."
HOUSTON: (thinks "Darn astronauts, it's all in their heads") "Well, we're all green here. You could wait a couple months to take a look at it last time. You can probably wait to do anything on this until the new guys arrive.
FOALE: (thinks "Dammit, I'm sitting here in a little pocket of air a foot away from hard vacuum and something is *breaking*.") "Yeah, how about we do a spacewalk sooner."
HOUSTON: "I'll 'escalate' the problem and we'll see what we can do. Just sit tight up there."
Genocide was certainly not a goal of Hitler that was clear to the public four years into his rule (it was not until well into World War II that genocide became a goal), though he exhibited stronger anti-racial standpoints at that time than Bush currently is.
Exile was the proposed answer instead.
I have, on my refrigerator, a picture of a man in the Netherlands. The picture dates back about a month ago. He has sewn his own eyes and mouth shut with heavy thread. He is being held his tearful Dutch wife. He is one of thousands of Iraqis that is now being forced out of the country, to go back to Iraq, thanks to pressure from Bush. I would like to know what more people have to do to get their message across -- that they they are desperate and being treated horribly badly.
The really awful thing is that a major thing we used to think despicable about Soviet Russia and Nazi Germany was the identification papers and the restrictions on travel.
With computer databases, your image and your fingerprints *are* identification papers, and now you are being forced to hand them over at checkpoints.
Seriously, it was all very funny when we *started* to point out the amazing number of similarities between Hitler and Bush's rise. There was a terrorist act on a national monument (and even, in the 9/11 case, *attempted* on the national legislature, same as Germany) that produced national fear, whipped up by leader, used to convince legislature to pass through critical bills granting extensive police powers. Political opponents were accused of being soft on terrorism. Fear and xenophobia against religious (Islamic/Christian) and racial (Arabic/Jewish) groups was used to greatly infringe those people's rights and persecute them. A number of undesireable people, in violation of national law, were locked up in a camp to isolate them from the rest of society (Guantanamo Bay/Nazi concentration camps). Nationalistic fervor was whipped up and whipped up again to build up a popular base. Personal vendettas were made good upon with the new power (Bush-Hussein/Hitler-a number of enemies). Other countries were invaded and occupied on poor pretexts, banking on the fact that other, less powerful, countries would not be willing to organize or do more than protest (Iraq/several countries). A primary motivation for the invasion was resources (and later Nazi invasion into the USSR was significantly for oil). Business and government had close ties, and war profiteer corporations did a number of nasty things to take advantage of cronyism with major political figures (Schindler's List is a nice example). Right now the third largest employer of armed forces in Iraq (after the US and Britain)are private corporations -- big companies that are answerable only to an extremely friendly occupational government that grants Iraqis almost no rights and consists mostly of people trying to curry favor with their US occupiers to try to get a more advantageous political position in the future. Neither leader is brilliant, but both are prone to violence and grudge-holding. Both managed to seize control of the legislature at about the time they gained office. Neither has much regard for the lives of the people they have conquered -- we have been using unarmed Iraqi guards as inspectors of cars into restricted areas before US personnel come close, making human shields out of them. Neither feels that international opinion is of much import. Both quickly established powerful police organizations with far stronger powers than their predecessors, little oversight, and the ability to bypass much of the judicial system (OHS/Gestapo). Both started their invasions based on punishing the terrorists that attacked their nation, and immediately spread out once they had the power they needed. Both had rising unemployment in their countries, and a growing degree of xenophobia towards foreign laborers.
There are some differences. Hitler respected and even idolized what Britain had done -- Bush treats Britain as a lapdog. Hitler actively physically intimidated his physical opponents -- Bush does not. Hitler invaded, occupied, and eliminated the governments of no countries within his first four years as ruler, whereas Bush invaded, occupied, and eliminated the governments of two countries within his first four years as ruler. Hitler wound up eventually killing many more people than Bush has thus far, though Bush is currently ahead for the first four years of rule. Hitler did not actively attempt to control other countries through diplomatic means -- Bush has a team that works hard to control other contries without needing to overthrow their government. Bush has computers and telecommunication monitoring systems, but Hitler did not.
Screw Goodwin's Law. The man didn't write it in 2004.
I'll leave
The first is the Novell copyright situation. To me, it's not clear who's in the right here.
McBride: Would you buy an operating system without the source-code copyright? If you don't have copyright, they can turn around the next day and screw you.
Sontag: Instead, they waited nine years.
McBride: We have no doubts that our Unix copyright claims are valid.
One must, of course, ask why SCO felt that they had to wait years before notifying Linux folks of their alleged horrific infringements, and then felt that it was necessary to avoid actually *telling* Linux folks what the alleged infringements year until months and multiple court orders forced them to do so.
Sontag: We don't have to knock out the GPL for us to succeed on the copyright issue. The GPL itself supports, in a lot of ways, our positions. Section 0 of the GPL states that the legit copyright holder has to place a notice assigning the copyright over to the GPL.
All these contributions of our IP did not have an assignment by SCO saying here, 'We assign these copyrights to the GPL.' The fact that we participated with Linux does not mean that we inadvertently contributed our code to the GPL. You can't contribute inadvertently to Linux. We feel we have a very strong position based on the GPL.
First, this tidbit:
'We assign these copyrights to the GPL.'
Okay, enough fun has been made of Sontag and McBride's lack of competence when it comes to IP, so I'll avoid the jokes. You don't "assign a copyright to a license" (though GNU contributors are required to assign their copyright to the FSF for a number of reasons, in addition to licensing it under the GPL -- Linux is not a GNU project.)
Uh, huh. The fact that you added them to a file containing a GPL header doesn't count, eh? It's been well understood for many years that one header works for multiple contributions. When it comes to licensing, intent matters, and there was very clearly intent to GPL this code. I can't understand how you could make any kind of a counterargument.
The fact that we participated with Linux does not mean that we inadvertently contributed our code to the GPL.
Well, the alternative you have is that you committed massive infringement of thousands of IP holders that licensed their Linux code under the GPL. It's one or the other, SCO. If you want to go after Linux (and it's a damned weak argument -- I can't see how you'd manage to win it), you're also admitting that you deliberately committed a far worse crime. The potential costs of years of theft of perhaps millions of copies of Linux would easily bankrupt your company. I would expect that a shrewd mediator would find that donation of your code's copyright to the IP holders as a group would be the most acceptable form of restitution (trying to work out monentary damages from a class action lawsuit by a mass of coders with no interest in your money would be hard to resolve), which would put you back at square one, except without your money.
McBride: We will admit the things we've contributed and that we can't claw them back.
Darl, your second-in-command just said otherwise five seconds ago. C'mon, guys. At least maintain a cohesive position.
We think we have protection under both the GPL and copyright law.
This makes no sense. Name one right granted you by the GPL to either your IP or anyone else's IP that would entitle you to "protection" from other people using this code. If your code or other people's code is GPLed, everyone is clearly in the right to use it.
the copyright holder must make an explicit assignment, typically in writing, in a contract.
No. Team-written software is a form of joint authorship, which does not require explicit copyright assignment. While SCO might be able to argue that perhaps they have sole copyright ownership of the patch itself, the patched work is also owned by all the other authors of Linux, who
So UNIX is better because you have to go out of your way to do something?
Because you have to go out of your way to do something less desireable, the "non-common case". If it was really easy to, say, blow away all your files instead of a user having to go "out of their way", I'd call that a misfeature.
Seriously, I think the no-sharing default is to help prevent anamalous behavior. Sharing a file requries extra planning; you wouldn't share a block of memory between two threads, read/write, without a lock; exclusive access provides a primitive kind of locking. If you don't want the file to be locked, you only have to pass one extra flag when opening the file. But you are right that it is silly to deny read access when it's only open for read anyways.
Tell me you've never gotten a sharing violation when using Windows. Describe to me under what circumstances you would want to avoid reading from a file by two processes at once. Tell me you haven't rebooted when installing software. I've used both systems, and haven't ever hit problems with the UNIX approach, and the Windows approach has caused me countless grief. I've worked with a Windows fileserver and a cluster of machines that were running MS Visual Studio and Explorer. Inevitably, MS Visual Studio on some machine would have a file locked or Explorer a directory locked, and to delete a directory on one machine I'd have to go to every machine killing off all the processes that might be using the file/directory. Incredibly stupid. On *IX, you blow away a file, and the OS refcounts the thing. It doesn't break any applications currently using the file -- the file just doesn't have a directory entry any more, and when the last application using a file goes away, so does the file.
If you don't want the file to be locked, you only have to pass one extra flag when opening the file
That's not the point. The problem is that *developers don't*. They plop a zero in that field and don't worry about it. The net effect is that a user can't delete a file because something has it open without passing the shared flag. He didn't write the program, he doesn't have the source, and there are roughly zero instances where not locking the file is going to cause problems. You don't see folks in the UNIX world walking around with corrupted files, you know?
Yeah, sharing violation errors. However, sharing violation errors are obvious and direct. Insufficent locking can result in corruption and intermittent behavior that is hard to diagnose.
Yes, in theory. And in theory memory overcommits can cause massive unpredictable system failure, but you don't see your typical Linux system dropping on its feet. In theory languages that use a calling stack (C, C++, etc) can have completely unavoidable and unpredictable deaths due to exhausting memory by growing their stack when there aren't any pages left, but it's not a real-world problem either. People don't write apps that try dumping data to the same file at once (if they did, say by writing to a log file, Linux users would probably see mysteriously corrupt entries and Windows users would probably see mysteriously missing entries). Hell, every 1/(very large number) times key generation fails uncatchably for any prime-factor based cryptosystem, but we don't worry about it. They aren't real-world issues.
I created a new directory, changed to it in a command prompt, and tried to delete it. Explorer told me that it couldn't be deleted because it was in use. This is on XP (not that Explorer is anywhere near perfect). If I didn't know what has something open, there is always proces explorer, where I can search for handles and force them to close.
Not in my testing, which was over a file server. XP failed silently, whereas 2000 failed with an error. I could be wrong, but I doubt that Process Explorer will let me kill off said handles from a remote system (and certainly not if the access is from a different account...I might even have to go sit in front of the file server to run Process Explorer...I'll admit that it could have been handy other times that I've run into issues though, and didn't know about it).
My advice is to stop making Linux the "elite" operating system. If an "instruction-manual-following monkey" can get the system up and doing what it's supposed to, mission accomplished.
Look, I'm not trying to criticize people that aren't using Linux. I'd say that it takes an equally long period of time to *really know* Windows. I just think that fewer people *really know* Windows. I know probably one person that *really knows* Windows, but I've run into many *IX admins that know an *IX system inside out.
And I've seen clueless *IX admins as well, so I'm certainly not trying to make a statement about all admins on either platforms.
How many Windows admins know (and I'm not a serious Windows guy, so I'm sure I'm missing crucial tools) Dependency Walker, Regmon, and Filemon around, for those times when *something* has changed and things aren't working? How many people could fix a system where Explorer has started refusing to boot, or would know what to start doing? What if a file association mysteriously didn't show up in Explorer and applications couldn't register that association all of a sudden? What if an admin password goes missing? What if some user tries installing Linux and blows away the Windows boot loader? I'm sure there are tons of Active Directory weirdnesses that a Real Windows Guru will know how to deal with that I can't even begin to describe.
The thing is, I really think that the only way to learn a system is to use it. A lot. And fix it when something goes wrong. And it's really hard to do that if your primary system is Windows and most of the problems someone has you fix are Windows-based. The same would go for a Mac OS X fan who maintains a Windows server or two or (me, a Linux guy) for maintaining a major Windows installation at a Fortune 500 company. It'd be silly. It'd not because I'm stupid, it's just because I can't learn everything about admining a Windows system in a week or probably even six months. And because Linux is new, a lot of admins are being handed a training course, and expected to be able to fix all problems. Their bosses are going to expect that if something goes down, the system will be back up again shortly. So suddenly there is this huge mass of newbie Linux admins expected to handle critical Linux systems. A lot of them have no interest in ever learning more, and are going to stay right at that level.
The fact that you're running Linux at home makes a pretty strong statement that you're interested in doing more than just yanking out the manual and never learning more than the ten things you had to do the last time something went wrong. Same would go for someone running MS Exchange or Solaris or Photoshop at home to learn it -- they aren't just doing the bare minimum to get by.
Could have easily eaten the entire thing in one bite. I took two, and they were possibly the best two bites i'd had in my life until then. Standard American food is... crap.
I'd just like to point out that this food was:
* freshly prepared (not under constraints that it exist in a freezer or on a shelf for weeks or months)
* You were paying attention to it and savoring it.
That doesn't mean that you're wrong that the food was good. However, the difference between fresh food (like fish just out of a lake or bread straight from the oven) made by an amateur and a master is much less (IMHO) than the difference between new and old food produced by the same person.
And the environment and your status matters a lot too. Dinty Moore stew is not what you'd call fine cuisine. I have a pretty low opinion of the stuff in an ordinary environment. However, I remember camping -- a long hike all day with no food, and I was starving, and waiting slowly as a can of the stuff heated up in a beat-up old pot in the cold. When I finally got the stew, I ate it slowly, in bliss. No spices, nothing fancy, but that stew was one of the most amazing experiences ever. I was just terribly hungry and had been anticipating and smelling the stew for some time. It was *astounding*, the difference being really hungry made. Normally, I just snack whenever I'm hungry, so I don't get really, really hungry, and I don't frequently exercise and work up an apetite. Try skipping lunch and then going for a jog before supper. It ratches up the quality of your dinner *immensely*.
My XP laptop boots into a useable GUI in about 8 seconds. SuSE takes 30 seconds.
How many servers and whatnot is it running?
Seriously, I've never figured out why Linux distributors don't have their Linux systems have the option to just start X11 early in the init process. Most users don't care about a failing service at startup (that's what syslog is for, and if something fails, you can always print an error onscreen once xdm/gdm/kdm is up and going).
XFree86 takes maybe a second or two to get to a login screen on my system. If grub/lilo is set to no delay, and XFree86 is started *first thing* in the init process (as opposed to last, as it is now), it should be possible to get to an X login prompt in something like seven or eight seconds after the grub screen appears. XP just lets all the other services start up in the background while the user is smacking away at the login screen. Linux can pretty easily be told to do the same thing.
Maybe I should just start up a business that does nothing but set up and train users with free software for a small fee. *grin* It'd still be cheaper than any of the solutions from Redmond...
If you do so, do it soon. For a couple of years, there's going to be an increasing flood of demand for this. It's easier to get into an expanding market than a static one...
A lot of people would be appalled to know that the crypto algorithms that run banks, military hardware, and the computer systems at Microsoft, Sun, Oracle, IBM, and Verisign (eeeeevil Verisign) are all open.
from many of text books, you would think that the US is the greatest perpetrator of evil in the world. or at least, no better than most other nations. (okay slash trolls, flame on)
Really? I've found that US textbooks (assuming you teach in the US) are reasonably pro-US, though not to the point of USSR textbooks. Mine included things that reflected negatively on the US, like the origins of the Panama Canal, the origins of the Spanish-American and Mexican-American wars, the fact that the US didn't care too much about helping Jews (and had its own anti-Semetic problems at the beginning of World War II), the fact that the Emancipation Proclamation was less about liberty than a political tool, US finagling in Korea and Iran, points out the fact that the US was the a major bastion of slavery and so on. It gets much more rah-rah-rah, Red, White and Blue during the Cold War, though, and doesn't poke very hard at US-induced nastiness during that period.
los, I would mention that Linux+Unices only have 8% of the marketplace while Windows occupies 85%
Not in the server world, that isn't the case. In the server world, *IX is dominant. For workstations, MS is dominant.
/dev/cdrom /mnt/cdrom auto user,ro,noauto,nosuid 0 0
/dev/cdrom to /dev/scd0 or /dev/hdc or whatever your cdrom drive is, as he should) in fstab will let you simply type the command "mount /mnt/cdrom" or "umount /mnt/cdrom". The distro should have set this up automatically. I'm not familiar with the automounting software that some people use, where the CD is mounted when accessed. The only major usability difference between Linux and Windows here is that you have to wait a bit (with automount) before hitting the eject button on Linux, or manually unmount or software eject a CD-ROM.
(assuming your administrator has a symlink pointing from
The admins probably failed to put the "user" bit in the file, so only root has access to mount a CD-ROM.
I dunno about printing -- I agree that I've run into some seriously stupid print filters in the past (lpr [pdf file] should *damn* well run the pdf file through a filter so that it prints nicely, not require you to manually convert the pdf to a ps).
His arguement was that it is not easy to use, it is not guarenteed to continue into the future, and there is no one to be held accountable for failures or for fixes.
.NET. This is all covering a span of under twenty years.
Wow.
not easy to use
I'd give it currently, from an end-user standpoint, about roughly equal to Windows. It is different, though, which means that for a user skilled in Windows, it is more difficult to use at first, until they become familiar with the differences.
it is not guarenteed to continue into the future
I will bet a million bucks that the Linux kernel will be around longer than the Windows NT kernel. There is one company working on the NT kernel -- there are many people working on Linux. Many companies have an investment and the ability and desire to continue using it, and nobody has the ability to "discontinue" Linux.
Or did he mean the APIs? UNIX system and library APIs have been more or less constant since the *'70*s. On Windows, a programmer has had to learn (get ready for it) DOS goodies, Win16, Win32, potentially the missing functionality in Windows CE and the added functionality in WinNT (which, frankly, is vastly more of a pain in the ass than the differences between even "different operating systems" like FreeBSD and Linux). Toss MFC into the mix. Now Microsoft's moving their developers to
Or maybe he was talking about the applications? Sysadmins might learn an application and then it's yanked out from under their feet...but sendmail (then called delivermail) shipped in the *'70*s. How about Apache? It started out as NCSA httpd, and was the second web server ever written.
there is no one to be held accountable for failures or for fixes
Absurd. Unless you are Dell or the US Government (and then only *maybe*), Microsoft does not *care* whether there's a bug in Windows. Name one instance where someone successfully sued Microsoft for a flaw in, say, Windows, and recieved damages for the problems caused by it. You can call Microsoft "accountable" all you want -- they are simply not.
In the Open Source world, I can sit down right now and email the main author, the development team, the maintainer, or the author of a particular feature (and usually *exact* line of code that I care about). I can generally enter bugs into the same bug-tracking system that the developers themselves use. If I'm in a hurry and need a contract for a fix within a certain time bound, I can hire a contractor to fix a bug or add a feature and send that fix to them, even if my company does not have any in-house developers capable of fixing the problem. I can discuss the problem at a technical level and point out the exact lines of code causing the problem publically, with every interested eye in the world trained on the bug. Linux has seen bug fix times for crucial bugs on the order of less than an hour ("there's a TCP bug that needs to be fixed *NOW*) "we need a fix out ASAP". Let's say you use Photoshop and report a bug to Adobe. Maybe, if you're lucky, they'll fix a bug. WilberWorks (a company formed by some GIMP developers) sells service contracts with guarantees that bugs you run into and require fixes for will be fixed within ten *days*. Try getting Adobe interested in doing something like that. Plus, if I don't like WilberWorks, I can hire anyone else to deal with my problem -- there are consultants and programmers-for-hire all over, and I can pay them whatever it takes or have them sign whatever contract I want to get them to fix my problem. Getting someone to be accountable to ensure that Open Source works is much easier than closed source products, where you have only one option -- the original vendor, which generally does not provide support on par with open source developers that provide support contracts (at least of the ones I've noticed). Most closed-source companies have churn, and do not keep developers on a single project. Microsoft, for exampl
Look at the CreateFile() documentation. While CreateFile() technically supports sharing, the default is off, which means that if you're doing the obvious thing, you'll get sharing violations if anyone tries to, say, open a file for reading that you have open for reading.
On UNIX, you have to go out of your way to lock the file if you want exclusive access (and even if you do so, the lock is merely advisory -- forced locks are done via the permission system). The default is to *not* have a file locked.
Guess which is a more common problem -- issues caused by files not being locked when they should, or the dreaded "sharing violation"?
Plus, the aforementioned "sharing violation" is why Explorer frequently can't delete directories (in XP this is "fixed" to become a silent failure rather than popping up a messagebox alerting the user that the directory could not be deleted) and Windows generally has to be rebooted after major updates, whereas Linux doesn't.
I agree that the reasoning in the textbook is that of a doped-up orangutan, but there's a kernel of truth. It isn't a real-world issue, but it's probably something that the article author heard from some Windows guy who had once run into *IX and been surprised by the way things work.
No, a PHB is not going to call Red Hat, he's going to be pitched on FOSS by a consultant like myself. My experience has been that these consultants are, by and large, pretty slanted in the first place.
Okay, fair enough -- but it's not as if slanted consulting is uncommon for other software products, either. Try getting a consultant (who primarily deals with one package) to recommend a CAD package, say. Ugh.
And you read Slashdot. That's one *hell* of a bizarre personality you have going there.