Windows XP SP2 is here... what you need to do

[Natalinasmpf]

What's the point? Linux has a "monlithic" kernel -- which is something that was declared obsolete by researchers and academics over a decade ago. It's a horrible architecture, everyone else is onto microkernels because they scale and port far better.

It doesn't matter if it's static or DLL, in the end it's running in kernel userspace... dynamic is just the better, more efficient approach.

Modules are dynamic, and whether a feature is modular or built-in or not is customisable.

It's an advantage to have a dynamic/modular/"fragmented" operating system.

What about the ability to choose? Linux can be both. I don't think its that much of an advantage, especially when it gets spread out and separated (thus loading slower during boot since it can't read all the sectors in one read).

I don't see any root vulnerabilities for Windows XP SP2, either. Don't see your point.

There are, just not discovered yet!

That's bull**** -- since Firefox 0.9 was released, there have been 6 security vulnerabilities fixed in it. It was released just over a month ago. And yes, they affect Linux as well.

Such as? Are they critical?

This is a myth that is easily debunked by common sense and reality...

How does it debunk it? Peer review seems to be the better option. Unless of course, you're the capitalistic type.

Ohhh...another great argument. Let's list a bunch of 5 year old games, then a company that went bankrupt. You did know that Loki went under because the game market for Linux is not supportable, right?

Aha, no. I just uh, took the free port of it (I had found it free!)...so basically its assets are open now.

Well, generally, its starting to be supportable, because the market growth is there.

See, there's your problem. Windows didn't cost me a damn penny. And it's legal. Didn't even need to burn a CD!

?!!!!

What, you are extremely lucky to get XP for free, or what?

The stupidest logic I've ever heard -- virus writers write to cause damage and to get famous for doing so. Why would they bother writing a virus for an OS with 2.2% marketshare?

Because that OS is harder to hack into?

Virus writers don't often do that to "get heard"; thats just a generalisation. Professional ones, do it to break into company information, extract credit card info...

Generally a lot of this information

Of course, google's servers are Linux-based, so is BBC's....wouldn't virus writers write viruses for them? They would get heard for bringing them down.

Btw, marketshare is 3.5%, not 2.2%. Its already above the Mac's marketshare.

I can rob old ladies, of which they are hundreds on the street, or rob a bank. If I'm a mafia criminal, guess which I would rob?

Virus writers, for big feats, would do the latter (in the form of electronics).

[Asher]

Originally posted by Natalinasmpf
Modules are dynamic, and whether a feature is modular or built-in or not is customisable.

What about the ability to choose? Linux can be both. I don't think its that much of an advantage, especially when it gets spread out and separated (thus loading slower during boot since it can't read all the sectors in one read).

If this was actually true, I wouldn't need to recompile the kernel when installing 3D-accelerated Nvidia drivers in Linux.

There are, just not discovered yet!

Duh, ditto for the latest Linux.

Such as? Are they critical?

Seeing as they can allow a remote server to run the code of their choice as someone views the website...I'm going to say yes.

How does it debunk it? Peer review seems to be the better option. Unless of course, you're the capitalistic type.

Peer review happens in commercial software as well, if not more often. The stupidity behind the "many eyes" theory is the common sense than very few people actually look at the code. You don't have many eyes looking at it, you have the possibility of many eyes looking at it. People seem to assume lots of people do, but they don't... meanwhile commercial software has extensive peer review processes (at IBM, each patch needs to be reviewed by 3 other programmers and a security team before it's accepted).

?!!!!

What, you are extremely lucky to get XP for free, or what?

School is part of the MSDN Academic Alliance.

Because that OS is harder to hack into?

Virus writers don't often do that to "get heard"; thats just a generalisation. Professional ones, do it to break into company information, extract credit card info...

Generally a lot of this information

Of course, google's servers are Linux-based, so is BBC's....wouldn't virus writers write viruses for them? They would get heard for bringing them down.

It's funny you bring that up, because as a matter of fact, even with Linux's minute marketshare, they make up the majority of actual business electronic intrusions:

http://www.zdnet.com.au/news/software/0,2000061733,39116229,00.htm

An analysis of hacker attacks on online servers in January by security consultancy mi2g found that Linux servers were the most frequently violated, accounting for 13,654 successful attacks, or 80 per cent of the survey total. Windows ran a distant second with 2,005 attacks. A more specific analysis of government servers also found Linux more susceptible, accounting for 57 per cent of all breaches.

Whoops, there goes your argument. Hackers go after what has the biggest marketshare in general -- Linux has a large-ish marketshare in the server area, and go figure it gets hacked a lot there. The reason why the numbers are so ****ing huge compared to Windows is because of ignorance like yours -- most Windows server admins jump on patches as they come out, most Linux people don't even know patches are out (for example, YOU).

[Verto]

MS vs. Apple v. etc OS arguments should be banned.

[Asher]

For being too easy?

[Verto]

For causing mass suicide.

[Asher]

I expected as much when I posted the thread.

Apple and Linux users always feel the need to prove themselves, even though it ends up backfiring.

[Ari Rahikkala]

[Asher]

I love you though, Ari.

[Natalinasmpf]

If this was actually true, I wouldn't need to recompile the kernel when installing 3D-accelerated Nvidia drivers in Linux.

Uh, sorry, you don't recompile the kernel when you install the drivers. You have to reinstall the drivers when you recompile the kernel. Which takes like what, 10 seconds?

Duh, ditto for the latest Linux.

At least its much more sparse. Like every major kernel revision, not every single little patch like in Windows...

Seeing as they can allow a remote server to run the code of their choice as someone views the website...I'm going to say yes.

Let me see, code like, win32 code? Especially when the user probably isn't a superuser? (Who is stupid to browse stuff on root anyway?)

You know, Windows users are superusers by default. My school installs NetOp monitoring programs, and "admin" stuff on XP, to restrict what sites we go to, make sure its "educational" and such.

Then suddenly I just use win32 API to tell a superuser program to launch the code of choice (like a backoffice shell to disable all the security) even when I'm not. Hey, at least THAT doesn't happen in X. Superuser programs in Linux don't have API and blindingly obey all API commands even when sent through a non-verified program.

(at IBM, each patch needs to be reviewed by 3 other programmers and a security team before it's accepted).

3 programmers! Wow! I bet they don't stay up 24/7, do they? At least a significant portion of the community is up, any time. Thousands.

School is part of the MSDN Academic Alliance.

But don't you know? TCO includes antivirus software....

It's funny you bring that up, because as a matter of fact, even with Linux's minute marketshare, they make up the majority of actual business electronic intrusions:

Do these companies know how to use iptables? And NOT run stuff as root?

[Asher]

Originally posted by Natalinasmpf
Uh, sorry, you don't recompile the kernel when you install the drivers.

Do you even use Linux? Yes, you do need to recompile the kernel to use 3D-accelerated Nvidia drivers.

I needed to enable GART or somesuch.

At least its much more sparse. Like every major kernel revision, not every single little patch like in Windows...

Rhetoric, pure rhetoric.

Let me see, code like, win32 code? Especially when the user probably isn't a superuser? (Who is stupid to browse stuff on root anyway?)

Like ANY code. Win32, Linux, MacOS...it doesn't matter if you're superuser, they could still wipe out all your documents...

3 programmers! Wow! I bet they don't stay up 24/7, do they? At least a significant portion of the community is up, any time. Thousands.

IBM employs 400,000 people around the world. Thank you for playing.

You do realize IBM is a huge multinational company with software labs in Europe, Asia, Africa, North America, and South America?

I don't know of ANY open source program which requires 3 qualified, skilled programmers to look at a patch before it's dropped into a build. Nor do I know of any that have a security team look at a patch every time as well. That is my point.

Do these companies know how to use iptables? And NOT run stuff as root?

This isn't about if they know how to run their OS. Virtually EVERY single Windows virus and intrusion can be prevented if the user followed directions, kept up to date, and properly configured their system.

In the end, it doesn't matter. Linux is hacked FAR more than Windows, with a smaller marketshare. Make up all the excuses you want, but your argument holds absolutely no merit.

[Natalinasmpf]

Filesystems?

Filesystems are much more better IMO. No defragging needed, much faster, journalising, I suppose you know reiserfs? (Both stable and the alpha filesystem that is Reiser4)

Lets see what Windows is coming up with for Longhorn: new revolutionary filesystem! Now you can put the type of file in the meta-data and a description of whats in it as well! Wow! Revolutionary!

Especially when konqueror in Linux can already do that, just not in the meta data. Which can get damaged, and then running something wrongly when Windows sees the type of file wrongly...

Do you even use Linux? Yes, you do need to recompile the kernel to use 3D-accelerated Nvidia drivers.

I use Nvidia drivers, and I don't have to. I just recompile the drivers, which takes 15 seconds. To stop X, uncompress the program, install, and such. 15 seconds.

Like ANY code. Win32, Linux, MacOS...it doesn't matter if you're superuser, they could still wipe out all your documents...

Yes, lets try passing a code for fdisking a partition without root access

I don't know of ANY open source program which requires 3 qualified, skilled programmers to look at a patch before it's dropped into a build. Nor do I know of any that have a security team look at a patch every time as well. That is my point.

.

A lot. Generally, a lot of people are tasked with obligations of having to regularly re-examine it, and check all details before actually passing it into the build.

Just recently, they caught someone trying to pass a backdoor into the kernel, but it didn't get through the build. Why? It was reviewed. Root vulnerability! OMG!

Fixed within hours.

In the end, it doesn't matter. Linux is hacked FAR more than Windows, with a smaller marketshare.

Consumeringly, no.

Corporately, perhaps, thats only because of a sharp rise on complacency when switching.

[Asher]

Originally posted by Natalinasmpf
Filesystems?

Filesystems are much more better IMO. No defragging needed, much faster, journalising, I suppose you know reiserfs? (Both stable and the alpha filesystem that is Reiser4)

Lets see what Windows is coming up with for Longhorn: new revolutionary filesystem! Now you can put the type of file in the meta-data and a description of whats in it as well! Wow! Revolutionary!

Especially when konqueror in Linux can already do that, just not in the meta data. Which can get damaged, and then running something wrongly when Windows sees the type of file wrongly...

That's just a journalling filesystem with meta-data -- NTFS does that as well.

What WFS is is more than stupid metadata...it's an SQL frontend to the filesystem. There's nothing even in development like that for Linux, so go fish.

I use Nvidia drivers, and I don't have to. I just recompile the drivers, which takes 15 seconds. To stop X, uncompress the program, install, and such. 15 seconds.

Is that because you've already added GART support to the kernel?

Yes, lets try passing a code for fdisking a partition without root access

That's not the point -- you can piss a user off mostly by deleting all of his files and settings. Deleting the programs themselves is but a minor inconvenience.

Consumeringly, no.

If you assert it, prove it.

Corporately, perhaps, thats only because of a sharp rise on complacency when switching.

Doesn't matter...

[Natalinasmpf]

Is that because you've already added GART support to the kernel?

I have no idea.

In fact, I used a very minimalist kernel. I didn't even compile ACPI in, or build it all. (Generally I don't see a need for it right now).......For graphics, I just enabled VESA and framebuffering.

That's not the point -- you can piss a user off mostly by deleting all of his files and settings. Deleting the programs themselves is but a minor inconvenience.

Especially when "deleting" it can be easily recovered? It simply erases it from the filesystem table. Actual shredding is about 100 times slower, noticed by the sudden increase in disk activity.

Aren't settings found in /etc/?

Thats a root problem. Custom user scripts = easily remade.

If you assert it, prove it.

So far, I believe its common knowledge that Windows intrusions into consumer PC"s outnumber consumer Linux desktops even when compared per capita.

That's just a journalling filesystem with meta-data -- NTFS does that as well.

What WFS is is more than stupid metadata...it's an SQL frontend to the filesystem. There's nothing even in development like that for Linux, so go fish.

Longhorn doesn't have journalling. Doesn't have dancing trees. Has fixed blocks, inefficient in handling small files, hmm....

it's an SQL frontend to the filesystem.

And merging a database with a filesystem helps functionality significantly how?

They are already quite good (with reiserfs) used separately.

[Fve Crathva]

Natalinasmpf, please stop using Linux.

SP

[Asher]

'nuf said, SP.