Only true to an extent. If they know I subscribe to BUGTRAQ, they might tailor their spam to look like BUGTRAQ messages.
But if I'm not a BUGTRAQ subscriber, BUGTRAQ terms like "exploit", "buffer overflow", "shellcode", etc are probably marked as spammy in my filter. The same applies to mailing lists on cooking, grabage disposal and the like. Pretty well every topic has its jargon.
And even so, there's no reason I need to filter my BUGTRAQ messages through a spam filter. I know they're clean, so I can catch them before they get checked for spamminess. So I have no need to train my filters to mark them good.
I know you're saying that public archives are a way of determining non-commercial speech, but my point is that most public archives have specialized jargon, so they don't look like email from my friends.
Actually, all they know is that all the mail they sent to you is now considered "spammy". Either it was detected as "spammy", or it was undetected and you manually marked it.
So web bugs are irrelevent. All they know is they can't send the same kind of thing twice.
They still don't know what your criteria are for non-spammy email. And they don't know your complete criteria for spammy email, either.
So they have no way to determine a pattern that evades the filter.
Please don't say that. PPPoE may buy you advantages at sign-on time, but only by making things slower & more complicated for the rest of the session. Quake gave much better ping before my DSL provider switched to PPPoE.
That's not a good example. Not only do all optical mice have microcontrolers, but all USB mice conform to the same protocol. You can just plug in a USB mouse, and it will work. (The same is true of PS/2 mice.)
And here's the problem with your "standardized protocol" solution-- Although mice conform to a standardized protocol, different mice have different feature sets. Even though the standard driver will work, you generally want to install mouse-specific drivers to enable extra buttons and configure things. Thus, standard protocols don't do away with the need for device-specific drivers.
Generally, being more responsive comes at the cost of being less efficient. They talk about "throughput" a lot, when determining the effects of the preempt patch.
So for servers, preempt is probably not the way to go.
A lot of computer programs these days have co-op programs. So at least SOME CS grads should have experience writing real code.
And anyone who's interested in web development should know what Apache is, even if they've never used it.
And I believe that good style should be taught as part learning a language. The number of C++ programmers who don't use auto_ptrs or initializer lists or references is truly shocking.
Expectations vary. The expectation in a mall or a museum is that people are allowed to enter an area unless they are specifically forbidden (or prevented) from entering it.
The expectation of a house is that people are forbidden from entering it unless they are specifically allowed.
I'd argue that a public web site is more like a mall or a museum (or park, or planetarium, etc) than a house. If you're already letting anybody in, you've got to be pro-active about preventing people from entering the areas where you don't want them. That's why they have locks and "Staff only" signs.
That's just silly. A lot of work this time around has gone into things like preemption and the O(N) scheduler that improve Linux's responsiveness as a desktop system.
commodity cpus increase in performance much faster than any other kind of processor
Have you been watching the 3D scene? In a fairly short time, we've gone from Voodoo and TNT cards to shader-based Radeon and GeForce cards. For some applications, custom procssors are just what the doctor ordered.
The 'testing' time frame-- probably the closest thing to that is the planned Oct 31 feature freeze. After that, the focus should be on getting it into a releaseable state.
It occurs to me that proxy voting might be a better form of representative democracy than the current approach.
If every citizen could vote on every matter, most people wouldn't. But they could select a proxy to vote for them. They could switch to another proxy whenever they liked. Proxies would have one vote for each person who selected them.
There would be no such thing as "throwing your vote away" by voting for an unpopular political party. Each proxy would have one vote per person they represented. And on issues where you felt strongly, you could vote directly, rather than letting your proxy handle it.
This kind of change should encourage more diversity in political thinking, and might even restore some faith in the political process. That might lead to less apathy and more intelligent proxy selection.
Depends where you looked. The ver command might have identified it as Windows 95, but the Windows GUI referred to the "DOS prompt" and "Full-screen DOS mode" and used the MSDOS icon for commandline programs. I've never noticed any mention of DOS in the Win2k gui.
The point is not that it was either Windows or DOS. The point is, it was confused.
Actually, you could configure Win95 to boot to the dos prompt. Then you typed "win" and it booted the Windows GUI. And it was referred to as a "DOS prompt" in the Windows 95 GUI.
Yes, it's all hair-splitting, but it's pretty clear that the Windows GUI was a separate thing, just like Gnome/KDE/TWM/foo.
And that's not such a bad idea. Why not keep separate functionality separate? You could even make a version of Windows Explorer that didn't contain a web browser. Oh wait-- they already did that.
Linux is a kernel, not a political platform, no matter how RMS and a subset of its users try to turn it into one.
Wrong. Linux is a kernel, and therefore is in competition with Microsoft operating systems. Anything which attempts to compete with Microsoft operating systems supports the cause of breaking the MS monopoly. The original poster did not say that Linux was written for the express purpose of destroying the MS monopoly, just that it served that end.
And remember that one of the goals of Linux, expressed by Linus, is "World Domination".
The commandline tools "apt-get" and "apt-cache" are utilities that use APT, not APT itself. It's also used in graphical projects like Synaptic and gnome-apt, as well as the console-menu tool Aptitude.
This quote sums up the reviewer's whole attitude: "you can't be a Unix and try to sell your product to plain users too."
Seems the reviewer's upset that Suse is, well, a Linux distro. Her prescriptions for dealing with dependencies suggest she's never used apt, either.
And pointing at Windows as a good example of installation behavior is just silly. On Windows, dependencies are shipped with the application, and sometimes you wind up with system libraries getting overwritten with older versions. And sometimes the older version's better, and gets overwritten with a newer one. Microsoft's had to write new features like "Windows File Protection" because of this.
On one point, I will agree: an installer or package manager should be as simple as it can be. If you install a package, any dependencies it requires should be automatically installed.
But all this stuff is a solved problem. It boggles the mind that people would rather use their own wierd solution than build on apt.
Slashdot Mirror
Only true to an extent. If they know I subscribe to BUGTRAQ, they might tailor their spam to look like BUGTRAQ messages.
But if I'm not a BUGTRAQ subscriber, BUGTRAQ terms like "exploit", "buffer overflow", "shellcode", etc are probably marked as spammy in my filter. The same applies to mailing lists on cooking, grabage disposal and the like. Pretty well every topic has its jargon.
And even so, there's no reason I need to filter my BUGTRAQ messages through a spam filter. I know they're clean, so I can catch them before they get checked for spamminess. So I have no need to train my filters to mark them good.
I know you're saying that public archives are a way of determining non-commercial speech, but my point is that most public archives have specialized jargon, so they don't look like email from my friends.
Scary thought: what if Microsoft started putting web bugs in Hotmail messages?
Actually, all they know is that all the mail they sent to you is now considered "spammy". Either it was detected as "spammy", or it was undetected and you manually marked it.
So web bugs are irrelevent. All they know is they can't send the same kind of thing twice.
They still don't know what your criteria are for non-spammy email. And they don't know your complete criteria for spammy email, either.
So they have no way to determine a pattern that evades the filter.
DHCP and plain old IP.
Please don't say that. PPPoE may buy you advantages at sign-on time, but only by making things slower & more complicated for the rest of the session. Quake gave much better ping before my DSL provider switched to PPPoE.
This tool does work with windows. It's probably also possible to set it up as an alternate mail server for your users.
That's not a good example. Not only do all optical mice have microcontrolers, but all USB mice conform to the same protocol. You can just plug in a USB mouse, and it will work. (The same is true of PS/2 mice.)
And here's the problem with your "standardized protocol" solution-- Although mice conform to a standardized protocol, different mice have different feature sets. Even though the standard driver will work, you generally want to install mouse-specific drivers to enable extra buttons and configure things. Thus, standard protocols don't do away with the need for device-specific drivers.
Generally, being more responsive comes at the cost of being less efficient. They talk about "throughput" a lot, when determining the effects of the preempt patch.
So for servers, preempt is probably not the way to go.
A lot of computer programs these days have co-op programs. So at least SOME CS grads should have experience writing real code.
And anyone who's interested in web development should know what Apache is, even if they've never used it.
And I believe that good style should be taught as part learning a language. The number of C++ programmers who don't use auto_ptrs or initializer lists or references is truly shocking.
Actually, they call it "hyperthreading" because there was already a more limited form called "superthreading".
Expectations vary. The expectation in a mall or a museum is that people are allowed to enter an area unless they are specifically forbidden (or prevented) from entering it.
The expectation of a house is that people are forbidden from entering it unless they are specifically allowed.
I'd argue that a public web site is more like a mall or a museum (or park, or planetarium, etc) than a house. If you're already letting anybody in, you've got to be pro-active about preventing people from entering the areas where you don't want them. That's why they have locks and "Staff only" signs.
Hyperthreading works well for certain types of software, and awful for others.
Here's an article from Ars Technica on HT/SMT.
That's just silly. A lot of work this time around has gone into things like preemption and the O(N) scheduler that improve Linux's responsiveness as a desktop system.
Glass/LCD isn't a good distinction. LCD monitors often sandwich the liquid crystals between layers of glass. (Which can shatter, which is no fun.)
commodity cpus increase in performance much faster than any other kind of processor
Have you been watching the 3D scene? In a fairly short time, we've gone from Voodoo and TNT cards to shader-based Radeon and GeForce cards. For some applications, custom procssors are just what the doctor ordered.
What's sadder is, no. It's not methane. It's methanol, which may have the same number of carbon atoms, but it's an alcohol.
Yes. Here's a paper (in PDF) that discusses phytoextraction as a means of reducing heavy metal contamination:
Of course, the practicality depends on the level of concentration and the efficiency of the plants. . .
The 'testing' time frame-- probably the closest thing to that is the planned Oct 31 feature freeze. After that, the focus should be on getting it into a releaseable state.
It occurs to me that proxy voting might be a better form of representative democracy than the current approach.
If every citizen could vote on every matter, most people wouldn't. But they could select a proxy to vote for them. They could switch to another proxy whenever they liked. Proxies would have one vote for each person who selected them.
There would be no such thing as "throwing your vote away" by voting for an unpopular political party. Each proxy would have one vote per person they represented. And on issues where you felt strongly, you could vote directly, rather than letting your proxy handle it.
This kind of change should encourage more diversity in political thinking, and might even restore some faith in the political process. That might lead to less apathy and more intelligent proxy selection.
Not only can the rich vote while the poor are disenfranchised, but rich organizations can sponsor those voters who will vote in their interest.
Depends where you looked. The ver command might have identified it as Windows 95, but the Windows GUI referred to the "DOS prompt" and "Full-screen DOS mode" and used the MSDOS icon for commandline programs. I've never noticed any mention of DOS in the Win2k gui.
The point is not that it was either Windows or DOS. The point is, it was confused.
Actually, you could configure Win95 to boot to the dos prompt. Then you typed "win" and it booted the Windows GUI. And it was referred to as a "DOS prompt" in the Windows 95 GUI.
Yes, it's all hair-splitting, but it's pretty clear that the Windows GUI was a separate thing, just like Gnome/KDE/TWM/foo.
And that's not such a bad idea. Why not keep separate functionality separate? You could even make a version of Windows Explorer that didn't contain a web browser. Oh wait-- they already did that.
Linux is a kernel, not a political platform, no matter how RMS and a subset of its users try to turn it into one.
Wrong. Linux is a kernel, and therefore is in competition with Microsoft operating systems. Anything which attempts to compete with Microsoft operating systems supports the cause of breaking the MS monopoly. The original poster did not say that Linux was written for the express purpose of destroying the MS monopoly, just that it served that end.
And remember that one of the goals of Linux, expressed by Linus, is "World Domination".
Apt isn't a front-end. It's more like a library.
The commandline tools "apt-get" and "apt-cache" are utilities that use APT, not APT itself. It's also used in graphical projects like Synaptic and gnome-apt, as well as the console-menu tool Aptitude.
This is why I said "build on" APT.
This quote sums up the reviewer's whole attitude: "you can't be a Unix and try to sell your product to plain users too."
Seems the reviewer's upset that Suse is, well, a Linux distro. Her prescriptions for dealing with dependencies suggest she's never used apt, either.
And pointing at Windows as a good example of installation behavior is just silly. On Windows, dependencies are shipped with the application, and sometimes you wind up with system libraries getting overwritten with older versions. And sometimes the older version's better, and gets overwritten with a newer one. Microsoft's had to write new features like "Windows File Protection" because of this.
On one point, I will agree: an installer or package manager should be as simple as it can be. If you install a package, any dependencies it requires should be automatically installed.
But all this stuff is a solved problem. It boggles the mind that people would rather use their own wierd solution than build on apt.