More seriously, I think a good idea would be for the various desktops to make an installer front-end that's agnostic about the package management system. So, you could have KInstaller and GInstaller and probably InstallerStep or something for the Windowmaker and GNUStep people. These installer systems would be able to use deb, rpm, portage, and tgz's (did I leave anybody out?), keep track of what's been put where, deal with dependencies/versioning, etc (yes, I know, apt, yum, and portage all do that but this way every Gnome user, for instance, has a single way to install and uninstall an application or library). This really is more of a desktop-level problem than a distro-level problem, if you ask me, and frankly Microsoft is pretty good at solving desktop-level problems -- it's just the stuff underneath that they don't do too well.
Damn guess I should have searched before I duped Sus3an's name on here. SJC also had a facilities guy named "Cobalt Blue". Not exactly 1337, but still weird.
Back in college my housing director's name was, no lie, "Sus3an". I figured at first that the "3" was just her trying to make a German "s" on a US keyboard but, no, it was in fact a 3.
Don't do it, dude. Your kids will have enough reasons to hate you without giving them stupid names.
No, dumbass, the difference is that closed-source companies keep it a secret (or doesn't know in the first place) when their servers are compromised while Gnome and Debian are very up-front about it.
If you think this kind of thing hasn't happened to Microsoft, Oracle, etc., you're wrong. They just like to keep it quiet.
Which do you think is a better decision-making process?
Oh, I don't know... how about picking a project manager in the first place who is familiar enough with the technology to make an informed decision about it?
Sadly the only way of implementing this with any confidence is a "trusted computing platform" sort of approach.
Well I don't know about that being the only way. When you install the security software, you do some sort of checksum on the executable file. This checksum gets put on some sort of write-once medium (PROM, etc.) and validated when the software is loaded into memory. Now, admittedly there could be malicious alteration of the validation process, but for that matter the same thing applies to trusted computing.
There hasn't been a real security-hole based epidemic since the Day the MSSQL Servers Died. All of these recent worms are still based on the tried and true Stupid User vulnerability: there are apparently still people who will actually double-click on an unrequested screensaver file they got in the email.
As much as I am a Linux snob, the fact remains that you could just as easily do that to stupid Linux users: if they're using their computer for anything useful, they're able to send mail and run executables; that's the only requirements for a worm like this.
I find Gentoo a lot easier than Debian to install. I've installed Gentoo on x86es and SPARCs without much of a headache. But try as I might I cannot get Debian to install right on any platform. I've tried dozens of times; I've gotten a bootable system maybe 5 times and never gotten X to work. For some reason installing Debian reminds me of programming a VCR, which I also can't do.
It's like on the one hand you have RedHat or SuSE-type installs where you get a nice GUI that makes installing easy. On the other end you have gentoo which gives you a full shell and I'm good with using a shell so that install was pretty easy too. But Debian lives in this weird in-between world (like a VCR's interface) where you don't have an intuitive GUI but you also don't have a shell's freedom to put what you want where you know it needs to go.
So, long story short, I can't get Debian to install, but I don't have problems with Gentoo. Maybe it's a matter of taste.
What weird, weird universe do a lot of these posters live in where you don't have to "fiddle" with Windows to get it to work? Hell, even SuSE required less fiddling than Win2k3.
I thought Kurtz got drummed out of the Homeland Security department (with no shortage of bad blood) after Congress gave his GovNet idea the cold shoulder. Maybe I'm remembering wrong; either way from what I remember of his proposals when he was in DHS they're all based around the idea of putting a (hopefully) impenetrable barrier (a Maginot Firewall?) around critical resources rather than constructing a compartmentalized defense-in-depth.
Am I wrong in remembering that Kurtz was politely but firmly fired? If so will he help CSIA or just make their lobbying efforts more awkward?
DRM simply cannot work without enforcement in the hardware. It it's in memory and it's an architecture remotely similary to what we now consider a "personal computer", I can copy it.
Object-orientation is primarily a design philosophy and only secondarily a language feature. If you want to write object oriented code, make your data structures contain the algorithms that process them -- this is very easy to do in Perl since Perl has first-class functions.
You don't have to call something an "instantiation of a class" to use it as an object: simply have it contain its own message handling methods and pass it messages.
So that's the answer then, we just change a few comments, recompile, and call it a derived work? Surly it can't be that simple?
I remember we were talking about something similar on the LDP a while ago when there was a big flap about the FDL vs. other documentation licenses. In fact, it's even simpler than you say: a "trivially derivative work" (ie, the exact same work) still counts as a derivative work. So, you can receive apache under the Apache license and then redistribute the same ("trivially derivative") package under the GPL.
Redistributions of source code must retain the above copyright notice, this list of conditions, and the following disclaimer.
Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution, and in the same place and form as other copyright, license and disclaimer information.
The end-user documentation included with the redistribution, if any, must include the following acknowledgment: "This product includes software developed by The XFree86 Project, Inc (http://www.xfree86.org/) and its contributors", in the same place and form as other third-party acknowledgments. Alternately, this acknowledgment may appear in the software itself, in the same form and location as other such third-party acknowledgments.
Except as contained in this notice, the name of The XFree86 Project, Inc shall not be used in advertising or otherwise to promote the sale, use or other dealings in this Software without prior written authorization from The XFree86 Project, Inc.
The GPL pretty much says that the only restrictions you can place on redistribution are the restrictions in the GPL. You can't add any other restrictions, however harmless they seem -- I guess the argument is that it's the start of a slippery slope, and IIRC some of the old UNIX systems had to have 3 screens worth of notices like that one.
I guess I can see RMS's point: if you let XFree require this notice, maybe Apache will want a notice too. And maybe Wall would want a similar notice for Perl. OK, that's annoying but we could live with it, even if we end up back in the UNIX days of multi-screen credit notices.
But then suppose NVidia releases a driver and says that you have to include an advertisement (not just a credit) for NVidia if you distribute it with the kernel -- or rather, they specify a credit message that many people would consider an advertisement. Well, now we've clearly crossed a line most Linux developers don't want to cross, but it's not exactly clear where that line was in the scenario I just mentioned. NVidia would say "all the other developers got to come up with their own credit text, why can't we say what we want to? If you don't let us have our free speech the terrorists have already won!"
And they'd have a point: if you let some people dictate terms to the GPL you don't have much grounds to keep others from dictating terms, and however innocent the first terms may seem, somebody will find a way to screw it up.
So, you have to kick the camel in the nose while that's still all he has in your tent.
But we don't have a record, in most cases. I'd say about 50% of human languages are pretty thoroughly documented. Another 30% have sketchy documentation. The final 20% are all but unknown to researchers (most of these are in New Guinea).
We really need to talk to the dying generation of New Guineans, Siberians, and Africans who speak these disappearing languages so that there will be a record, like you say. But do you have the money to send out a few thousand linguists? Me neither.
Are we assuming that the Verisign key is not available to the FBI?
Ummm... so?
All that means is that if *every single* communication using SSL were constantly being joe-jobbed by the FBI, they could alter what we read. They *still* couldn't decrypt something encrypted with my public key.
Anyone who can access a host computer that is hosted in a free society can set up their own free press on the Internet.
You haven't read your ISP's terms of service lately, have you? Home (and even small office) users usually have to agree that they will only be consumers of online content. This is only sporadicly enforced, but I imagine that enforcement is going to increase.
The simple fact is most "Internet Service Providers" do not actually provide Internet access because they want to limit your ability to host content; they want you to only view it.
Slashdot Mirror
Don't you mean "emerge"?
More seriously, I think a good idea would be for the various desktops to make an installer front-end that's agnostic about the package management system. So, you could have KInstaller and GInstaller and probably InstallerStep or something for the Windowmaker and GNUStep people. These installer systems would be able to use deb, rpm, portage, and tgz's (did I leave anybody out?), keep track of what's been put where, deal with dependencies/versioning, etc (yes, I know, apt, yum, and portage all do that but this way every Gnome user, for instance, has a single way to install and uninstall an application or library). This really is more of a desktop-level problem than a distro-level problem, if you ask me, and frankly Microsoft is pretty good at solving desktop-level problems -- it's just the stuff underneath that they don't do too well.
Obviously. What's it from?
Ummm... yeah I still have the letter from her saying that if I intoxicate another freshman I'm getting kicked out. S-U-S-3-A-N.
Huh?
Nope, she's real. Maybe she got the idea from Pratchett. I don't know. Google for her.
Damn guess I should have searched before I duped Sus3an's name on here. SJC also had a facilities guy named "Cobalt Blue". Not exactly 1337, but still weird.
Back in college my housing director's name was, no lie, "Sus3an". I figured at first that the "3" was just her trying to make a German "s" on a US keyboard but, no, it was in fact a 3.
Don't do it, dude. Your kids will have enough reasons to hate you without giving them stupid names.
It's the first day of April. Ring any bells?
No, dumbass, the difference is that closed-source companies keep it a secret (or doesn't know in the first place) when their servers are compromised while Gnome and Debian are very up-front about it.
If you think this kind of thing hasn't happened to Microsoft, Oracle, etc., you're wrong. They just like to keep it quiet.
Oh, I don't know... how about picking a project manager in the first place who is familiar enough with the technology to make an informed decision about it?
I kid, I kid...
Well I don't know about that being the only way. When you install the security software, you do some sort of checksum on the executable file. This checksum gets put on some sort of write-once medium (PROM, etc.) and validated when the software is loaded into memory. Now, admittedly there could be malicious alteration of the validation process, but for that matter the same thing applies to trusted computing.
There hasn't been a real security-hole based epidemic since the Day the MSSQL Servers Died. All of these recent worms are still based on the tried and true Stupid User vulnerability: there are apparently still people who will actually double-click on an unrequested screensaver file they got in the email.
As much as I am a Linux snob, the fact remains that you could just as easily do that to stupid Linux users: if they're using their computer for anything useful, they're able to send mail and run executables; that's the only requirements for a worm like this.
I find Gentoo a lot easier than Debian to install. I've installed Gentoo on x86es and SPARCs without much of a headache. But try as I might I cannot get Debian to install right on any platform. I've tried dozens of times; I've gotten a bootable system maybe 5 times and never gotten X to work. For some reason installing Debian reminds me of programming a VCR, which I also can't do.
It's like on the one hand you have RedHat or SuSE-type installs where you get a nice GUI that makes installing easy. On the other end you have gentoo which gives you a full shell and I'm good with using a shell so that install was pretty easy too. But Debian lives in this weird in-between world (like a VCR's interface) where you don't have an intuitive GUI but you also don't have a shell's freedom to put what you want where you know it needs to go.
So, long story short, I can't get Debian to install, but I don't have problems with Gentoo. Maybe it's a matter of taste.
I think postgres requires a kernel that doesn't have its thumb up its ass, so probably not
What weird, weird universe do a lot of these posters live in where you don't have to "fiddle" with Windows to get it to work? Hell, even SuSE required less fiddling than Win2k3.
Really, I'm not an idiot. I remember from CS 201 that the definition of "database" is broad enough to cover almost any set of facts.
What is a "database" for the purpose of this act? (yes I'm too lazy to RTFA)
I thought Kurtz got drummed out of the Homeland Security department (with no shortage of bad blood) after Congress gave his GovNet idea the cold shoulder. Maybe I'm remembering wrong; either way from what I remember of his proposals when he was in DHS they're all based around the idea of putting a (hopefully) impenetrable barrier (a Maginot Firewall?) around critical resources rather than constructing a compartmentalized defense-in-depth.
Am I wrong in remembering that Kurtz was politely but firmly fired? If so will he help CSIA or just make their lobbying efforts more awkward?
DRM simply cannot work without enforcement in the hardware. It it's in memory and it's an architecture remotely similary to what we now consider a "personal computer", I can copy it.
Object-orientation is primarily a design philosophy and only secondarily a language feature. If you want to write object oriented code, make your data structures contain the algorithms that process them -- this is very easy to do in Perl since Perl has first-class functions.
You don't have to call something an "instantiation of a class" to use it as an object: simply have it contain its own message handling methods and pass it messages.
I remember we were talking about something similar on the LDP a while ago when there was a big flap about the FDL vs. other documentation licenses. In fact, it's even simpler than you say: a "trivially derivative work" (ie, the exact same work) still counts as a derivative work. So, you can receive apache under the Apache license and then redistribute the same ("trivially derivative") package under the GPL.
IAAL(ayman)
From the new XFree license (emphasis added):
- Redistributions of source code must retain the above copyright notice, this list of conditions, and the following disclaimer.
- Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution, and in the same place and form as other copyright, license and disclaimer information.
- The end-user documentation included with the redistribution, if any, must include the following acknowledgment: "This product includes software developed by The XFree86 Project, Inc (http://www.xfree86.org/) and its contributors", in the same place and form as other third-party acknowledgments. Alternately, this acknowledgment may appear in the software itself, in the same form and location as other such third-party acknowledgments.
- Except as contained in this notice, the name of The XFree86 Project, Inc shall not be used in advertising or otherwise to promote the sale, use or other dealings in this Software without prior written authorization from The XFree86 Project, Inc.
The GPL pretty much says that the only restrictions you can place on redistribution are the restrictions in the GPL. You can't add any other restrictions, however harmless they seem -- I guess the argument is that it's the start of a slippery slope, and IIRC some of the old UNIX systems had to have 3 screens worth of notices like that one.I guess I can see RMS's point: if you let XFree require this notice, maybe Apache will want a notice too. And maybe Wall would want a similar notice for Perl. OK, that's annoying but we could live with it, even if we end up back in the UNIX days of multi-screen credit notices.
But then suppose NVidia releases a driver and says that you have to include an advertisement (not just a credit) for NVidia if you distribute it with the kernel -- or rather, they specify a credit message that many people would consider an advertisement. Well, now we've clearly crossed a line most Linux developers don't want to cross, but it's not exactly clear where that line was in the scenario I just mentioned. NVidia would say "all the other developers got to come up with their own credit text, why can't we say what we want to? If you don't let us have our free speech the terrorists have already won!"
And they'd have a point: if you let some people dictate terms to the GPL you don't have much grounds to keep others from dictating terms, and however innocent the first terms may seem, somebody will find a way to screw it up.
So, you have to kick the camel in the nose while that's still all he has in your tent.
But we don't have a record, in most cases. I'd say about 50% of human languages are pretty thoroughly documented. Another 30% have sketchy documentation. The final 20% are all but unknown to researchers (most of these are in New Guinea).
We really need to talk to the dying generation of New Guineans, Siberians, and Africans who speak these disappearing languages so that there will be a record, like you say. But do you have the money to send out a few thousand linguists? Me neither.
Ummm... so?
All that means is that if *every single* communication using SSL were constantly being joe-jobbed by the FBI, they could alter what we read. They *still* couldn't decrypt something encrypted with my public key.
Eh? It was created by the US Military and public universities. Politicians were all over it.
You haven't read your ISP's terms of service lately, have you? Home (and even small office) users usually have to agree that they will only be consumers of online content. This is only sporadicly enforced, but I imagine that enforcement is going to increase.
The simple fact is most "Internet Service Providers" do not actually provide Internet access because they want to limit your ability to host content; they want you to only view it.