The CSS Advanced Layout Module is being designed to fix that.
Thanks for the intresting link; to me, that upcoming draft looks like an attempt to embedd a layout language inside CSS. (More precisely: inside a "template string" CSS property called 'display' that uses what appears to be a rather non-CSS-like syntax...)
However, I would still really like to know if he agrees with you that something like that upcoming draft is greatly needed, and (as you say) current use of CSS for layout are "hacks".
A reply similar to yours from the person proposing CSS would be an intresting contribution in the debate on the use of (current) CSS for layout. There are a lot of people out there that advocates that what you call "hacks" are 'the right way of doing things'.
Re:Why is CSS such a good idea but a pain to use?
on
Ask Håkon About CSS or...?
·
· Score: 5, Interesting
... There are literally hundreds of design decisions that feel like they were implemented in a way to just make things as difficult as they can be.
I'd like to try to turn your observation into a more specific question:
CSS is clearly very useful for separating style from content. But apparently people tend to have problems when using it for layouts. Would you say this is because people have not yet understood how to properly do layout in CSS, or is it CSS that is lacking in this area? What can be done to improve the situation? --- Would the web benefit from HTML and CSS being complemented with some kind of "layout language"?
I remember in particular a biplane shoot-em-up that ran at ludicrous speed on a 486/33 in "turbo" mode.
Would that by any chance be 'sopwith'? http://sopwith.classicgaming.gamespy.com/ I remember trying to actually play that game with turbo enabled -- that was some hardcore fps:-)
For people who have not understood this: the Motorola Linux phones are running a *complete* Linux distribution:/proc,/dev,/usr,/etc. When you turn the phone on, its hidden startup is done sysV style with/etc/init.d scripts! There are ways to turn on telnet support. You do not have an X-server, but you have a framebuffer which can be used for QT programs etc. So, in short, they are even more than "like a PDA", they are pocket-versions of a system originally developed for multi-user servers (!).
Given this, it is frustrating to see how Motorola works hard to convince us to ignore this; but rather just pushes the phones as java platforms.
... even then, I'd doubt that the linux source is necessary, so much as just the development headers and utilities in order to build applications that can run upon it.
Right, and, as the article say, the UI is built on a proprietary api EZX (built upon QT), and the development headers for this library is not publically available. Hence, limiting developers who for example want to write applications that integrate with the standard UI.
I really doubt there's a large enough demand for the actual source to the kernel itself that it's killing development on the platform.
It might not be "killing development", but the kernel source is very useful for advanced hacks. When the source finally was released, it was not long until people in the "hack" community made avaialble kernel modules for e.g. loopback device mounting, new filesystems, and even overclocking (!) for these phones. These in turn enabled new advanced projects using loopback mounted filesystems etc.
Seeing as how you can access all the functions of the phone via their Java SDK, including camera, WiFi, multimedia, bluetooth, IR, and text messaging, I don't see what the difference would be.
First: this is simply not true. For example the E680i is at least missing both JSR75 (filesystem access) and JSR82 (bluetooth access) optional java apis.
Second: java applications don't have the look and feel of native apps. You can only have one java app running at any time and they don't integrate nicely into the UI.
Third: To say "why not just do it in java instead" is blatantly ignoring the most promising aspect with Linux devices: to port existing open source applications. These phones have been hacked to port VisualBoy Advance (gameboy emulator), Qonsole, Flashplayer, Doom, etc. To try to re-program these under java instead of porting them is just silly.
The Motorola SDK for their mobile phones is available right now, both the linux and non-linux varieties of phones.
Having experience with one of the Motorola phones myself, I belive the article describes the current situation very accurately. As the article explains: the public SDK is only for java development. The intresting thing with having a Linux phone is to develop native applications. There is no public SDK from Motorola for native applications. That is the problem.
Of course, in the days before discount cards they would have given those discounts to everyone and called them "sales". So, actually, they are giving out fewer discounts than they would have and getting marketing info for free.
This isn't strictly true: in theory the market info they get from the discount cards allows for tageted advertizing, which increase their sales. A procentage of the extra income goes towards extra sales. => Some of the sales given to people with discount cards wouldn't have been possible if there had not been discount cards.
But then, how big this effect is in practice can probably be discussed...
The following things are paranormal by definition: [...] Violations of Newton's Laws of Motion [...]
Eh?... If this is the case, then even I belive in the paranormal... Looks like Einstein has some cash to withdraw here. How do they look on discoveries of exotic particles? Seems like there might be some easy money here afterall...
As the voting form requires to vote for all categories it is not a good thing to do this if you have no clue who all these people are.
So, can someone with insight please line up the best patent-antanagonist choices in each categories (for extra credit: include motivations), so the crowd can wild...
... an "identify all the test cases" exercise with a simple problem: a function which classifies 3 side lengths with whichever type of triangle they could form. ... MAX_INT...
Eh... I've read your story and are left wondering; are you sure the exercise function used *integer* parameters rather than reals/floats? Few sets of 3 integers will form a triangle, and you'd have to implement a test for integer triangles. This seems to be a bit much for an excercise about limiting cases...
That said, if the exercise used reals, I suppose the solution still should have tested for overfolows; like checking for NaN:s and INFs.
If, by some miracle some company produces more than 9gb of content for their game, it isn't exactly difficult to put it onto a second disc (recall FFVII). As long as the disc doesn't have to be swapped every 3 hours it isn't a big deal.
I wonder if they will allow disc-swaps. Previous xbox hard-reset on eject. If this xbox don't do that, this could be a great opporunity for hackers...
Anyone want to explain the real world applications of this[?]
You leave us guessing at your background, but in case you are an "xbox hacker", here is an explanation for you:):
Let say "ax" and "by" happens to be colliding in SHA1. (They are not, but just pretend)
Lets say that an "insider" working on an xbox game on purpose put this into the code:
if(!strcmp("ax" == "by")) {
startup linux or something
} else {
run computer game
}
Lets say Microsoft signed that game.
THEN the "evil linux hackers" could take this game and exchange "by" for "ax" inside the executable. The executable would *still* confirm to Microsofts signature; the SHA1 would not have changed.
Result: Microsoft signed executable starting linux.
But --- I guess this isn't really that different from an "insider" inserting some other backdoor... So, its not that useful as an attack I guess...
Re:The provided exploit documents can be edited!
on
Meaningful MD5 Collisions
·
· Score: 4, Informative
It is the same document, just relying on differences in the document name (it appears) to generate the different pages.
No, you have missed the point. Go back and rtfa again. The attack still works if you rename the documents to the same filename.
The difference lies in a generated "binary cookie" in the beginning of the postscript documents. This "cookie" makes the postscript intepreter either select to show document 'A' or 'B'. The "thing" with the cookies are that they are carefully selected to be md5-colliding. Result: both documents have the same md5sum.
You can change the rest of the documents freely if you make the same changes in both documents. The md5sum will change, but it will still be the same for both documents.
OTOH, in the real world, one would be hard pressed to change the contents of an existing document.
You make it sound like the attack isn't feasible. I belive that is far from the truth in the *actual* real world.
All you need is to pull off some lame excuse: "Oh, they needed it in this format -- I had to convert your letter of recomendation to postscript; can you please sign this version too?".
Combine 1) This attack, 2) A little social engineering, 3) Someone who does not know about the recent discoveries of md5 collisions; and you have all you need for a successful attack.
The biggest problem I see is that this attack is done in *postscript* => something that your average Windows user cannot read at all. *That* will make your boss suspicious to sign it. Now, if it was 'pdf' on the other hand...
No, you and others who have replied to my post still don't seem to get it!
You open your favorite editor that don't destroy binary data (emacs is fine). You open the two postscript files. You will see how each contains a different "header", but after that *both* files include *both* messages(!)
Now, the headers are different and uses some postscript trickery to select which of the messages is going to be shown on screen, but yet "magically" manages to "preserve" the md5sum.
The rest of the two documents can be edited at will. You can for example change the word "Orders" to "Obligation to pay" if you wish. You just have to change this in *both* files. In file 'A' where it is not actually shown on screen, and in file 'B' where it is shown on screen.
Then check the md5sum's of the files. The md5sum will have *changed* relative to the old md5sum, but they will still be *exactly equal* between the two files.
It works. I have tested it. Please test it before you comment to say that it does not work.
What I haven't seen mentioned yet, and people perhaps haven't realized, is that in providing these two postscript files, they have essentially provided you with an postscript signature exploit kit!
All you need to do is download the two postscript documents and do *exactly corresponding edits* in both of them, and you get two documents saying different things and still have the same md5sums!
I just tried exchanging Alice's name for my own, and surely it did work.
Now, if they released a pdf-file hack, I would be genuinely worried:)...
Rolling distros are perfect for boys living at their parent's basament which think they have a life because they know what the last version of KDE or Gnome or whatever program is. Heck, adult people even haven't a "favourite program" to desire to upgrade as soon as its next version publishes!
Being in the position of part-time managing computers at an academic workplace I can state for a fact that what you say simply is not true. In this environment there are working adult people that sometimes have use for bleeding-edge features to get their work done. Contrary to what you think, the world just isn't divided into 1) grandmothers who are happy with a windows machine where wordpad and solitare works, 2) Kids who live in their parents basement.
What about people who happen to need a new mathematical function from the latest GNU scientific library? Should they wait for the next distro release? Of course -- they COULD spend the half day downloading the source and compiling it themselves -- but why waste time when you can be part of a rolling distro? Just install testing, and pull packages from unstable or experimental when needed.
To give another example: the other day I was messing around with the drawing program Inkscape for some figures to be used in an academic work. We wanted those figures exported as pdf so they could be used with pdfLatex. This "simple" task turned out to require the very latest beta-releases of both Inkscape and the desktop publishing program Scribus! What are your suggestion? Redo uglier figures in a simpler drawing program until our distro release a newer version?
"Instead, let some kind of specialized "release team" take care of making debian stable releases at whatever rate they feel apropriate simply by freezing the moving distribution and working on it until it is stable enough to be released as a new "debian stable"
Well, it seem you don't know what you talk about either! what do you think is the way Sarge is making its way into new Stable?
The way I described it is just not how it works today. Packages have for a longer time been frozen in testing/unstable while waiting for the sarge release to happen. My POINT was to suggest the moving relase to keep moving, with stable releases basically happening through: 1. Take a snapshot of the moving release. 2. Apply bugfixes and backports to the packages until they are stable enough, while the moving distro KEEP MOVING. 3. Release "stable". Again, this is not how it works today.
The stable distribution has a quite slow release cycle, but there is no reason for a desktop user to run the stable distribution. You can run either the unstable distribution, that regardless of its name is quite stable, or you can run the testing distribution.
There is a sad reason to not run testing: the testing distribution is the last one to get security updates; as I have understood this, unsecure packages from unstable can overwrite security fixed packages in testing. And I've seen people at debian-devel rave about how the reason for this is that "testing/unstable is not really meant to be used like this."...
But I wish the debian community would unite behind the picture you have, and with testing as the recommended desktop distribution; becuase then the focus would be on what debian does best of all distributions: to provide a rocking *moving* desktop distribution.
For a broadband user, a moving desktop distribution makes a lot of sense. You don't have to wait for the next release (as with Ubuntu, Fedora) to get the next version of your favourite program -- instead your desktop will be continously updated as new versions becomes available. (With just a small time-lag due to the package dependencies etc. being before being moved to testing). And the thing is: I belive this is how most people currently are using debian! Why not make this the supported way?
If it was my decision (which it clearly isn't), I would work for making the moving Linux distribution provided by testing, unstable, and experimental fully independent from the concept of "releases". Instead, let some kind of specialized "release team" take care of making debian stable releases at whatever rate they feel apropriate simply by freezing the moving distribution and working on it until it is stable enough to be released as a new "debian stable".
You couldn't go back in time and shoot Hilter before he got into power for the very simple reason that it didn't happen.
No, no, no; you got it all wrong. It was just exactly because someone went back and shot that Hilter you speak of; that the much worse dictator Hitler we do remember could come to power.:)
This is a total misconception of how evolution works. [...] The mutation described in the article is clearly a coincidental mutation that came about much earlier, not one that evolved due to the current 25-year-old aids epidemic.
But isn't this exactly how evolution works? There is never any "active adaption", but mutations happens randomly and may by pure luck give a beneficial property. These benefits makes the survival rate of the children higher, and after a looooong time decendants of a few mutated individuals have replaced the entire species.
Notice that the "looooong time" in this is the time it takes for the gene to spread to the whole population, not the time it takes from an outbreak of a disease until it is possible to find a gene that protects against that disease. As you say is "clear" here, the latter probably usually happens long *before* the former.
I think by "rejected" the parent means there is not an example of a single society ever having existed that didn't have a form of government or social hierarchy. Even the anarchist communes of the 60's and 70's ended up with a ruling class (and often started out with one). It's an inevitable consequence of human psyche that you have government.
With this way of seeing things I guess one could say that we still have anarchy everywhere in the world today. Only that given anarchy, people have chosen to group, come up with rules, and violently enforce them upon others. But in an environment with a total absence of rules, people should be allowed to do that too.
Given how people are very willing to group in "governments" when left to their own devices, I guess for anarchy to really mean "absence of rules" one would have to create a system that enforces "no rules". Like, having a rule that say "nobody is allowed to set up any rules." But, but < head explodes... >
Slashdot Mirror
The CSS Advanced Layout Module is being designed to fix that.
Thanks for the intresting link; to me, that upcoming draft looks like an attempt to embedd a layout language inside CSS. (More precisely: inside a "template string" CSS property called 'display' that uses what appears to be a rather non-CSS-like syntax...)
However, I would still really like to know if he agrees with you that something like that upcoming draft is greatly needed, and (as you say) current use of CSS for layout are "hacks".
A reply similar to yours from the person proposing CSS would be an intresting contribution in the debate on the use of (current) CSS for layout. There are a lot of people out there that advocates that what you call "hacks" are 'the right way of doing things'.
... There are literally hundreds of design decisions that feel like they were implemented in a way to just make things as difficult as they can be.
I'd like to try to turn your observation into a more specific question:
CSS is clearly very useful for separating style from content. But apparently people tend to have problems when using it for layouts. Would you say this is because people have not yet understood how to properly do layout in CSS, or is it CSS that is lacking in this area? What can be done to improve the situation? --- Would the web benefit from HTML and CSS being complemented with some kind of "layout language"?
I remember in particular a biplane shoot-em-up that ran at ludicrous speed on a 486/33 in "turbo" mode.
:-)
Would that by any chance be 'sopwith'? http://sopwith.classicgaming.gamespy.com/
I remember trying to actually play that game with turbo enabled -- that was some hardcore fps
She said yes !! Look I got an email !!
Subject: ILOVEYOU
Well, technically, that is not a yes.
You should look for another email: "improve your stamina" for the reason...
Now, if the phone is actually more like a PDA...
/proc, /dev, /usr, /etc. When you turn the phone on, its hidden startup is done sysV style with /etc/init.d scripts! There are ways to turn on telnet support. You do not have an X-server, but you have a framebuffer which can be used for QT programs etc. So, in short, they are even more than "like a PDA", they are pocket-versions of a system originally developed for multi-user servers (!).
... even then, I'd doubt that the linux source is necessary, so much as just the development headers and utilities in order to build applications that can run upon it.
For people who have not understood this: the Motorola Linux phones are running a *complete* Linux distribution:
Given this, it is frustrating to see how Motorola works hard to convince us to ignore this; but rather just pushes the phones as java platforms.
Right, and, as the article say, the UI is built on a proprietary api EZX (built upon QT), and the development headers for this library is not publically available. Hence, limiting developers who for example want to write applications that integrate with the standard UI.
I really doubt there's a large enough demand for the actual source to the kernel itself that it's killing development on the platform.
It might not be "killing development", but the kernel source is very useful for advanced hacks. When the source finally was released, it was not long until people in the "hack" community made avaialble kernel modules for e.g. loopback device mounting, new filesystems, and even overclocking (!) for these phones. These in turn enabled new advanced projects using loopback mounted filesystems etc.
Seeing as how you can access all the functions of the phone via their Java SDK, including camera, WiFi, multimedia, bluetooth, IR, and text messaging, I don't see what the difference would be.
First: this is simply not true. For example the E680i is at least missing both JSR75 (filesystem access) and JSR82 (bluetooth access) optional java apis.
Second: java applications don't have the look and feel of native apps. You can only have one java app running at any time and they don't integrate nicely into the UI.
Third: To say "why not just do it in java instead" is blatantly ignoring the most promising aspect with Linux devices: to port existing open source applications. These phones have been hacked to port VisualBoy Advance (gameboy emulator), Qonsole, Flashplayer, Doom, etc. To try to re-program these under java instead of porting them is just silly.
The Motorola SDK for their mobile phones is available right now, both the linux and non-linux varieties of phones.
Having experience with one of the Motorola phones myself, I belive the article describes the current situation very accurately. As the article explains: the public SDK is only for java development. The intresting thing with having a Linux phone is to develop native applications. There is no public SDK from Motorola for native applications. That is the problem.
quite big in Japan.
No, wait, the joke is on you; Big in Japan was one of the big hits of the German synthpop/-rock music trio Alphaville.
Of course, in the days before discount cards they would have given those discounts to everyone and called them "sales". So, actually, they are giving out fewer discounts than they would have and getting marketing info for free.
This isn't strictly true: in theory the market info they get from the discount cards allows for tageted advertizing, which increase their sales. A procentage of the extra income goes towards extra sales. => Some of the sales given to people with discount cards wouldn't have been possible if there had not been discount cards.
But then, how big this effect is in practice can probably be discussed...
From the site's FAQ:
The following things are paranormal by definition:
[...] Violations of Newton's Laws of Motion [...]
Eh?... If this is the case, then even I belive in the paranormal... Looks like Einstein has some cash to withdraw here.
How do they look on discoveries of exotic particles? Seems like there might be some easy money here afterall...
As the voting form requires to vote for all categories it is not a good thing to do this if you have no clue who all these people are.
So, can someone with insight please line up the best patent-antanagonist choices in each categories (for extra credit: include motivations), so the crowd can wild...
... an "identify all the test cases" exercise with a simple problem: a function which classifies 3 side lengths with whichever type of triangle they could form.
... MAX_INT ...
Eh... I've read your story and are left wondering; are you sure the exercise function used *integer* parameters rather than reals/floats? Few sets of 3 integers will form a triangle, and you'd have to implement a test for integer triangles. This seems to be a bit much for an excercise about limiting cases...
That said, if the exercise used reals, I suppose the solution still should have tested for overfolows; like checking for NaN:s and INFs.
If, by some miracle some company produces more than 9gb of content for their game, it isn't exactly difficult to put it onto a second disc (recall FFVII). As long as the disc doesn't have to be swapped every 3 hours it isn't a big deal.
I wonder if they will allow disc-swaps. Previous xbox hard-reset on eject. If this xbox don't do that, this could be a great opporunity for hackers...
Anyone want to explain the real world applications of this[?]
:) :
You leave us guessing at your background, but in case you are an "xbox hacker", here is an explanation for you
Let say "ax" and "by" happens to be colliding in SHA1. (They are not, but just pretend)
Lets say that an "insider" working on an xbox game on purpose put this into the code:
if(!strcmp("ax" == "by")) {
startup linux or something
} else {
run computer game
}
Lets say Microsoft signed that game.
THEN the "evil linux hackers" could take this game and exchange "by" for "ax" inside the executable. The executable would *still* confirm to Microsofts signature; the SHA1 would not have changed.
Result: Microsoft signed executable starting linux.
But --- I guess this isn't really that different from an "insider" inserting some other backdoor... So, its not that useful as an attack I guess...
It is the same document, just relying on differences in the document name (it appears) to generate the different pages.
No, you have missed the point. Go back and rtfa again. The attack still works if you rename the documents to the same filename.
The difference lies in a generated "binary cookie" in the beginning of the postscript documents. This "cookie" makes the postscript intepreter either select to show document 'A' or 'B'. The "thing" with the cookies are that they are carefully selected to be md5-colliding. Result: both documents have the same md5sum.
You can change the rest of the documents freely if you make the same changes in both documents. The md5sum will change, but it will still be the same for both documents.
So. No. It is indeed a md5 collission attack.
OTOH, in the real world, one would be hard pressed to change the contents of an existing document.
You make it sound like the attack isn't feasible. I belive that is far from the truth in the *actual* real world.
All you need is to pull off some lame excuse: "Oh, they needed it in this format -- I had to convert your letter of recomendation to postscript; can you please sign this version too?".
Combine 1) This attack, 2) A little social engineering, 3) Someone who does not know about the recent discoveries of md5 collisions; and you have all you need for a successful attack.
The biggest problem I see is that this attack is done in *postscript* => something that your average Windows user cannot read at all. *That* will make your boss suspicious to sign it. Now, if it was 'pdf' on the other hand...
No, you and others who have replied to my post still don't seem to get it!
You open your favorite editor that don't destroy binary data (emacs is fine). You open the two postscript files. You will see how each contains a different "header", but after that *both* files include *both* messages(!)
Now, the headers are different and uses some postscript trickery to select which of the messages is going to be shown on screen, but yet "magically" manages to "preserve" the md5sum.
The rest of the two documents can be edited at will. You can for example change the word "Orders" to "Obligation to pay" if you wish. You just have to change this in *both* files. In file 'A' where it is not actually shown on screen, and in file 'B' where it is shown on screen.
Then check the md5sum's of the files. The md5sum will have *changed* relative to the old md5sum, but they will still be *exactly equal* between the two files.
It works. I have tested it. Please test it before you comment to say that it does not work.
What I haven't seen mentioned yet, and people perhaps haven't realized, is that in providing these two postscript files, they have essentially provided you with an postscript signature exploit kit!
:)...
All you need to do is download the two postscript documents and do *exactly corresponding edits* in both of them, and you get two documents saying different things and still have the same md5sums!
I just tried exchanging Alice's name for my own, and surely it did work.
Now, if they released a pdf-file hack, I would be genuinely worried
Rolling distros are perfect for boys living at their parent's basament which think they have a life because they know what the last version of KDE or Gnome or whatever program is. Heck, adult people even haven't a "favourite program" to desire to upgrade as soon as its next version publishes!
Being in the position of part-time managing computers at an academic workplace I can state for a fact that what you say simply is not true. In this environment there are working adult people that sometimes have use for bleeding-edge features to get their work done. Contrary to what you think, the world just isn't divided into 1) grandmothers who are happy with a windows machine where wordpad and solitare works, 2) Kids who live in their parents basement.
What about people who happen to need a new mathematical function from the latest GNU scientific library? Should they wait for the next distro release? Of course -- they COULD spend the half day downloading the source and compiling it themselves -- but why waste time when you can be part of a rolling distro? Just install testing, and pull packages from unstable or experimental when needed.
To give another example: the other day I was messing around with the drawing program Inkscape for some figures to be used in an academic work. We wanted those figures exported as pdf so they could be used with pdfLatex. This "simple" task turned out to require the very latest beta-releases of both Inkscape and the desktop publishing program Scribus! What are your suggestion? Redo uglier figures in a simpler drawing program until our distro release a newer version?
"Instead, let some kind of specialized "release team" take care of making debian stable releases at whatever rate they feel apropriate simply by freezing the moving distribution and working on it until it is stable enough to be released as a new "debian stable"
Well, it seem you don't know what you talk about either! what do you think is the way Sarge is making its way into new Stable?
The way I described it is just not how it works today. Packages have for a longer time been frozen in testing/unstable while waiting for the sarge release to happen. My POINT was to suggest the moving relase to keep moving, with stable releases basically happening through:
1. Take a snapshot of the moving release.
2. Apply bugfixes and backports to the packages until they are stable enough, while the moving distro KEEP MOVING.
3. Release "stable".
Again, this is not how it works today.
The stable distribution has a quite slow release cycle, but there is no reason for a desktop user to run the stable distribution. You can run either the unstable distribution, that regardless of its name is quite stable, or you can run the testing distribution.
There is a sad reason to not run testing: the testing distribution is the last one to get security updates; as I have understood this, unsecure packages from unstable can overwrite security fixed packages in testing. And I've seen people at debian-devel rave about how the reason for this is that "testing/unstable is not really meant to be used like this."...
But I wish the debian community would unite behind the picture you have, and with testing as the recommended desktop distribution; becuase then the focus would be on what debian does best of all distributions: to provide a rocking *moving* desktop distribution.
For a broadband user, a moving desktop distribution makes a lot of sense. You don't have to wait for the next release (as with Ubuntu, Fedora) to get the next version of your favourite program -- instead your desktop will be continously updated as new versions becomes available. (With just a small time-lag due to the package dependencies etc. being before being moved to testing). And the thing is: I belive this is how most people currently are using debian! Why not make this the supported way?
If it was my decision (which it clearly isn't), I would work for making the moving Linux distribution provided by testing, unstable, and experimental fully independent from the concept of "releases". Instead, let some kind of specialized "release team" take care of making debian stable releases at whatever rate they feel apropriate simply by freezing the moving distribution and working on it until it is stable enough to be released as a new "debian stable".
You couldn't go back in time and shoot Hilter before he got into power for the very simple reason that it didn't happen.
:)
No, no, no; you got it all wrong. It was just exactly because someone went back and shot that Hilter you speak of; that the much worse dictator Hitler we do remember could come to power.
The sad thing is that the built-in tv-tuner only supports NTSC, so this product is obviously not targeted for europe...
spyware almost always hides its true intentions deeply into some EULA nobody reads
spyware usually is very hard to uninstall
Wouldn't this qualify directX?
< ducks >
This is a total misconception of how evolution works. [...] The mutation described in the article is clearly a coincidental mutation that came about much earlier, not one that evolved due to the current 25-year-old aids epidemic.
But isn't this exactly how evolution works? There is never any "active adaption", but mutations happens randomly and may by pure luck give a beneficial property. These benefits makes the survival rate of the children higher, and after a looooong time decendants of a few mutated individuals have replaced the entire species.
Notice that the "looooong time" in this is the time it takes for the gene to spread to the whole population, not the time it takes from an outbreak of a disease until it is possible to find a gene that protects against that disease. As you say is "clear" here, the latter probably usually happens long *before* the former.
I think by "rejected" the parent means there is not an example of a single society ever having existed that didn't have a form of government or social hierarchy. Even the anarchist communes of the 60's and 70's ended up with a ruling class (and often started out with one). It's an inevitable consequence of human psyche that you have government.
With this way of seeing things I guess one could say that we still have anarchy everywhere in the world today. Only that given anarchy, people have chosen to group, come up with rules, and violently enforce them upon others. But in an environment with a total absence of rules, people should be allowed to do that too.
Given how people are very willing to group in "governments" when left to their own devices, I guess for anarchy to really mean "absence of rules" one would have to create a system that enforces "no rules". Like, having a rule that say "nobody is allowed to set up any rules." But, but < head explodes... >