"That's non-security. Make a user type his password n times a week and he'll type it in every single dialog window that asks for his password. Even the malicious ones."
Except when n == 1 or n == 2 (most of the tmes, n is much lower than 1)! You don't need to install programs or change system settings so often on Linux. Also, all the programs that ask for the root password are located on the same menu (or 2 menus) with a name like "setings" and have the clear goal of modifying the system configurations.
He said approximate solutions. Quantum computers can't solve NP-complete problems, but can apply some heuristucs in a much faster way. Also, you don't make a dent in NP-complete, if you can make a dent, you can solve it all.
Binary compatibility is very important if you want to send your program via internet and be sure it opens on the other point, doesn't matter what system it is using. But it doesn't help to port things. It doesn't really matters if you have to compile your program again, most of the time is spend modifying the behaviour of the interface, that is not adequade for the new system.
And, wile I am on it, most of Oo is written in java.
Don't generalise so much, GPU are usual nowadays, and some systems see great improvements from FPUs.
Historicaly, almost every time that someone developped some co-processor, everybody uses it* (the most clear exceptions are math and network co-processors). But a lot of times the first one to develop the co-processor can't sell it because it is done too early and is very expensive by the time.
* Or don't you use disk controllers, bus controllers, DMA, sound cards...
It is more a consequence of free software being new and little accepted than of it being out there. Inhouse programing is so expensive that companies prefer to use s**t tools that don't fill their needs than make inhouse development. To solve this, we need more free software, not less.
"OSX comes with 3-4 high quality applications that you really need (iPhoto, iTunes). What I use most often extra: OO.org (actually, NeoOffice), FireFox."
You may use it 95% of the time, but not having that tool that you need on the other 5% is also painfull.
And you are not the only one. Thinking a little more, I guess I wasn't right about this. It'll probably be usefull.
About the conventions, yes, XML is overkill (I sugested something faster), but just pairs key:value aren't enogh. Maybe, what is missing here is some type declaration, where the types define keys and their values are on the key:value pairs. That will make objects that contain other objects possible and would still be fast to parse.
Another point is that the terminal may know how to display the data if you declare its type, so the 'format' step would not be necessary.
Have you never noted the similarity with "those who don't understand history are doomed to repeat it"? UNIX people dismiss every revolutionary attempty to modify what they know because history has already showed that they are quite right about the way they do stuff (they wheren't always right, but fixed several things already).
If someone starts to create something completely new, he will make much of the mistakes that we already made after UNIX and on the earlier systems, so will come with something almost useless. If he, however, start from UNIX and improves it bit by bit, he'll probably get a better system.
"Having a data structure passed along a pipe like MSH does is a huge advantage and very efficient, but I think most UNIX people can agree that it's not worth it to bring everything into the same process. What's an alternative? Serialize the data structure (in some human readable form to stay true to UNIX tradition) and pass that down the pipe, from one process to another. That would work with the pipes we have now and the shells we have now, we just need new tools and a serialization protocol."
We would need a hole new convention on how to serialize those objects (A faster XML?). Also, it is not so clear if it would be an advantaje or if it will reduce the usefullnes of the CLI, because every use must be though at programming time (like GUI)
I think it worth trying, but I am not sure if it will be a good experience.
May be foundede by MS, but this kind of result is not so good for them, so, if it is biased I can't understand how. Discovering that the traditional approaches are becomming useless to solve Windos security problems is not something MS would pay for publishing (makes more sense if you think of this as internal research).
"Being good on a command line doesn't make you "better" or "more in tune" with a machine. It just means you are good on a command line."
It also means that you know how to automate tasks and you'll survive if something happens to X. Not that desktop users should know how to use CLI, but administrators should.
I think it is reasonable to make such exceptions, but only if you really trust a lot the marketer, and lying about the income range... Anyway, do you realy know a trustworth marketer?
The people are supposed to be inside the asteroid. Spin is a good think, but it need to be adjusted to match earth's gravity, and you can't toss the rubble so often because you are inside the thing, and have to do a great effort to send it into the center (where there is no gravty and, probably, a hole).
If your settlement is big enogth, like many thousands of kylometers wide, and taller than earth's atmosphere, and if light comes from above, you may not see the diference.
You just have more restrictions. You alread can't chose to not fall when jump from a building, or to create a hot and a cold objects from two of the same temperature without putting extra energy. If they are right you can't kill yourself on the past either, but can chose where you are going tomorow night.
Nobody has ever proved that. What happens is that everybody accepts this fact as a base, and build physics from there. If you assume that a consequence can never happen before the cause, you end up with a model where is impossible to build a time machine that goes to the past.
Debian said KDE was illegal, and suddenly, KDE's problem become public. That was a necessary step to have it fixed, as it is now. Now, Debian is making GFDL and Mozilla's problems public, maybe Mozilla will not fix it (it is not so harmfull), but you can bet that Debian move will lead to a new GFDL.
Re:Can We Get Firefox Developers To Do This, Too?
on
Hackers, Meet Microsoft
·
· Score: 2, Insightful
"Flash-forward to a couple of years ago, when Bill sent out yet another all-hands memo, pointing the company in the direction of security."
That is the problem, security can't be achieved the same way that browser market domination was. To fix security, MS will need the following:
A lot of rewritting, that is expensive. But can be done.
A lot of testing, that FOSS gets for free and MS pays a lot. But can be done.
Also, they'll need to modify the relationship they have with their customers. That is a hard one, MS will need to respect their clients. They'll need a complete reestruturation, but can be done.
And, finaly, the problem: MS will need to discontinue bad projects, breaking past compatibility.
Lets face it, Windows, IE and Office are kept on top because of the net effect. The advantage that people get when running those products is to get something that is compatible with everything else, so they don't need to care about that. If MS suddenly break past compatibility, they'll see their market suddenly vanish.
This is why MS will not develop secure products so soon, their software projects are flawed and they can't correct it. Those events are good PR, but will not make MS programs better than FOSS.
One good way is, very like you said, break it into MSWindows+IE+player, Office+visio+etc and Xbox+hardware. The other smaler stuff can go to any of them or be turned into another company (althoug, I think it should go to Windows owner).
It breaks MS into 3 lucrative companies. Two of them are monopolies, but they will not enforce each other like they do now.
You can't compare release time of Windows and Linux, they are completelly different. First, Linux comes with all the programs you should need, so freesing the release will freese all your applications. Also, the FOSS world moves much faster than the closed one. And, finaly, Linux (applications, not the kernel) was not really usable until recently, you can't compare the release cicle of a mature technology with a new one.
For the user view, Linux upgrades are free (beer) - if you by support, you pay for support, not software - while Windows ones are expensive. Linux upgraes don't generaly break running applications (altough you should still check it) and are easier to apply - you don't need to format your hard driver, or change all the system at once.
When you fill so compelled to compare both, remember that Debian Woody is near as old as WinXP (not even 2000) and was so old that it was unusable when Sarge was released.
Slashdot Mirror
The browser market is not even near mature, but MS is holding it stable since the last bowser wars.
It is a matter of education, I, for instance, am educating my parents into running Linux. Once I am done, the problem won't exist anymore...
"That's non-security. Make a user type his password n times a week and he'll type it in every single dialog window that asks for his password. Even the malicious ones."
Except when n == 1 or n == 2 (most of the tmes, n is much lower than 1)! You don't need to install programs or change system settings so often on Linux. Also, all the programs that ask for the root password are located on the same menu (or 2 menus) with a name like "setings" and have the clear goal of modifying the system configurations.
"hint: if you really want to have some fun introduce a freshly hatched MCSE to vim"
Hey, have you meet ed?
He said approximate solutions. Quantum computers can't solve NP-complete problems, but can apply some heuristucs in a much faster way. Also, you don't make a dent in NP-complete, if you can make a dent, you can solve it all.
Well, maybe is not the right answer...
"will it play Duke Nukem Forever??"
The answer is yes, and no. But you'll can never see the game without turning it into a black screen.
Binary compatibility is very important if you want to send your program via internet and be sure it opens on the other point, doesn't matter what system it is using. But it doesn't help to port things. It doesn't really matters if you have to compile your program again, most of the time is spend modifying the behaviour of the interface, that is not adequade for the new system.
And, wile I am on it, most of Oo is written in java.
Don't generalise so much, GPU are usual nowadays, and some systems see great improvements from FPUs.
Historicaly, almost every time that someone developped some co-processor, everybody uses it* (the most clear exceptions are math and network co-processors). But a lot of times the first one to develop the co-processor can't sell it because it is done too early and is very expensive by the time.
* Or don't you use disk controllers, bus controllers, DMA, sound cards...
It is more a consequence of free software being new and little accepted than of it being out there. Inhouse programing is so expensive that companies prefer to use s**t tools that don't fill their needs than make inhouse development. To solve this, we need more free software, not less.
But, again, it is Gatner...
"OSX comes with 3-4 high quality applications that you really need (iPhoto, iTunes). What I use most often extra: OO.org (actually, NeoOffice), FireFox."
You may use it 95% of the time, but not having that tool that you need on the other 5% is also painfull.
"No, I don't agree."
And you are not the only one. Thinking a little more, I guess I wasn't right about this. It'll probably be usefull.
About the conventions, yes, XML is overkill (I sugested something faster), but just pairs key:value aren't enogh. Maybe, what is missing here is some type declaration, where the types define keys and their values are on the key:value pairs. That will make objects that contain other objects possible and would still be fast to parse.
Another point is that the terminal may know how to display the data if you declare its type, so the 'format' step would not be necessary.
All this seems to be a very interesting project!
Have you never noted the similarity with "those who don't understand history are doomed to repeat it"? UNIX people dismiss every revolutionary attempty to modify what they know because history has already showed that they are quite right about the way they do stuff (they wheren't always right, but fixed several things already).
If someone starts to create something completely new, he will make much of the mistakes that we already made after UNIX and on the earlier systems, so will come with something almost useless. If he, however, start from UNIX and improves it bit by bit, he'll probably get a better system.
"Having a data structure passed along a pipe like MSH does is a huge advantage and very efficient, but I think most UNIX people can agree that it's not worth it to bring everything into the same process. What's an alternative? Serialize the data structure (in some human readable form to stay true to UNIX tradition) and pass that down the pipe, from one process to another. That would work with the pipes we have now and the shells we have now, we just need new tools and a serialization protocol."
We would need a hole new convention on how to serialize those objects (A faster XML?). Also, it is not so clear if it would be an advantaje or if it will reduce the usefullnes of the CLI, because every use must be though at programming time (like GUI)
I think it worth trying, but I am not sure if it will be a good experience.
May be foundede by MS, but this kind of result is not so good for them, so, if it is biased I can't understand how. Discovering that the traditional approaches are becomming useless to solve Windos security problems is not something MS would pay for publishing (makes more sense if you think of this as internal research).
"Being good on a command line doesn't make you "better" or "more in tune" with a machine. It just means you are good on a command line."
It also means that you know how to automate tasks and you'll survive if something happens to X. Not that desktop users should know how to use CLI, but administrators should.
I think it is reasonable to make such exceptions, but only if you really trust a lot the marketer, and lying about the income range... Anyway, do you realy know a trustworth marketer?
The people are supposed to be inside the asteroid. Spin is a good think, but it need to be adjusted to match earth's gravity, and you can't toss the rubble so often because you are inside the thing, and have to do a great effort to send it into the center (where there is no gravty and, probably, a hole).
If your settlement is big enogth, like many thousands of kylometers wide, and taller than earth's atmosphere, and if light comes from above, you may not see the diference.
You just have more restrictions. You alread can't chose to not fall when jump from a building, or to create a hot and a cold objects from two of the same temperature without putting extra energy. If they are right you can't kill yourself on the past either, but can chose where you are going tomorow night.
Yes, but remember 9/11? That could have been you...
Nobody has ever proved that. What happens is that everybody accepts this fact as a base, and build physics from there. If you assume that a consequence can never happen before the cause, you end up with a model where is impossible to build a time machine that goes to the past.
Debian said KDE was illegal, and suddenly, KDE's problem become public. That was a necessary step to have it fixed, as it is now. Now, Debian is making GFDL and Mozilla's problems public, maybe Mozilla will not fix it (it is not so harmfull), but you can bet that Debian move will lead to a new GFDL.
"Flash-forward to a couple of years ago, when Bill sent out yet another all-hands memo, pointing the company in the direction of security."
That is the problem, security can't be achieved the same way that browser market domination was. To fix security, MS will need the following:
A lot of rewritting, that is expensive. But can be done.
A lot of testing, that FOSS gets for free and MS pays a lot. But can be done.
Also, they'll need to modify the relationship they have with their customers. That is a hard one, MS will need to respect their clients. They'll need a complete reestruturation, but can be done.
And, finaly, the problem: MS will need to discontinue bad projects, breaking past compatibility.
Lets face it, Windows, IE and Office are kept on top because of the net effect. The advantage that people get when running those products is to get something that is compatible with everything else, so they don't need to care about that. If MS suddenly break past compatibility, they'll see their market suddenly vanish.
This is why MS will not develop secure products so soon, their software projects are flawed and they can't correct it. Those events are good PR, but will not make MS programs better than FOSS.
One good way is, very like you said, break it into MSWindows+IE+player, Office+visio+etc and Xbox+hardware. The other smaler stuff can go to any of them or be turned into another company (althoug, I think it should go to Windows owner).
It breaks MS into 3 lucrative companies. Two of them are monopolies, but they will not enforce each other like they do now.
You can't compare release time of Windows and Linux, they are completelly different. First, Linux comes with all the programs you should need, so freesing the release will freese all your applications. Also, the FOSS world moves much faster than the closed one. And, finaly, Linux (applications, not the kernel) was not really usable until recently, you can't compare the release cicle of a mature technology with a new one.
For the user view, Linux upgrades are free (beer) - if you by support, you pay for support, not software - while Windows ones are expensive. Linux upgraes don't generaly break running applications (altough you should still check it) and are easier to apply - you don't need to format your hard driver, or change all the system at once.
When you fill so compelled to compare both, remember that Debian Woody is near as old as WinXP (not even 2000) and was so old that it was unusable when Sarge was released.