I am absolutely shocked that somebody would actually think this is true:
if you think back to the old typewriter, you have to have a Carriage Return, and a Line Feed to get to the start of the next line when typing.
Obviously you have never even seen a typewriter. On old typewriters the big silver bar on the left did both cr+lf. Electric ones had a key (where "Enter" is on your computer) that did both cr and lf. If you wanted to overprint, you did the return action, then turned the big knob on the left to basically do an "inverse-lf". If typewriters were the inspriation, we would have newline and reverse-lf characters.
It's true that early teletypes using baudot standardized on the two characters in their communication. This is because the mechanical return action was so slow that if it started doing the lf after the return started no time was lost. The lf character forced a delay to be added so the system would work, printing after a cr would never work, the next character would appear somewhere in the middle because the carriage was still returning. You had to add delay nulls to get overprinting. Believe me, at 50 baud, if they could have gotten it to return & lf in one character time they would have saved that character!
I think on early machines there was a key to generate a cr+lf pair. Also every computer system I ever saw or heard of would convert a single key into both codes internally, you never needed to type it.
Microsoft could fix their system in one day if they wanted to (just change "write as text" to be identical to "write as binary" but leave reading alone). However it is in their interest to make sure their files break when used on other systems, though almost all Unix programs have been fixed to treat CR as whitespace because of this. They also have made sure the default application you get when you double-clicik a file (textedit?) will not work for plain LF, so that Unix files look like crap. Notice that every other program they have can handle plain-LF just fine, this is pretty positive proof that they did this on purpose to make interoperability look difficult.
Also, OS/X uses LF, just like everything in the world except Microsoft.
Making each device have an individual imbedded private key is the only plausable way to make DRM work. However there seem to be a few problems with this:
1. There is concern about privacy, since the person sending you the entertainment needs your public key and can use that to track you. This can be fixed by having the device have several hundred or thousand keys so you can send a different one each time.
2. Any really secure system would be open-source so that any flaws in the encryption can be pointed out and fixed. However this means that anybody can generate "fake" public keys where they know the private key. This can be fixed by having a trusted organization record the public keys of every device manufactured by trusted manufacturers. If somebody sends a content distributor a public key not on the list they can ignore it.
3. Fair use is completly broken. You lose all your entertainment if your playback device breaks. This could be solved by requiring distributors to re-encode the data for any new device provided you send them the old copy (they can check if this is one they encrypted for pay before). But it does not help if the distributor goes out of business.
However I think the most serious problem is that Microsoft (and a few others) want to hide this solution from everybody. This is because it cuts out software completely and they lose their lock on the market.
Did not mean to claim anything I said was scientifically accurate. I was trying to report what I remember being the "common wisdom" at various times when I was in Elementary and middle school.
Oddly enough, the original poster's claim is pretty true: there is bogus science being pushed to get through an agenda, and people are willing to make "facts" up to support their arguments. My main problem is that all these people saying "all those environmentalists reversed themselves" is in fact itself a made-up fact, actaully a lie, to try to make their argument simpler. What really has happened is the *reason* the temperature may go up has changed, and also the amount has vastly decreased.
Also there was a true reversal when the "nuclear winter" hysteria happened, which even if the proponents did not claim so, convinced a lot of people that pollution would cause it to become cold. But unless I am severly mistaken, this was started by Carl Sagan when he was popular, which was certainly around 1979 to 1982, not the early 70's that the original poster is claiming.
Notheless, I distinctly remember this. "nuclear winter" was considered a radical theory precisely because it contradicted conventional wisdom about what dirt in the air would do.
It is plausible that the theory has switched more times than you think. In the 40's and 50's a "coming ice age" appears to have been popular, though not caused by humans but a natural thing (witness many science fiction stories and magazine covers illustrating humans fighting the ice with nuclear power). Possibly in the 60's (i was a little too young to remember anything then) perhaps the "obvious" answer that dirt in the air would shade the ground and make it colder was considered popular. What I do remember was that it was "cool" to know that in fact dirt in the air would *trap* heat and make the world heat up, and thus human pollution would cause the air to heat up, you were then a science geek because you understood this.
Nuclear winter, as far as I remember, appeared in my last year of high school, and thus in 1979. Exactly the same response as you came out then: "but you were all saying it would get warm!!!"
I would like you to find anything other than a single Newsweek article from the 70's in the popular literature that predicts the world getting colder. I
Where I work we have bought hundreds of thousands of dollars worth of Linux software. It's all closed source proprietary stuff with slick manuals and customer support lines and so on.
Claims that Linux users won't buy this stuff is false. They will buy it if it was available. In computer graphics it is available.
Here is the real answer:
on
Linux, Inc.
·
· Score: 2, Interesting
This was certainly considered and many, many people tried. But nothing really came of it. It now looks like the way to progress is to modify X into a modern system.
The biggest impediment was in fact the opposite of what you are claiming: using X gets you the modern hardware, and alternatives don't. Any and all advanced drivers are written for X, and the X driver interface was so complex and linked to how X worked that it was impossible to reuse them for anything other than X. Thus the very first thing any replacement for X would encounter was that it would not run in anything better that SVGA mode and would run quite slow without even rudimentary 2D acceleration. This has killed Berlin, Fresco, KGI, even kernel framebuffer interfaces, and every other alternative to X on Linux. Every one of these projects has been forced to make a "run in an X window" implementation so that they can at least write tests yet run their screens at their natural resolution, which means they will never work better than X and will all be subject to any and all glitches in X.
The truly annoying thing is that a real alterantive would probably have a much better driver interface and if it succeeded we would have many more and better drivers and hardware support than X has. Not only that, more alternatives would be easy to write because these new drivers could probably be reused easier. So it is all a catch-22.
The other lesser impediment was the need to provide compatability with X so you could run your existing applications. Ideally this should be done by an emulation library atop your new system, however Xlib is such a mess that this is incredibly difficult and will not be done by a new development. So at least initially, either X programs don't work, or they work in an incredible kludge where the hardware is shared by two low-level systems (like the OS/X solution) which makes the windows behave in annoyingly imcompatable ways that produce a worse user experience than plain X does.
I personally thought this was an excellent idea and fully expected it to happen, may favorite solution would be to return to IrixGL days, where OpenGL was used to draw every single pixel you see on the screen. But it never materialized, I followed development on a dozen systems but nothing ever came of it. It now looks like the future is in Xorg and the freedesktop standards, where they are trying, however painful it is, to modify X into a modern system, by replacing the driver level and adding new X calls to do modern graphics.
Either I had a completely different liberal school system than everybody else, or this is not true.
I grew up in the 70's. "Pop science" clearly stated that humans were going to make the world warmer. Now it did say this was due to CO2 emmissions and opaque pollution such as particulates causing the greenhouse effect. You were a science geek if you understood why clouds would make the world warmer, when it seems obvious that the shade would make it colder. I certainly remember plenty of warnings in pop culture such as "Earth Day" that our pollution would "turn the world into Venus". I remember warnings that even clean fusion energy would not save us, as even the waste heat would trip the world into heating up. I remember "pop" science fiction talking plenty about a future where civilization overheated the world, for instance Larry Niven's Ringworld had an alien civilization moving their planet further from their sun to avoid overheating.
Now either I grew up in a completely different liberal school system (this was Massachusetts public schools) or you are trying to distort history.
It's quite possible you are being confused by "nuclear winter" which came out in 1978 or so. This was initially attacked precisely because it violated conventional wisdom that "bad things humans do will make the earth hot". It is probably true that "nuclear winter" became so popular that many people did start to think that "pollution will make the world cold" but this would have been around 1982.
Now "pop science" is certainly quite wrong (saying CO2 would cause the warming is just as wrong compared to today's science as claiming cooling would happen). But I do find it very alarming that people are willing to lie about history to try to make their point. A single Newsweek article reporting on some scientists saying conventional wisdom is wrong does not change history.
On early screens the pixel actually lit up when the signal came from the computer and faded over time, exactly like a CRT, so refresh rate mattered.
I believe all modern LCD screens however have pixels that change when the signal comes in and then stay constant until the next retrace. On these the refresh rate does not affect pixels blinking at all.
"ClearType" means the new anti-aliasing algorithim being used by XP, verses "font smooothing" which is the old algorithim. The new algorithim is much more like the ones being used by Xft on Linux and by OS/X.
ClearType can do sub-pixel rendering for LCD screens, but that is not it's only function. It also does much better antialiasing than the old scheme even when limited to gray pixels.
This is a *real* problem with X window managers
on
Xfce 4.2.0 Released
·
· Score: 1
No amount of excuses. This problem exists and is caused by the window manager being a seperate program from the rest of the GUI.
Want proof: notice that you can move & resize those "skinned" MP3 and similar little programs quite well. That is because those programs draw all the pixels inside the window.
What you are seeing is the asynchronous updating of window borders and the window contents. This looks annoying and makes contents blink white or jiggle relative to the resize border. Double buffering does not fix it (it fixes blinking, which was certainly just as visible on Windows) since there is nothing to swap the border and interior double buffers at the same time. Speed is also not an issue because this has been true and constant despite a 100x speedup in the systems running X.
We need a new X window manager idea where the program is responsible for drawing everything inside the edge of the frame and handling all the events. Ie it draws the window border and handles drag & resize and raise. The "window manager" would just draw taskbars and present lists of running programs.
Yes a lot of people will panic that somehow differnt window borders will "confuse the user". This has to be solved for all the buttons inside the app anyway, so this is not a problem!
Most such bugs are in libraries that are generally available, like libjpeg.
The things that are included in an app "bundle" are things like the correct version of Qt. As far as I know there has never been a "security bug fix" of these. Instead all updates are "the new version" which requires the program to be recompiled.
In any case, if they remain shared libraries, the knowledgable user can delete the instance from the bundle and it will use the main shared one.
Even if you assumme the documents were planted by the Republicans, CBS was a bunch or morons for believing them.
And they certainly were just acting like idiots, not like the "vast left-wing conspiracy" so many believe. If they were evil they certainly would not have released the document images if even one person there said "these look fake".
Um, reading the home page, I see this in the VERY FIRST PARAGRAPH:
It is based on the portable LGPL libmspack library. cabextract supports all special features and all compression formats of Microsoft cabinet files.
But you are probably right that some useful code is in this GPL program. However I know exactly why they made it GPL, to discourage use of a proprietary Microsoft format. The fact that you were unable to make your software means the writers of this achieved exactly what they wanted to, and I am not going to cry for you, the world is much better off without your non-GPL code for reading Microsoft's stuff. Try learning how to read/write a portable format, which you will notice is all under LGPL or freer licenses.
RMS wants you to be physically or altering the art. Same as for code. This has nothing to do with copyright and certainly does not mean you can violate it.
The GPL relies on copyright, if you took Emacs and changed it and gave away your version without source code, you are violating the copyright (the GPL is really just a grant that says "you can violate my copyright if you follow these rules"). If it were not for copyright, the GPL would be meaningless.
So the ability to modify Emacs obviously then does not mean you are free to violate the copyright on it. Same thing with the ability to modify the art in the game, which pretty much means that the source code to the game is available.
I agree that this was RMS being an ass, but it is over now, just like "hacker". Though it sure seemed like "open source" was a better term, or "free source", or something.
You have to call your stuff "freeware" or "free (as in beer)" or "it's free" any of the other terms that have been used to disambiguate this. Be warned that this will make everybody think your stuff is loaded with viruses.
Okay, name the GPL code you really really wanted to use but you couldn't. Don't say "linux". NAME THE PRECISE MODULE YOU INVESTIGATED AND WERE DISAPPOINTED THAT YOU COULD NOT USE. And it had better be real GPL and not something you "accidentally" did not realize was BSD or LGPL.
If you don't do this then I will conclude that you are talking out of your ass. Tons and tons of posts from people like you who just want to say "the GPL is bad" without giving an actual example.
And don't say "readline". Everybody says that, but I have NEVER seen a commercial application use "readline" or anything like it. In case you don't know, readline is for running applications in a terminal and taking commands from stdin.
So could this be the bug, is Firefox setting the permission on the files it creates wrong? Could be it, in which case shame on them, especially if they are doing something other than what Windows does by default.
Like other readers here, I am confused about what Firefox could possibly be doing that is different than other programs. This could be it.
MDI style interfaces are crap. We had "tiled" in about 1982 (check out the Andrew window system) and it was rejected for overlapping windows for very good reasons.
Overlapping windows allow the windows to have arbitrary sizes. This is an incredible huge advantage, so large that it makes it almost incredible that anybody would consider anything else.
Unfortunatley modern systems are incredibly broken so that trying to get overlapping windows to work is almost impossible. This leads to compromises such as "put a tiled window manager inside a window" or even horrid things such as MDI, which is really "put an overlapping window manager inside a tiled one and put it inside a window". (note that real MDI has been pretty much rejected, most people when they say MDI mean the tiled-in-a-window api)
It also leads to huge numbers of people believing the problem is overlapping windows, not that the system prevents them from working properly.
What does the system need? Here it is:
Absolutely #1 priority: STOP RAISING THE F**KING WINDOWS ON CLICK!!!!! Hey, you idiots, my program is perfectly capable of calling window->raise() on click. My program is incapable of magically stopping that behavior. PLEASE, at least on Linux, get a CLUE! And don't tell me "it can be turned off in metacity" because I need a GUARANTEE that it is off.
Number 2 priority: when the user tries to raise a window by clicking on the title bar, just tell me, so I can raise other windows as well. Again I am quite capable of raising the window myself. Of course I can somewhat simulate the behavior I want with "child" or "transient-for" windows but it is a nightmare, with this simple change I can ignore all that, and do things like have multiple "main" windows with all the tools remaining above them.
Less important but nice:
3. Let me create the icon/taskbar entries myself, completely independent of any windows. I should be able to create one of these, and when I create any window say "this window belongs to this taskbar entry". Tell me when the user clicks on it, and I will pick what to show. Currently you are also forced to use "child" windows to avoid extra taskbar entries. (on X there is a "window group" interface, but I have never seen a window manager pay attention to this).
4. If the user tries to iconize a window, don't do anything except tell me (but perhaps turn on the icon/taskbar entry if you have some old WM design where they are missing when the window is not iconized). I can hide the window (and all the others) myself. Currently you can fake this by watching for show/hide but it is ugly.
Currently the only solution is to use "child" windows. This at least somewhat mitigates the goddamm "click to top". However you cannot make any structure other than a branch-less stack work correctly. You certainly cannot have more than one "bottom" window (such as two different paintings in photoshop). You also can't have more than one child because click-to-top still happens between them, making overlapping them useless.
These limitations have pretty much forced the "tiled window" interface. And they have also forced people into the mistaken notion that only tiled interfaces can work. This is a very bad state of affairs.
It would work for downloaded data as well directly to the home user's machine. The user would have to send their public key to the download service and it would encrypt it.
It does seem an actual working "DRM" scheme could be done, and it would work much like you say. It does have privacy problems, sure. And people have to buy new hardware. And it does not allow fair use. But it does have the useful fact that it does not violate the laws of physics to accomplish what it wants.
As I see it there are well-documented decoder chips that are parts of display devices. They are encased in plastic and well-glued directly to the display so it is impossible to extract them without destroying them. Each has a randomly-generated public/private key pair, and an interface so the public key can easily be read.
You would buy entertainment off the net. Your software would read the public key from the display and send that. This key would be used by the download service to encrypt the data. The result could not be decrypted without the private key, and only your display knows that.
To enhance privacy, the chips really ought to have a few hundred key pairs on them. One is selected at random.
The decoder chips should be completly open-source designed so there is no barrier for people to make the displays or to encode for them, and so it can be very certain that there is no math error such that it is possible to decrypt without knowing the private key.
To prevent people from making up a fake key pair, there would be services that would list all keys registered by manufacturers of the chips. If a key is not listed there the encrypters would not use it. Private individuals who don't want to pay for these services could just accept any key and assumme that since fake keys don't work for the majority of services, and get reasonably secure DRM for their own content.
Anyway you are right about that source code not being "FOSS" back then. I don't think *anybody* gave out source code where they said it was free for reuse. Everything was copyrighted.
I suspect if companies had just kept doing this, it would not have pissed off RMS and other people, and there would be no GPL-style open source now. Microsoft could have existed almost exactly as-is except they would probably, for free, gets lots of bug fixes from end users. In addition cloning Windows would be virtually impossible because Microsoft could easily sue for copyright violations, since the code would be so widely available.
Ikea really has some smart people in their advertising department. That stuff is really funny. Does anybody know if this is actually printed in English newspapers? The jokes would never work in the USA, people here are too stupid to get it.
Slashdot Mirror
I am absolutely shocked that somebody would actually think this is true:
if you think back to the old typewriter, you have to have a Carriage Return, and a Line Feed to get to the start of the next line when typing.
Obviously you have never even seen a typewriter. On old typewriters the big silver bar on the left did both cr+lf. Electric ones had a key (where "Enter" is on your computer) that did both cr and lf. If you wanted to overprint, you did the return action, then turned the big knob on the left to basically do an "inverse-lf". If typewriters were the inspriation, we would have newline and reverse-lf characters.
It's true that early teletypes using baudot standardized on the two characters in their communication. This is because the mechanical return action was so slow that if it started doing the lf after the return started no time was lost. The lf character forced a delay to be added so the system would work, printing after a cr would never work, the next character would appear somewhere in the middle because the carriage was still returning. You had to add delay nulls to get overprinting. Believe me, at 50 baud, if they could have gotten it to return & lf in one character time they would have saved that character!
I think on early machines there was a key to generate a cr+lf pair. Also every computer system I ever saw or heard of would convert a single key into both codes internally, you never needed to type it.
Microsoft could fix their system in one day if they wanted to (just change "write as text" to be identical to "write as binary" but leave reading alone). However it is in their interest to make sure their files break when used on other systems, though almost all Unix programs have been fixed to treat CR as whitespace because of this. They also have made sure the default application you get when you double-clicik a file (textedit?) will not work for plain LF, so that Unix files look like crap. Notice that every other program they have can handle plain-LF just fine, this is pretty positive proof that they did this on purpose to make interoperability look difficult.
Also, OS/X uses LF, just like everything in the world except Microsoft.
Making each device have an individual imbedded private key is the only plausable way to make DRM work. However there seem to be a few problems with this:
1. There is concern about privacy, since the person sending you the entertainment needs your public key and can use that to track you. This can be fixed by having the device have several hundred or thousand keys so you can send a different one each time.
2. Any really secure system would be open-source so that any flaws in the encryption can be pointed out and fixed. However this means that anybody can generate "fake" public keys where they know the private key. This can be fixed by having a trusted organization record the public keys of every device manufactured by trusted manufacturers. If somebody sends a content distributor a public key not on the list they can ignore it.
3. Fair use is completly broken. You lose all your entertainment if your playback device breaks. This could be solved by requiring distributors to re-encode the data for any new device provided you send them the old copy (they can check if this is one they encrypted for pay before). But it does not help if the distributor goes out of business.
However I think the most serious problem is that Microsoft (and a few others) want to hide this solution from everybody. This is because it cuts out software completely and they lose their lock on the market.
Did not mean to claim anything I said was scientifically accurate. I was trying to report what I remember being the "common wisdom" at various times when I was in Elementary and middle school.
Oddly enough, the original poster's claim is pretty true: there is bogus science being pushed to get through an agenda, and people are willing to make "facts" up to support their arguments. My main problem is that all these people saying "all those environmentalists reversed themselves" is in fact itself a made-up fact, actaully a lie, to try to make their argument simpler. What really has happened is the *reason* the temperature may go up has changed, and also the amount has vastly decreased.
Also there was a true reversal when the "nuclear winter" hysteria happened, which even if the proponents did not claim so, convinced a lot of people that pollution would cause it to become cold. But unless I am severly mistaken, this was started by Carl Sagan when he was popular, which was certainly around 1979 to 1982, not the early 70's that the original poster is claiming.
Notheless, I distinctly remember this. "nuclear winter" was considered a radical theory precisely because it contradicted conventional wisdom about what dirt in the air would do.
It is plausible that the theory has switched more times than you think. In the 40's and 50's a "coming ice age" appears to have been popular, though not caused by humans but a natural thing (witness many science fiction stories and magazine covers illustrating humans fighting the ice with nuclear power). Possibly in the 60's (i was a little too young to remember anything then) perhaps the "obvious" answer that dirt in the air would shade the ground and make it colder was considered popular. What I do remember was that it was "cool" to know that in fact dirt in the air would *trap* heat and make the world heat up, and thus human pollution would cause the air to heat up, you were then a science geek because you understood this.
Nuclear winter, as far as I remember, appeared in my last year of high school, and thus in 1979. Exactly the same response as you came out then: "but you were all saying it would get warm!!!"
I would like you to find anything other than a single Newsweek article from the 70's in the popular literature that predicts the world getting colder. I
Where I work we have bought hundreds of thousands of dollars worth of Linux software. It's all closed source proprietary stuff with slick manuals and customer support lines and so on.
Claims that Linux users won't buy this stuff is false. They will buy it if it was available. In computer graphics it is available.
This was certainly considered and many, many people tried. But nothing really came of it. It now looks like the way to progress is to modify X into a modern system.
The biggest impediment was in fact the opposite of what you are claiming: using X gets you the modern hardware, and alternatives don't. Any and all advanced drivers are written for X, and the X driver interface was so complex and linked to how X worked that it was impossible to reuse them for anything other than X. Thus the very first thing any replacement for X would encounter was that it would not run in anything better that SVGA mode and would run quite slow without even rudimentary 2D acceleration. This has killed Berlin, Fresco, KGI, even kernel framebuffer interfaces, and every other alternative to X on Linux. Every one of these projects has been forced to make a "run in an X window" implementation so that they can at least write tests yet run their screens at their natural resolution, which means they will never work better than X and will all be subject to any and all glitches in X.
The truly annoying thing is that a real alterantive would probably have a much better driver interface and if it succeeded we would have many more and better drivers and hardware support than X has. Not only that, more alternatives would be easy to write because these new drivers could probably be reused easier. So it is all a catch-22.
The other lesser impediment was the need to provide compatability with X so you could run your existing applications. Ideally this should be done by an emulation library atop your new system, however Xlib is such a mess that this is incredibly difficult and will not be done by a new development. So at least initially, either X programs don't work, or they work in an incredible kludge where the hardware is shared by two low-level systems (like the OS/X solution) which makes the windows behave in annoyingly imcompatable ways that produce a worse user experience than plain X does.
I personally thought this was an excellent idea and fully expected it to happen, may favorite solution would be to return to IrixGL days, where OpenGL was used to draw every single pixel you see on the screen. But it never materialized, I followed development on a dozen systems but nothing ever came of it. It now looks like the future is in Xorg and the freedesktop standards, where they are trying, however painful it is, to modify X into a modern system, by replacing the driver level and adding new X calls to do modern graphics.
Either I had a completely different liberal school system than everybody else, or this is not true.
I grew up in the 70's. "Pop science" clearly stated that humans were going to make the world warmer. Now it did say this was due to CO2 emmissions and opaque pollution such as particulates causing the greenhouse effect. You were a science geek if you understood why clouds would make the world warmer, when it seems obvious that the shade would make it colder. I certainly remember plenty of warnings in pop culture such as "Earth Day" that our pollution would "turn the world into Venus". I remember warnings that even clean fusion energy would not save us, as even the waste heat would trip the world into heating up. I remember "pop" science fiction talking plenty about a future where civilization overheated the world, for instance Larry Niven's Ringworld had an alien civilization moving their planet further from their sun to avoid overheating.
Now either I grew up in a completely different liberal school system (this was Massachusetts public schools) or you are trying to distort history.
It's quite possible you are being confused by "nuclear winter" which came out in 1978 or so. This was initially attacked precisely because it violated conventional wisdom that "bad things humans do will make the earth hot". It is probably true that "nuclear winter" became so popular that many people did start to think that "pollution will make the world cold" but this would have been around 1982.
Now "pop science" is certainly quite wrong (saying CO2 would cause the warming is just as wrong compared to today's science as claiming cooling would happen). But I do find it very alarming that people are willing to lie about history to try to make their point. A single Newsweek article reporting on some scientists saying conventional wisdom is wrong does not change history.
On early screens the pixel actually lit up when the signal came from the computer and faded over time, exactly like a CRT, so refresh rate mattered.
I believe all modern LCD screens however have pixels that change when the signal comes in and then stay constant until the next retrace. On these the refresh rate does not affect pixels blinking at all.
"ClearType" means the new anti-aliasing algorithim being used by XP, verses "font smooothing" which is the old algorithim. The new algorithim is much more like the ones being used by Xft on Linux and by OS/X.
ClearType can do sub-pixel rendering for LCD screens, but that is not it's only function. It also does much better antialiasing than the old scheme even when limited to gray pixels.
No amount of excuses. This problem exists and is caused by the window manager being a seperate program from the rest of the GUI.
Want proof: notice that you can move & resize those "skinned" MP3 and similar little programs quite well. That is because those programs draw all the pixels inside the window.
What you are seeing is the asynchronous updating of window borders and the window contents. This looks annoying and makes contents blink white or jiggle relative to the resize border. Double buffering does not fix it (it fixes blinking, which was certainly just as visible on Windows) since there is nothing to swap the border and interior double buffers at the same time. Speed is also not an issue because this has been true and constant despite a 100x speedup in the systems running X.
We need a new X window manager idea where the program is responsible for drawing everything inside the edge of the frame and handling all the events. Ie it draws the window border and handles drag & resize and raise. The "window manager" would just draw taskbars and present lists of running programs.
Yes a lot of people will panic that somehow differnt window borders will "confuse the user". This has to be solved for all the buttons inside the app anyway, so this is not a problem!
Most such bugs are in libraries that are generally available, like libjpeg.
The things that are included in an app "bundle" are things like the correct version of Qt. As far as I know there has never been a "security bug fix" of these. Instead all updates are "the new version" which requires the program to be recompiled.
In any case, if they remain shared libraries, the knowledgable user can delete the instance from the bundle and it will use the main shared one.
Even if you assumme the documents were planted by the Republicans, CBS was a bunch or morons for believing them.
And they certainly were just acting like idiots, not like the "vast left-wing conspiracy" so many believe. If they were evil they certainly would not have released the document images if even one person there said "these look fake".
It is based on the portable LGPL libmspack library. cabextract supports all special features and all compression formats of Microsoft cabinet files.
But you are probably right that some useful code is in this GPL program. However I know exactly why they made it GPL, to discourage use of a proprietary Microsoft format. The fact that you were unable to make your software means the writers of this achieved exactly what they wanted to, and I am not going to cry for you, the world is much better off without your non-GPL code for reading Microsoft's stuff. Try learning how to read/write a portable format, which you will notice is all under LGPL or freer licenses.
RMS wants you to be physically or altering the art. Same as for code. This has nothing to do with copyright and certainly does not mean you can violate it.
The GPL relies on copyright, if you took Emacs and changed it and gave away your version without source code, you are violating the copyright (the GPL is really just a grant that says "you can violate my copyright if you follow these rules"). If it were not for copyright, the GPL would be meaningless.
So the ability to modify Emacs obviously then does not mean you are free to violate the copyright on it. Same thing with the ability to modify the art in the game, which pretty much means that the source code to the game is available.
I agree that this was RMS being an ass, but it is over now, just like "hacker". Though it sure seemed like "open source" was a better term, or "free source", or something.
You have to call your stuff "freeware" or "free (as in beer)" or "it's free" any of the other terms that have been used to disambiguate this. Be warned that this will make everybody think your stuff is loaded with viruses.
Okay, name the GPL code you really really wanted to use but you couldn't. Don't say "linux". NAME THE PRECISE MODULE YOU INVESTIGATED AND WERE DISAPPOINTED THAT YOU COULD NOT USE. And it had better be real GPL and not something you "accidentally" did not realize was BSD or LGPL.
If you don't do this then I will conclude that you are talking out of your ass. Tons and tons of posts from people like you who just want to say "the GPL is bad" without giving an actual example.
And don't say "readline". Everybody says that, but I have NEVER seen a commercial application use "readline" or anything like it. In case you don't know, readline is for running applications in a terminal and taking commands from stdin.
So could this be the bug, is Firefox setting the permission on the files it creates wrong? Could be it, in which case shame on them, especially if they are doing something other than what Windows does by default.
Like other readers here, I am confused about what Firefox could possibly be doing that is different than other programs. This could be it.
MDI style interfaces are crap. We had "tiled" in about 1982 (check out the Andrew window system) and it was rejected for overlapping windows for very good reasons.
Overlapping windows allow the windows to have arbitrary sizes. This is an incredible huge advantage, so large that it makes it almost incredible that anybody would consider anything else.
Unfortunatley modern systems are incredibly broken so that trying to get overlapping windows to work is almost impossible. This leads to compromises such as "put a tiled window manager inside a window" or even horrid things such as MDI, which is really "put an overlapping window manager inside a tiled one and put it inside a window". (note that real MDI has been pretty much rejected, most people when they say MDI mean the tiled-in-a-window api)
It also leads to huge numbers of people believing the problem is overlapping windows, not that the system prevents them from working properly.
What does the system need? Here it is:
Absolutely #1 priority: STOP RAISING THE F**KING WINDOWS ON CLICK!!!!! Hey, you idiots, my program is perfectly capable of calling window->raise() on click. My program is incapable of magically stopping that behavior. PLEASE, at least on Linux, get a CLUE! And don't tell me "it can be turned off in metacity" because I need a GUARANTEE that it is off.
Number 2 priority: when the user tries to raise a window by clicking on the title bar, just tell me, so I can raise other windows as well. Again I am quite capable of raising the window myself. Of course I can somewhat simulate the behavior I want with "child" or "transient-for" windows but it is a nightmare, with this simple change I can ignore all that, and do things like have multiple "main" windows with all the tools remaining above them.
Less important but nice:
3. Let me create the icon/taskbar entries myself, completely independent of any windows. I should be able to create one of these, and when I create any window say "this window belongs to this taskbar entry". Tell me when the user clicks on it, and I will pick what to show. Currently you are also forced to use "child" windows to avoid extra taskbar entries. (on X there is a "window group" interface, but I have never seen a window manager pay attention to this).
4. If the user tries to iconize a window, don't do anything except tell me (but perhaps turn on the icon/taskbar entry if you have some old WM design where they are missing when the window is not iconized). I can hide the window (and all the others) myself. Currently you can fake this by watching for show/hide but it is ugly.
Currently the only solution is to use "child" windows. This at least somewhat mitigates the goddamm "click to top". However you cannot make any structure other than a branch-less stack work correctly. You certainly cannot have more than one "bottom" window (such as two different paintings in photoshop). You also can't have more than one child because click-to-top still happens between them, making overlapping them useless.
These limitations have pretty much forced the "tiled window" interface. And they have also forced people into the mistaken notion that only tiled interfaces can work. This is a very bad state of affairs.
It would work for downloaded data as well directly to the home user's machine. The user would have to send their public key to the download service and it would encrypt it.
It does seem an actual working "DRM" scheme could be done, and it would work much like you say. It does have privacy problems, sure. And people have to buy new hardware. And it does not allow fair use. But it does have the useful fact that it does not violate the laws of physics to accomplish what it wants.
As I see it there are well-documented decoder chips that are parts of display devices. They are encased in plastic and well-glued directly to the display so it is impossible to extract them without destroying them. Each has a randomly-generated public/private key pair, and an interface so the public key can easily be read.
You would buy entertainment off the net. Your software would read the public key from the display and send that. This key would be used by the download service to encrypt the data. The result could not be decrypted without the private key, and only your display knows that.
To enhance privacy, the chips really ought to have a few hundred key pairs on them. One is selected at random.
The decoder chips should be completly open-source designed so there is no barrier for people to make the displays or to encode for them, and so it can be very certain that there is no math error such that it is possible to decrypt without knowing the private key.
To prevent people from making up a fake key pair, there would be services that would list all keys registered by manufacturers of the chips. If a key is not listed there the encrypters would not use it. Private individuals who don't want to pay for these services could just accept any key and assumme that since fake keys don't work for the majority of services, and get reasonably secure DRM for their own content.
The paint was stopped to save weight, not money.
Sorry.
Anyway you are right about that source code not being "FOSS" back then. I don't think *anybody* gave out source code where they said it was free for reuse. Everything was copyrighted.
I suspect if companies had just kept doing this, it would not have pissed off RMS and other people, and there would be no GPL-style open source now. Microsoft could have existed almost exactly as-is except they would probably, for free, gets lots of bug fixes from end users. In addition cloning Windows would be virtually impossible because Microsoft could easily sue for copyright violations, since the code would be so widely available.
You said "I did a number of projects on PDP 11s in the late 1970's and early 1980s and often had to buy software for it."
You did not say ".. and often software did not have source code with it". You said "...and often had to buy software for it".
Please explain why I should not conclude that you think open source / FOSS means free of cost to you.
Ikea really has some smart people in their advertising department. That stuff is really funny. Does anybody know if this is actually printed in English newspapers? The jokes would never work in the USA, people here are too stupid to get it.
I heard it is pretty lousy in "Lousiana".
So you are saying if you disagree with Communism, you must be a Marxist?
I think you missed some other possibilities!