Most OS API's are C or C++, so they already do reference counting to manage their objects
How using C or C++ means using refcounting? Ans yes, refocunting is pretty good and efficient memory management strategy, but has its own problems, like circular links, etc. Garbage collection memory managers, like Java, don't have these problems (they have their own:)
they've certainly demonstrated that ability well by relying not only on linux, but also on x86, and X11.. doesn't seem very independant,
That says exactly nothing. They had to start from somewhere, right? And if their inner workings are not dependant by OS (I don't know if they are or not - they are only source of info on this) they could do it on the other OSes later. Though facts are I have yet to see achieving a true platform-independancy without a lot of ugly hacks on the way and a ton of compromises (like, the core works on Linux, Solaris and Win32 but multimedia is Windows-only and this funky module works only for Linux, etc.)
You know, stone axes were once "most successful, most widely deployed techology". That not saves them from being totally unaproppriate for the modern world, from any point of view except being displayed at historical museum. HTML was fine for academical knowledge-exchange networks. It's not appropriate for data-representation or webpage-building of today. And quality of current HTML (go find how many of them validates as HTML 4.0?) and nasty tricks you have to do to get something useful of it (browser detection, etc) is the best evidence of this.
And yes, quantum physics of today won't exist without alchemy too. But it's not the reason for quest of the philosopher's stone in 21th century. This quest was a total failure, though brought very significant advances on the way. The case of HTML is alike - it didn't reach it's purpose, but brought a lot of useful things on the way. But the time has come to bury it.
HTML in it's current state is a terrible failure. It did not serve it's purpose as content markup (what XML successfully does now), because of browser-makers' addition of non-content tags and too late introduction of CSS (if CSS was out there when HTML started, we won't have FONT tags and all that sillyness). Neither it served it's purpose as visual markup language - try to do something really smart with HTMl and you get or into browser-dependant DHTML/Javascript, or into Java or Flash.
Ih general, HTML became total failure. That's why the urge for replacement was so strong that W3C even created some useful stadards before Microsoft or some other marketing juggernaut had to come and do it as they like.
But, misuderstanding those new standards can do you a bad service, as it did to HTML. XML not an HTML replacement. XML is purely content-markup language. You can generate visual markup from XML content markup with XSLT or DSSSL and CSS, but XML itself has only logical markup, that's it. It is returning to SGML principles, enriched by the expirience of real application and real developers' needs. It adds a bit of complication by requiring you to know all those standards, but when you grok it (and believe me, it's not so hard - one can do it in a matter of days), it vbecomes clear and logical hierachy of content representation standards.
What I see there in filemanager screenshot is the same old MSexplorer panel sans scrollbar. The idea of stripping scrollbar is pretty dumb - now I couldn't do fast steps with "Directional lock". Also, an idea of notifying applications on update doesn't look good - why just not to check directory timestamps?
No signs of "3D file hieararchy" or even principles on which it will be based are shown. Mostly, since Unix file hierarchy (if we ignore links) is a planar graf (a tree), it doesn't need any 3D representation. Also, I guess it would be a PITA to work with 3D filesystem on 2D monitor. The clever use of the mouse buttons for directory tree movements is definitely cute, but I fail to see here innovative 3D design. Using mouse wheel doesn't make an application 3D, as well as creative use of bevels and shadows in window borders.
The only real 3D thing is that cube, but I'm afraid users won't like this. Just imagine your phone dialer pad in cubic shape. Would you like dialing on such a thing? Maybe there could be a trainign that will allow user to effectively use such gadget, but I fear this will be no better than plain old planar button dialer.
Now, if we had some kind of 3D control device (gloves?) and 3D display (goggles, I guess), we might have some use of such an interface. However, I don't see it (i.e., widely distributed use of gloves and goggles as primary I/O devices, replacing displays and keyboards/mouses) happening in the next 50 years.
The situation with advertisement gets really comical. From one side, advertisement is meant to sell you something, i.e., to convince you that advertised product is better than alternative. Or it meant to make you consider buying this product, i.e., to convince you it's something good.
On the other hand, most people are really annoyed with current advertisement, and are actively seeking means to get rid of tehm, either with mental measures (such as ignoring all content that looks like banner or advertisement) or passive filtering (switching channel on TV, scrolling down web page, turning off the radio, turning the page of the newspaper) or active filtering (TiVo, junkbuster proxies, Mozilla image loading settings, spam filters). So, advertisement agencies become engaged in war with the same people they have to please and make them love the advertized material. What do you think average user will think when new wave of the advertisement appears and he'll be unable to block them? "Oh, they do it so good that I must buy it?". Hell no. The reaction will be "I need better protection measures, those spammers got me again".
So, the question is - why businessmen continue to pay advertisers to wage war on consumers? Shouldn't they instead vote with their money for more consumer-friendly (and less annoying) ways of advertising? I don't believe that it's impossible to find - if people are able to make 1Ghz processors and Mars landers, why aren't they able to think out advertizement that won't be hated by the consumers? The only thing needed for that is active wallet-voting from the side of the advertisers.
I guess the lack of Hurd developers has something to do with existance of Linux. If I had to choose free OS project I wanted to join, I'd always choose working OS with millions of users against obscure OS without users, purpose or acceptance. This is not to bash Hurd, just "Linux kernel developer" in resume weights much more than "Hurd kernel developer" in the eyes of people.
Well, that's a question - what should we do. Move out, that's what. Find another ISP. If you paid in advance, sue to get money back. There are a real lot of ISPs that would not delete your files just out of the blue because they don't like them.
Actually, this is good thing, used properly. Allows to create "trusted installations", which are known to run only authorized software. Might be good for many applications, like public-access computers, security-sensitive installations, etc.
Not every idea of Microsoft is necessarily bad, y'know...
Zend will be releasing PHP IDE in coming monthes (When It Will Be Ready (TM) of cousre:). It would be Java-based, and would have an IDE and Debugger.
It would not be free, and neither Open Source, I guess.
If someone detects a Great Big Hole in the GPL, the statement "or any newer version" is the only way how this can be plugged: by issuing a better version!
No. That's the user who chooses which GPL version to use. So I can always use old version with GBH in it.
Not exactly true. This is not only my changes to your code (that would be Qt license), but entire my code that uses your code should be GPL. That means, even iif I use your code unchanged (GPL library), I still probably have to make my entire application GPL.
I agree that you can impose whatever restriction you like on your code. It just occures to me that GPL becomes much less of a "free" software once you happen to disagree with FSF on the ways of how free software should be. With providing standard license, FSF acquired significant power behind it's GPL (most people don't really check licenses when they want to release something, they just stamp "GPL" on it because that's what they know is "free"). But once you are not in the FSF ways, all this power turns on you, and software that you could use for your and people's benefit (we are still talking open source, let's say) becomes forbidden for you. Probably, if people would realize this implication of putting GPL stamp on their package, they would use some less restrictive license (provided such a standard license would be available), because they don't really against some non-GPL project using their product. But now you need to come out and ask permissions from each of these persons - which negates all the "freedom" effect for you.
Summing up, as Stallman says on GNU pages, "GPL is not a Mr. Nice Guy". It's pretty restrictive license created for certain political purpose (a good one, but still), and if you agree with this purpose - you are fine, but if you are not sure - you should understand all the implications.
Everything can be regarded as a derived work. For example, GNU pages clearly state, when explaining readline case, that I cannot use this library in non-GPL program. Even that my work is in no way related to readline, I just want to provide user with comfortable way of typing in. It's still "derived work" and it still should be GPL or nothing.
Generally, from what I heard the official position is "GPL or nothing". Meaning, you cannot mix GPL-ed software with software under non-GPL-compatible license. You can insert non-GPL code into GPL project, but not vice-versa. Needless to say, it's very hurting to the open source projects that for various reasons can not or want not to use GPL. Needless to say, that's not exactly what you might think about "free software" term when you first hear it - it's more of "free software iff you understand freedom in the exactly same way as FSF does".
But this means that copyright holder will control not only how I copy/distribute software, but actually how do I run it. Isn't this exactly the case of DVD where so much fighting was? Isn't this exactly what liberty advocates is struggling against when they criticize restrictive EULA's? Now comes Stallman and says "Guys, let's make control on program runs common and OK". I don't really like it, and I don't consider it to be a wise move - it will hurt people's freedom much more than improve it. Now every corporate lawyer can refer to Stallman when he assures public that the program usage is copyright-controlled and is out of the filed of the fair use.
To take improvements back - means something along the lines of Qt license. Qt license is GPL-incompatible, means you cannot touch any GPL code in your works.
I don't see here any "diminishing of the scope". Basically, if you want to use GPLed code (and Stallman regards "use" as "do anything with it that is of use to your application"), your code must be GPL (or GPL-compatible license) too. No other open-source license applies, let alone closed-source. So GPL is just as restrictive as Microsoft EULA, only it benefits the other side of the deal. But to say it's "more free" on every front would be plain wrong - it's more free on hacker side, but less free on business side. For hacker it's good, for business - well, it not so good as if GPL was not so restrictive, but stil better that common commerical license.
Microsoft reserves the right to terminate any user
Now that's scary. I won't object termination of the license agreement, but sending death-squad to the user who peeked into M$ code seems to me a bit like over-reaction...
Slashdot moderation doesn't cease to surprise me. The jms guy got +4 Informative for stating what is stated at least three times in the previous posts, and does stupid mistake even in those 2 lines, and the he gets +3 Informative for correcting that stupid mistake. D'oh.
I also had engaged in a conversation with Mr. Stallman on a subject of different non-GPL open source software, and my impression was that he considers any non-GPL software as evil, and every commercial software as deadly sin. While appreciating integrity of his position, it appears to me too far from the Real World as we know it. Though, he appears to be in this point a politician, meaning it's not too useful to ask him about something if your primary goal is to get the work done, but only if your primary goal is to get a "kosher certificate" from FSF/GNU/whatever. If these goals contradict, one should choose what he wants first - certificate or thing done.
Also, we should note that while their products still have a lot of bugs, "critical" (e.g. security) bug are fixed pretty quick and no longer delayed until "next service pack". And their documentation is not bad too, especially compared to some other vendors. I'd say this is due to public pressure Microsoft is under last years. Every "bad thing" they do is publicized quickly, while their competitors can do thousand of "bad things" and still go relatively unnoticed. This also means we should continue pressing them in hope they'll do better and sometimes Windows platform will be soo good that I could work on it again:)
Nothing is too silly for a legal system. Legal system works by the law, not by "common sense", and this is probably good because some 90% of earth population is lacking that sense completely and the remaining 10% have very different view on what it is.
You won't blame your computer when it does something stupid - you don't blame legal system for it. You blame the law (or lack of it) which made it to work this way. Your americans should call (or write) your government representative and make the law change, not blame some poor guy at the patent office because he doesn't know is the "one click" obvious or not - most probably for him it's just half of "two click", whatever that be.
I'm almost sure now that there's some very pro-Linux persons in Microsoft management. Now, all world knows that Linux has working NTFS, and that even Microsoft recognizes it works so good that is scares them. Now think, how much would it require to get this recognition by conventional means? And with stupid legal threat (which is in fact very clever trick from deeply hidden Linux agent) this is done in days. Good work, Microsoft people.
Slashdot Mirror
Most OS API's are C or C++, so they already do reference counting to manage their objects
:)
How using C or C++ means using refcounting? Ans yes, refocunting is pretty good and efficient memory management strategy, but has its own problems, like circular links, etc. Garbage collection memory managers, like Java, don't have these problems (they have their own
they've certainly demonstrated that ability well by relying not only on linux, but also on x86, and X11.. doesn't seem very independant,
That says exactly nothing. They had to start from somewhere, right? And if their inner workings are not dependant by OS (I don't know if they are or not - they are only source of info on this) they could do it on the other OSes later. Though facts are I have yet to see achieving a true platform-independancy without a lot of ugly hacks on the way and a ton of compromises (like, the core works on Linux, Solaris and Win32 but multimedia is Windows-only and this funky module works only for Linux, etc.)
Well, why not if it's useful. World is large enough to have competitors even for allmighty GNOME.
You know, stone axes were once "most successful, most widely deployed techology". That not saves them from being totally unaproppriate for the modern world, from any point of view except being displayed at historical museum. HTML was fine for academical knowledge-exchange networks. It's not appropriate for data-representation or webpage-building of today. And quality of current HTML (go find how many of them validates as HTML 4.0?) and nasty tricks you have to do to get something useful of it (browser detection, etc) is the best evidence of this.
And yes, quantum physics of today won't exist without alchemy too. But it's not the reason for quest of the philosopher's stone in 21th century. This quest was a total failure, though brought very significant advances on the way. The case of HTML is alike - it didn't reach it's purpose, but brought a lot of useful things on the way. But the time has come to bury it.
HTML in it's current state is a terrible failure. It did not serve it's purpose as content markup (what XML successfully does now), because of browser-makers' addition of non-content tags and too late introduction of CSS (if CSS was out there when HTML started, we won't have FONT tags and all that sillyness). Neither it served it's purpose as visual markup language - try to do something really smart with HTMl and you get or into browser-dependant DHTML/Javascript, or into Java or Flash.
Ih general, HTML became total failure. That's why the urge for replacement was so strong that W3C even created some useful stadards before Microsoft or some other marketing juggernaut had to come and do it as they like.
But, misuderstanding those new standards can do you a bad service, as it did to HTML. XML not an HTML replacement. XML is purely content-markup language. You can generate visual markup from XML content markup with XSLT or DSSSL and CSS, but XML itself has only logical markup, that's it. It is returning to SGML principles, enriched by the expirience of real application and real developers' needs. It adds a bit of complication by requiring you to know all those standards, but when you grok it (and believe me, it's not so hard - one can do it in a matter of days), it vbecomes clear and logical hierachy of content representation standards.
What I see there in filemanager screenshot is the same old MSexplorer panel sans scrollbar. The idea of stripping scrollbar is pretty dumb - now I couldn't do fast steps with "Directional lock". Also, an idea of notifying applications on update doesn't look good - why just not to check directory timestamps?
No signs of "3D file hieararchy" or even principles on which it will be based are shown. Mostly, since Unix file hierarchy (if we ignore links) is a planar graf (a tree), it doesn't need any 3D representation. Also, I guess it would be a PITA to work with 3D filesystem on 2D monitor. The clever use of the mouse buttons for directory tree movements is definitely cute, but I fail to see here innovative 3D design. Using mouse wheel doesn't make an application 3D, as well as creative use of bevels and shadows in window borders.
The only real 3D thing is that cube, but I'm afraid users won't like this. Just imagine your phone dialer pad in cubic shape. Would you like dialing on such a thing? Maybe there could be a trainign that will allow user to effectively use such gadget, but I fear this will be no better than plain old planar button dialer.
Now, if we had some kind of 3D control device (gloves?) and 3D display (goggles, I guess), we might have some use of such an interface. However, I don't see it (i.e., widely distributed use of gloves and goggles as primary I/O devices, replacing displays and keyboards/mouses) happening in the next 50 years.
The situation with advertisement gets really comical. From one side, advertisement is meant to sell you something, i.e., to convince you that advertised product is better than alternative. Or it meant to make you consider buying this product, i.e., to convince you it's something good.
On the other hand, most people are really annoyed with current advertisement, and are actively seeking means to get rid of tehm, either with mental measures (such as ignoring all content that looks like banner or advertisement) or passive filtering (switching channel on TV, scrolling down web page, turning off the radio, turning the page of the newspaper) or active filtering (TiVo, junkbuster proxies, Mozilla image loading settings, spam filters). So, advertisement agencies become engaged in war with the same people they have to please and make them love the advertized material. What do you think average user will think when new wave of the advertisement appears and he'll be unable to block them? "Oh, they do it so good that I must buy it?". Hell no. The reaction will be "I need better protection measures, those spammers got me again".
So, the question is - why businessmen continue to pay advertisers to wage war on consumers? Shouldn't they instead vote with their money for more consumer-friendly (and less annoying) ways of advertising? I don't believe that it's impossible to find - if people are able to make 1Ghz processors and Mars landers, why aren't they able to think out advertizement that won't be hated by the consumers? The only thing needed for that is active wallet-voting from the side of the advertisers.
I guess the lack of Hurd developers has something to do with existance of Linux. If I had to choose free OS project I wanted to join, I'd always choose working OS with millions of users against obscure OS without users, purpose or acceptance. This is not to bash Hurd, just "Linux kernel developer" in resume weights much more than "Hurd kernel developer" in the eyes of people.
Well, that's a question - what should we do. Move out, that's what. Find another ISP. If you paid in advance, sue to get money back. There are a real lot of ISPs that would not delete your files just out of the blue because they don't like them.
Actually, this is good thing, used properly. Allows to create "trusted installations", which are known to run only authorized software. Might be good for many applications, like public-access computers, security-sensitive installations, etc.
Not every idea of Microsoft is necessarily bad, y'know...
Zend will be releasing PHP IDE in coming monthes (When It Will Be Ready (TM) of cousre :). It would be Java-based, and would have an IDE and Debugger.
It would not be free, and neither Open Source, I guess.
I was talking about the "old" Qt license. The QPL version 1.
If someone detects a Great Big Hole in the GPL, the statement "or any newer version" is the only way how this can be plugged: by issuing a better version!
No. That's the user who chooses which GPL version to use. So I can always use old version with GBH in it.
Not exactly true. This is not only my changes to your code (that would be Qt license), but entire my code that uses your code should be GPL. That means, even iif I use your code unchanged (GPL library), I still probably have to make my entire application GPL.
I agree that you can impose whatever restriction you like on your code. It just occures to me that GPL becomes much less of a "free" software once you happen to disagree with FSF on the ways of how free software should be. With providing standard license, FSF acquired significant power behind it's GPL (most people don't really check licenses when they want to release something, they just stamp "GPL" on it because that's what they know is "free"). But once you are not in the FSF ways, all this power turns on you, and software that you could use for your and people's benefit (we are still talking open source, let's say) becomes forbidden for you. Probably, if people would realize this implication of putting GPL stamp on their package, they would use some less restrictive license (provided such a standard license would be available), because they don't really against some non-GPL project using their product. But now you need to come out and ask permissions from each of these persons - which negates all the "freedom" effect for you.
Summing up, as Stallman says on GNU pages, "GPL is not a Mr. Nice Guy". It's pretty restrictive license created for certain political purpose (a good one, but still), and if you agree with this purpose - you are fine, but if you are not sure - you should understand all the implications.
Everything can be regarded as a derived work. For example, GNU pages clearly state, when explaining readline case, that I cannot use this library in non-GPL program. Even that my work is in no way related to readline, I just want to provide user with comfortable way of typing in. It's still "derived work" and it still should be GPL or nothing.
Generally, from what I heard the official position is "GPL or nothing". Meaning, you cannot mix GPL-ed software with software under non-GPL-compatible license. You can insert non-GPL code into GPL project, but not vice-versa. Needless to say, it's very hurting to the open source projects that for various reasons can not or want not to use GPL. Needless to say, that's not exactly what you might think about "free software" term when you first hear it - it's more of "free software iff you understand freedom in the exactly same way as FSF does".
But this means that copyright holder will control not only how I copy/distribute software, but actually how do I run it. Isn't this exactly the case of DVD where so much fighting was? Isn't this exactly what liberty advocates is struggling against when they criticize restrictive EULA's? Now comes Stallman and says "Guys, let's make control on program runs common and OK". I don't really like it, and I don't consider it to be a wise move - it will hurt people's freedom much more than improve it. Now every corporate lawyer can refer to Stallman when he assures public that the program usage is copyright-controlled and is out of the filed of the fair use.
To take improvements back - means something along the lines of Qt license. Qt license is GPL-incompatible, means you cannot touch any GPL code in your works.
I don't see here any "diminishing of the scope". Basically, if you want to use GPLed code (and Stallman regards "use" as "do anything with it that is of use to your application"), your code must be GPL (or GPL-compatible license) too. No other open-source license applies, let alone closed-source. So GPL is just as restrictive as Microsoft EULA, only it benefits the other side of the deal. But to say it's "more free" on every front would be plain wrong - it's more free on hacker side, but less free on business side. For hacker it's good, for business - well, it not so good as if GPL was not so restrictive, but stil better that common commerical license.
Microsoft reserves the right to terminate any user
Now that's scary. I won't object termination of the license agreement, but sending death-squad to the user who peeked into M$ code seems to me a bit like over-reaction...
Slashdot moderation doesn't cease to surprise me. The jms guy got +4 Informative for stating what is stated at least three times in the previous posts, and does stupid mistake even in those 2 lines, and the he gets +3 Informative for correcting that stupid mistake. D'oh.
I also had engaged in a conversation with Mr. Stallman on a subject of different non-GPL open source software, and my impression was that he considers any non-GPL software as evil, and every commercial software as deadly sin. While appreciating integrity of his position, it appears to me too far from the Real World as we know it. Though, he appears to be in this point a politician, meaning it's not too useful to ask him about something if your primary goal is to get the work done, but only if your primary goal is to get a "kosher certificate" from FSF/GNU/whatever. If these goals contradict, one should choose what he wants first - certificate or thing done.
Also, we should note that while their products still have a lot of bugs, "critical" (e.g. security) bug are fixed pretty quick and no longer delayed until "next service pack". And their documentation is not bad too, especially compared to some other vendors. I'd say this is due to public pressure Microsoft is under last years. Every "bad thing" they do is publicized quickly, while their competitors can do thousand of "bad things" and still go relatively unnoticed. This also means we should continue pressing them in hope they'll do better and sometimes Windows platform will be soo good that I could work on it again :)
Nothing is too silly for a legal system. Legal system works by the law, not by "common sense", and this is probably good because some 90% of earth population is lacking that sense completely and the remaining 10% have very different view on what it is.
You won't blame your computer when it does something stupid - you don't blame legal system for it. You blame the law (or lack of it) which made it to work this way. Your americans should call (or write) your government representative and make the law change, not blame some poor guy at the patent office because he doesn't know is the "one click" obvious or not - most probably for him it's just half of "two click", whatever that be.
I'm almost sure now that there's some very pro-Linux persons in Microsoft management. Now, all world knows that Linux has working NTFS, and that even Microsoft recognizes it works so good that is scares them. Now think, how much would it require to get this recognition by conventional means? And with stupid legal threat (which is in fact very clever trick from deeply hidden Linux agent) this is done in days. Good work, Microsoft people.