Sorry that's actually not true; Romania offered Moldova a free step-in right after 1989 but it became obvious that the Moldavian pro-Romanian movement was vastly outnumbered by the pro-Russians.
If they wanted to unite with Romania right now, Europe and NATO would have a problem -- they don't want conflict zones within their borders, and the Transdniester Republic is one. Plus, we would really be stepping on Russian toes. Penalty for that is usually invasion, sometimes annexation and, in some rare cases, world war 3.
Please let me know how the statement "Every religion is a cult" contradicts, in any way, the last definition of the word "cult" as found in the above link.
Actually Sony did much more than that. They installed a rootkit on your PC. It was a proper rootkit, invisible to all windows tools, and it did stuff that was abusive in many ways. You do get to know about warden in advance from the eula, and also you do get to uninstall warden when you uninstall the game (actually warden doesnt even run until you log in the game, as opposed to a rootkit that starts up with your OS). Warden does not facilitate other software bypass security measures on your PC (such as antiviruses), Sony's rootkit did that too. Warden is protecting a service you are paying for _monthly_ (well, once every 2 months for game cards, once every 1, 6 or 12 months for online purchase). Sony's rootkit was preventing you from accessing digital data you had a _perpetual_ license for.
Warden is not a rootkit. Sony's rootkit was just that.
Also warden does not infringe upon anybody's GPL'd IP as far as we know.
"I don't play WOW, I don't get why people are obsessed with it"
Heh:D your first sentence is the explanation for your second. Sorry, I know it's not invalidating your points (and I agree with them) just couldn't help myself. The funny thing was not too long ago, a few months back, I was exactly in your position. I made the mistake of trying it out. I have been addicted since then. So the moral of the story is - if you wanna keep it that way (not understanding the obsession) don't change a thing (don't start playing).
There are too many things I love about WoW and I just don't want to get started, I don't want to have your social life on my consicence too. I used to have a life, you know...:D
"Some searches (very specific phrases, and I won't list any of them right now - Google knows which they are) return results with a large number of.cn (Chinese) sites."
"The.cn sites don't appear to be hosted ANYWHERE." (wow!)
"[...] the Word-Confirm on all of their sites, including the one I will have to use to post this, generate a large number of rogue responses, and the HELPDESK facilities with thousands of consoles and employees each all over the planet watch the responses and other traffic characteristics [...]"
How the HECK did _this_ get on/.? It's a new low, I swear.
The issue has been addressed already. That was an unnoticed mistake that was quickly rectified. The parts of the code that were not dual-licensed were treated separately. However there was dual-licensed code there as well, which was re-licensed as GPL. Theo took an issue with this.
"The BSD licence *was* removed in the case that started all this discussion. You cannot do that."
Yes You Can. I've explained how and why and it's silly for you to continue stating that was illegal without refuting my arguments first.
"I have read the GPL, it says you MUST give away at least as many rights as you had when you received the code. Seeing as you received the right to use either licence, it follows that you must also give your distributees the same right."
Then you have _not_ read the GPL. Because there is no such requirement in the GPL. All the GPL requires is that derivative works must be distributed, in whole, under the GPL (they use unambiguous language "to be licensed as a whole at no charge to all third parties under this License", nowhere do they say "as many right as you had when you received").
"I think that dual-licenced code doesn't actually give you the choice, this is a fallacy. In practise, you only have the right to distrbute it as the same dual-licence."
The fallacy lies in your faulty interpretation alone.
Dual licensing explained: http://en.wikipedia.org/wiki/Dual_license. My favourite quote: "When software is dual-licensed, recipients can choose which terms they want to use or distribute the software under."
Every sane person involved has accepted that dual-licensing does not mean that. Theo's the only one to contest what dual-licensing really means.
If Theo's crazy theory would be true it would throw away all dual-licensed GPL/proprietary code out there. MySQL and others would be undistributable. Are you saying I cannot operate a change on the MySQL code and redistribute under GPL?
You are wrong about one very important thing: copyright notices have _never_ been removed.
Obviously you mean permission/license notes, not copyright notices. And no, there is nothing to say that copyright notices cannot be removed, other than the license itself. Given that you don't have to follow _both_ licenses, what you just said is nothing more than a bunch of gibberish.
Each of the two licenses talks about the rights passed by virtue of said license. One license cannot possibly interfere with the other. The BSD note cannot prevent you from removing a GPL note and vice versa. Go read the GPL. There is nothing about "do not remove the BSD clause" there or about "do not distribute under GPL when code is dual-licensed".
You all Theo apologists just fail to understand what dual licensing means in the first place. It's a freaking choice! It gives you an option - distribute under license A (and _only_ license A neeeds to apply) or distribute under license B (and _only_ license B needs to apply). If one of the licenses gives you the right to redistribute (both GPL and BSD do) then you can _safely_ redistribute under _only_ that linces. No need to keep the code dual-licensed, unless the license you chose to use specifically asks for it.
The removal of copyright attribution is, afaik, forbidden by both BSD and GPL, so no, it does not let you do that as whichever license you choose you still have to follow this rule.
The BSD license cannot restrict your freedoms granted under a different license. Period. I am aware that the BSD license states that "you're not allowed to do this", but I don't _have_ to abide the BSD license when I have a choice between two licenses.
WRT "why upset the apple cart" -- are we still talking about what's legal or we've switched back to morality talk?
Tehcnicalities is what makes the difference between GPL and BSD so yes, they are very important. And the choice of license is _very_ important. Make sure you choose the right license when your distribute your own code. It will certainly affect what people do with your code.
That's only because Theo, in his latest rant, is conveniently ignoring the fact that the code was dual-licensed, even though he did acknowledge that not too long ago.
Theo's last post stated that dual-licensed code cannot be distributed with only one of the licenses still attached. BSD/GPL dual code, he says, cannot be distributed under the GPL unless you keep it BSD/GPL dual. That's a theory of it's own, which deserves some clinical therapy imho.
Theo's current post deals with the aspect of re-licensing pure BSD code. His theory is that, because the license says you can distribute/copy/modify only if you keep copyright notice and "this" notice intact, this means that you are not allowed to distribute under a different license. I will now prove him wrong.
Here's a file with the BSD clause attached. It's copyrighted to me under the BSD:
------ Copyright (c) 2007 myself Permission to use, copy, modify, and distribute this software for any purpose with or without fee is hereby granted, provided that the above copyright notice and this permission notice appear in all copies.
Some text that is supposedly copyrightable. ------
Now, say that somebody wants to GPL it. Here's how he would do it:
------ Copyright (c) 2007 myself This file is distributed under the provisions of the GPLv2 license. This file is based on work under the following permission notice: Permission to distribute this file was granted by the following BSD clause:
> Permission to use, copy, modify, and distribute this software for any
> purpose with or without fee is hereby granted, provided that the above
> copyright notice and this permission notice appear in all copies. ------
What I did was legal: I kept the copyright notice intact, I kept the BSD clause intact. This does _not_ mean that I also kept the distribution license as BSD. Compare the BSD notice with the GPL requirement:
You may convey a work based on the Program, or the modifications to produce it from the Program, in the form of source code under the terms of section 4, provided that you also meet all of these conditions:
[...]
c) You must license the entire work, as a whole, under this License to anyone who comes into possession of a copy. [...]
Notice that BSD's requirement is to keep a text attached but does not say anything about keeping the license. GPL on the other hand explicitly requires that redistribution be done under the GPL.
This is the equivalent of taking a GPL file, adding another license, then taking away the GPL because "the second license allows it". Of course if anybody did this, there would be 1000 posts in this thread instead of 100.
Just that nobody would think there was any wrondoing. If the copyright owner dual-licensed a file under both GPL and a second license, then certainly anybody wishing to use that second license could take the code, strip out the GPL and do whatever they want, as long as they obey that second lincese.
Your 4-step scenario is flawed. The dual-licensing was a choice of the developers. Thus they agreed in advance to allow what just happened. Here's the thing: DON'T DELETE SOMEONE'S LICENSE!!!"
As far as I know there is nothing but the license itself that can forbid that. Considering the files in question were not BSD but BSD/GPL dual-licensed, this prohibition was lifted.
Everybody still has that choice of license. The code hasn't gone anywhere. It's still dual-licensed, but that does not mean that it has to be distributed with both licenses still attached.
You dont _need_ to be a copyright holder. It suffices to have a redistribution license. You have two: the GPL and the BSD. If you don't want to abide one of them you can still use the other. Thus you _can_ redistribute, modifications and all.
"I think the GPL crowd would be miffed if, due to the dual licensing, the BSD crowd took it as ok to remove GPL attribution and put it somewhere as purely BSD code."
Funny thing is the BSD crowd _can_ do that, legally and without offending anyone. As long as the code is dual-licensed, the GPL part can be removed by anyone.
If the BSD people were not allowed to do this then the whole concept of dual-licensing would be in vain.
It certainly was one of the kindest we've seen from Theo lately.
But it _was_ a rant. First of all it proposes some strange legal theory I've never heard about before. Second it uses the occasion to brag about license superiority and community inferiority. "Mine is bigger than yours!" on a grand new level.
I don't really care if Theo cannot stand fools, proven or not. I do care about how he reacts _publicly_ about stuff. He can curse and swear at the GPL and Linux all he wants, in private, but when making a public statement he bears all the responsbility for it and it certainly affects his public image.
Also he seems to mix up two separate things: legality and morality. While the first one can be fairly objectively assessed in the end, the latter varies widely from person to person, from community to community. I find it strange that Theo is suprised by this difference in perceived morality considering the very difference between the GPL and the BSD communities _is_ generated by differences of opinion about what is moral.
Theo does _not_ say it's allowed. Quite the opposite: 1) he states that dual-licensed code cannot be redistributed without following both licenses and 2) he states that altering a license is prohibited. Both statements are WRONG as far as I can tell, but he certainly claims Alan Cox is pushing people to break the law.
That's a load of crap. Being dual-licensed means that I can choose to abide whichever license suits me best. This _includes_ the right to redistribute the whole code under the GPLv2, ignoring the BSD license and all it's requirements altogether.
Re:As a longtime(past tense) PHP developer I can s
on
PHP 5 in Practice
·
· Score: 1
I am sorry to disappoint you. Unfortunately I have quite vast experience with PHP, more that I should have, used in very large applications, both in number of users and in terms of complexity (code and workflow). I have been using PHP since PHP3 on and with each iteration of the package I have been waiting for the inconsistencies to go away and for the mess to get organized.
I fail to see how one can treasure the shortcomings and inconsistencies of a development language (especially give that you expect them to be addressed). Usually a mature well organized language is a much more attractive basis for a development framework.
The shortcomings of PHP fall in three distinct areas:
Language issues - these are the more fundamental complaints about the failed OOP implemementation, lack of modern mechanisms, core incosistencies. They indicate that PHP is not a good choice because of the most central choices the developers made. History tells us they will never go away. The developers simply don't want to change the way the language works and feels, most of the time citing the mantra of simplicity.
Framework issues - these include the whole set of inconsistencies, failed implementations and huge limitations in the core function set. Take a look at the number of core functions in PHP and compare that with the number of methods available in.NET's mscorlib and system assemblies alone (if we were to be fair we would have to include more assemblies in the list). By far PHP's the one with the thinner framework, yet PHP suffers from a lot of clutter and inconsistencies. Only the most recent additions actually use objects and, because of the first point, they are not very good. The PHP folks do not want to break backwards compatbility at all and this means they will have to carry all the namespace clutter and all the irreconcilable conventions with them with each version.
Native or platform, which is better? - because of the nature of the PHP language (interpreted language, not compiled) there are considerable performance benefits when using platform code. The whole core framework is written in C because of this. Yet writing extensions in C is a tedious task that bears tremendous stability risks and this can be easily observed by the very reduced functionality offered by the core framework. At the same time this means that most extension frameworks (I include PEAR here) are written natively, which brings the disadvantage of using a poorly designed (see pt1) interpreted language, thus rendering these extension frameworks slower than if platform implemented. In contrast most of the.NET framework (at least with mono) is natively implemented (even the c# compiler is written in c#).
You suggest that I should use a good MVC and give PHP a fair shot. First MVC is not necessarily what I would like to use. I find MVC to be overly simplistic. Then which of the multitude of MVC's out there will actually survive and will still be actively developed this time next year? Same goes for non-MVC web application frameworks out there. Yes, choice is a good thing, but some areas require standardization more than choice. A good web application framework is a very complex thing and learning more than one can be difficult. Having to learn a dozen is absurd.
Also a web application framework is not the only thing missing in PHP. Admittedly most so-called "web application frameworks" actually include functionality outside the scope implied by the nomenclature. Some functionality though is hard to come by. Writing a self-hosted PHP web application is something I would not recommend to anyone, even though it's quite common in Perl, Python or.NET.
Re:As a longtime(past tense) PHP developer I can s
on
PHP 5 in Practice
·
· Score: 1
Amusingly enough I have encountered the same sort of frustration that the grandparent post author has. Just never ever try to write a SOAP server using the core SOAP module:D I mean it, you will go insane in a matter of days.
Even for day-to-day tasks, because PHP makes simplicity it's creed you will run into all sorts of problems the minute you grow your project into something bigger. Most of the problems I have run into were undocumented or poorly documented "features", I'll admit, but on occasion the dent my head left in the wall was caused by pure and simple brain damage on the part of the PHP developers. The SOAP core module is a good example. PDO is another. I was once considering switching from the PEAR DB layer to PDO, given that it was advertised as a superior solution and given that PEAR in particular is a very bad piece of programming one should avoid at all costs. On paper PDO looks great. The implementation sucks. Half the time the database was returning an error the PHP process would segfault.
Consider this, many a year has passed since Apache mpm-worker has been around as a stable and reliable solution. Even to this day I am forced to rely on mpm-prefork. PHP would randomly segfault otherwise, because some of the core modules are not yet multithreading aware.
Re:As a longtime(past tense) PHP developer I can s
on
PHP 5 in Practice
·
· Score: 3, Interesting
With regard to the link you provided under "no usable object-model" - I would have to agree with the PHP developers on that one. self::$var or self::CONSTANT is supposed to bind statically. That is how self is supposed to work and actually what the bug reporter was trying to achieve is impossible in C# too AFAIK.
Namespaces are upcoming in PHP6 (again, AFAIK).
The rest of the observations are correct to the best of my knowledge and make PHP quite a horrible choice for all projects beyond a certain complexity threshold. Personally I wouldn't think of Perl as a substitute, even though I understand in many respects Perl is superior to PHP, I just find it difficult to make the right choices when it comes to picking a module for a given job. I think Perl suffers from a lot of duplicated effort, there is no concerted effort to establish a de facto framework of modules.
PHP is dying a slow death in my eyes because of all the inconsistencies. The object orientation feels more like a patch or a workaround rather than a core architectural choice. The total mess created by the ton of functions in the core is not going to go anywhere with the devs maintaining backwards compatibility indefinitely.
Sure, l7filter works well for what it was designed to do.
However, it won't allow you to mark traffic coming from, say, firefox and konqueror separately, as both are http traffic.
That's where owner matching comes into play. It's the right tool for the right job, really...
Slashdot Mirror
Sorry that's actually not true; Romania offered Moldova a free step-in right after 1989 but it became obvious that the Moldavian pro-Romanian movement was vastly outnumbered by the pro-Russians. If they wanted to unite with Romania right now, Europe and NATO would have a problem -- they don't want conflict zones within their borders, and the Transdniester Republic is one. Plus, we would really be stepping on Russian toes. Penalty for that is usually invasion, sometimes annexation and, in some rare cases, world war 3.
There are some people out there who would get sexual or some other kind of gratification out of pictures of anything, including those of banana peels.
you mean this?
Please let me know how the statement "Every religion is a cult" contradicts, in any way, the last definition of the word "cult" as found in the above link.
I see no misuse of terms.
Actually Sony did much more than that. They installed a rootkit on your PC. It was a proper rootkit, invisible to all windows tools, and it did stuff that was abusive in many ways. You do get to know about warden in advance from the eula, and also you do get to uninstall warden when you uninstall the game (actually warden doesnt even run until you log in the game, as opposed to a rootkit that starts up with your OS). Warden does not facilitate other software bypass security measures on your PC (such as antiviruses), Sony's rootkit did that too. Warden is protecting a service you are paying for _monthly_ (well, once every 2 months for game cards, once every 1, 6 or 12 months for online purchase). Sony's rootkit was preventing you from accessing digital data you had a _perpetual_ license for. Warden is not a rootkit. Sony's rootkit was just that. Also warden does not infringe upon anybody's GPL'd IP as far as we know.
"I don't play WOW, I don't get why people are obsessed with it"
:D your first sentence is the explanation for your second. Sorry, I know it's not invalidating your points (and I agree with them) just couldn't help myself. The funny thing was not too long ago, a few months back, I was exactly in your position. I made the mistake of trying it out. I have been addicted since then. So the moral of the story is - if you wanna keep it that way (not understanding the obsession) don't change a thing (don't start playing).
:D
Heh
There are too many things I love about WoW and I just don't want to get started, I don't want to have your social life on my consicence too. I used to have a life, you know...
Quotes:
.cn (Chinese) sites."
.cn sites don't appear to be hosted ANYWHERE." (wow!)
/.? It's a new low, I swear.
"Some searches (very specific phrases, and I won't list any of them right now - Google knows which they are) return results with a large number of
"The
"[...] the Word-Confirm on all of their sites, including the one I will have to use to post this, generate a large number of rogue responses, and the HELPDESK facilities with thousands of consoles and employees each all over the planet watch the responses and other traffic characteristics [...]"
How the HECK did _this_ get on
The issue has been addressed already. That was an unnoticed mistake that was quickly rectified. The parts of the code that were not dual-licensed were treated separately. However there was dual-licensed code there as well, which was re-licensed as GPL. Theo took an issue with this.
"The BSD licence *was* removed in the case that started all this discussion. You cannot do that."
Yes You Can. I've explained how and why and it's silly for you to continue stating that was illegal without refuting my arguments first.
"I have read the GPL, it says you MUST give away at least as many rights as you had when you received the code. Seeing as you received the right to use either licence, it follows that you must also give your distributees the same right."
Then you have _not_ read the GPL. Because there is no such requirement in the GPL. All the GPL requires is that derivative works must be distributed, in whole, under the GPL (they use unambiguous language "to be licensed as a whole at no charge to all third parties under this License", nowhere do they say "as many right as you had when you received").
"I think that dual-licenced code doesn't actually give you the choice, this is a fallacy. In practise, you only have the right to distrbute it as the same dual-licence."
The fallacy lies in your faulty interpretation alone.
Dual licensing explained: http://en.wikipedia.org/wiki/Dual_license. My favourite quote: "When software is dual-licensed, recipients can choose which terms they want to use or distribute the software under."
Every sane person involved has accepted that dual-licensing does not mean that. Theo's the only one to contest what dual-licensing really means.
If Theo's crazy theory would be true it would throw away all dual-licensed GPL/proprietary code out there. MySQL and others would be undistributable. Are you saying I cannot operate a change on the MySQL code and redistribute under GPL?
You are wrong about one very important thing: copyright notices have _never_ been removed.
Obviously you mean permission/license notes, not copyright notices. And no, there is nothing to say that copyright notices cannot be removed, other than the license itself. Given that you don't have to follow _both_ licenses, what you just said is nothing more than a bunch of gibberish.
Each of the two licenses talks about the rights passed by virtue of said license. One license cannot possibly interfere with the other. The BSD note cannot prevent you from removing a GPL note and vice versa. Go read the GPL. There is nothing about "do not remove the BSD clause" there or about "do not distribute under GPL when code is dual-licensed".
You all Theo apologists just fail to understand what dual licensing means in the first place. It's a freaking choice! It gives you an option - distribute under license A (and _only_ license A neeeds to apply) or distribute under license B (and _only_ license B needs to apply). If one of the licenses gives you the right to redistribute (both GPL and BSD do) then you can _safely_ redistribute under _only_ that linces. No need to keep the code dual-licensed, unless the license you chose to use specifically asks for it.
The removal of copyright attribution is, afaik, forbidden by both BSD and GPL, so no, it does not let you do that as whichever license you choose you still have to follow this rule.
The BSD license cannot restrict your freedoms granted under a different license. Period. I am aware that the BSD license states that "you're not allowed to do this", but I don't _have_ to abide the BSD license when I have a choice between two licenses.
WRT "why upset the apple cart" -- are we still talking about what's legal or we've switched back to morality talk?
Tehcnicalities is what makes the difference between GPL and BSD so yes, they are very important. And the choice of license is _very_ important. Make sure you choose the right license when your distribute your own code. It will certainly affect what people do with your code.
According to who/what?
And yes, I'm saying that what you describe is actually OK.
IANAL and this is getting into a lot of legal details, but I _think_ the "Additional Terms" section of the GPL would forbid removing that notice.
This time it's the GPL that prevents you from removing the notice, not the BSD (anymore).
I don't have a problem accepting what you just said, but then please tell me who did dual-license them? And was it a legal change?
That's only because Theo, in his latest rant, is conveniently ignoring the fact that the code was dual-licensed, even though he did acknowledge that not too long ago.
Theo's last post stated that dual-licensed code cannot be distributed with only one of the licenses still attached. BSD/GPL dual code, he says, cannot be distributed under the GPL unless you keep it BSD/GPL dual. That's a theory of it's own, which deserves some clinical therapy imho.
Theo's current post deals with the aspect of re-licensing pure BSD code. His theory is that, because the license says you can distribute/copy/modify only if you keep copyright notice and "this" notice intact, this means that you are not allowed to distribute under a different license. I will now prove him wrong.
Here's a file with the BSD clause attached. It's copyrighted to me under the BSD:
------
Copyright (c) 2007 myself
Permission to use, copy, modify, and distribute this software for any
purpose with or without fee is hereby granted, provided that the above
copyright notice and this permission notice appear in all copies.
Some text that is supposedly copyrightable.
------
Now, say that somebody wants to GPL it. Here's how he would do it:
------
Copyright (c) 2007 myself
This file is distributed under the provisions of the GPLv2 license.
This file is based on work under the following permission notice:
Permission to distribute this file was granted by the following BSD clause:
> Permission to use, copy, modify, and distribute this software for any
> purpose with or without fee is hereby granted, provided that the above
> copyright notice and this permission notice appear in all copies.
------
What I did was legal: I kept the copyright notice intact, I kept the BSD clause intact. This does _not_ mean that I also kept the distribution license as BSD. Compare the BSD notice with the GPL requirement:
You may convey a work based on the Program, or the modifications to produce it from the Program, in the form of source code under the terms of section 4, provided that you also meet all of these conditions:
[...]
c) You must license the entire work, as a whole, under this License to anyone who comes into possession of a copy. [...]
Notice that BSD's requirement is to keep a text attached but does not say anything about keeping the license.
GPL on the other hand explicitly requires that redistribution be done under the GPL.
This is the equivalent of taking a GPL file, adding another license, then taking away the GPL because "the second license allows it". Of course if anybody did this, there would be 1000 posts in this thread instead of 100.
Just that nobody would think there was any wrondoing. If the copyright owner dual-licensed a file under both GPL and a second license, then certainly anybody wishing to use that second license could take the code, strip out the GPL and do whatever they want, as long as they obey that second lincese.
Your 4-step scenario is flawed. The dual-licensing was a choice of the developers. Thus they agreed in advance to allow what just happened.
Here's the thing: DON'T DELETE SOMEONE'S LICENSE!!!"
As far as I know there is nothing but the license itself that can forbid that. Considering the files in question were not BSD but BSD/GPL dual-licensed, this prohibition was lifted.
Everybody still has that choice of license. The code hasn't gone anywhere. It's still dual-licensed, but that does not mean that it has to be distributed with both licenses still attached.
You dont _need_ to be a copyright holder. It suffices to have a redistribution license. You have two: the GPL and the BSD. If you don't want to abide one of them you can still use the other. Thus you _can_ redistribute, modifications and all.
"I think the GPL crowd would be miffed if, due to the dual licensing, the BSD crowd took it as ok to remove GPL attribution and put it somewhere as purely BSD code."
Funny thing is the BSD crowd _can_ do that, legally and without offending anyone. As long as the code is dual-licensed, the GPL part can be removed by anyone.
If the BSD people were not allowed to do this then the whole concept of dual-licensing would be in vain.
It certainly was one of the kindest we've seen from Theo lately.
But it _was_ a rant. First of all it proposes some strange legal theory I've never heard about before. Second it uses the occasion to brag about license superiority and community inferiority. "Mine is bigger than yours!" on a grand new level.
I don't really care if Theo cannot stand fools, proven or not. I do care about how he reacts _publicly_ about stuff. He can curse and swear at the GPL and Linux all he wants, in private, but when making a public statement he bears all the responsbility for it and it certainly affects his public image.
Also he seems to mix up two separate things: legality and morality. While the first one can be fairly objectively assessed in the end, the latter varies widely from person to person, from community to community. I find it strange that Theo is suprised by this difference in perceived morality considering the very difference between the GPL and the BSD communities _is_ generated by differences of opinion about what is moral.
Theo does _not_ say it's allowed. Quite the opposite: 1) he states that dual-licensed code cannot be redistributed without following both licenses and 2) he states that altering a license is prohibited. Both statements are WRONG as far as I can tell, but he certainly claims Alan Cox is pushing people to break the law.
That's a load of crap. Being dual-licensed means that I can choose to abide whichever license suits me best. This _includes_ the right to redistribute the whole code under the GPLv2, ignoring the BSD license and all it's requirements altogether.
I fail to see how one can treasure the shortcomings and inconsistencies of a development language (especially give that you expect them to be addressed). Usually a mature well organized language is a much more attractive basis for a development framework.
The shortcomings of PHP fall in three distinct areas:
- Language issues - these are the more fundamental complaints about the failed OOP implemementation, lack of modern mechanisms, core incosistencies. They indicate that PHP is not a good choice because of the most central choices the developers made. History tells us they will never go away. The developers simply don't want to change the way the language works and feels, most of the time citing the mantra of simplicity.
- Framework issues - these include the whole set of inconsistencies, failed implementations and huge limitations in the core function set. Take a look at the number of core functions in PHP and compare that with the number of methods available in
.NET's mscorlib and system assemblies alone (if we were to be fair we would have to include more assemblies in the list). By far PHP's the one with the thinner framework, yet PHP suffers from a lot of clutter and inconsistencies. Only the most recent additions actually use objects and, because of the first point, they are not very good. The PHP folks do not want to break backwards compatbility at all and this means they will have to carry all the namespace clutter and all the irreconcilable conventions with them with each version.
- Native or platform, which is better? - because of the nature of the PHP language (interpreted language, not compiled) there are considerable performance benefits when using platform code. The whole core framework is written in C because of this. Yet writing extensions in C is a tedious task that bears tremendous stability risks and this can be easily observed by the very reduced functionality offered by the core framework. At the same time this means that most extension frameworks (I include PEAR here) are written natively, which brings the disadvantage of using a poorly designed (see pt1) interpreted language, thus rendering these extension frameworks slower than if platform implemented. In contrast most of the
.NET framework (at least with mono) is natively implemented (even the c# compiler is written in c#).
You suggest that I should use a good MVC and give PHP a fair shot. First MVC is not necessarily what I would like to use. I find MVC to be overly simplistic. Then which of the multitude of MVC's out there will actually survive and will still be actively developed this time next year? Same goes for non-MVC web application frameworks out there. Yes, choice is a good thing, but some areas require standardization more than choice. A good web application framework is a very complex thing and learning more than one can be difficult. Having to learn a dozen is absurd.Also a web application framework is not the only thing missing in PHP. Admittedly most so-called "web application frameworks" actually include functionality outside the scope implied by the nomenclature. Some functionality though is hard to come by. Writing a self-hosted PHP web application is something I would not recommend to anyone, even though it's quite common in Perl, Python or
Amusingly enough I have encountered the same sort of frustration that the grandparent post author has. Just never ever try to write a SOAP server using the core SOAP module :D I mean it, you will go insane in a matter of days.
Even for day-to-day tasks, because PHP makes simplicity it's creed you will run into all sorts of problems the minute you grow your project into something bigger. Most of the problems I have run into were undocumented or poorly documented "features", I'll admit, but on occasion the dent my head left in the wall was caused by pure and simple brain damage on the part of the PHP developers. The SOAP core module is a good example. PDO is another. I was once considering switching from the PEAR DB layer to PDO, given that it was advertised as a superior solution and given that PEAR in particular is a very bad piece of programming one should avoid at all costs. On paper PDO looks great. The implementation sucks. Half the time the database was returning an error the PHP process would segfault.
Consider this, many a year has passed since Apache mpm-worker has been around as a stable and reliable solution. Even to this day I am forced to rely on mpm-prefork. PHP would randomly segfault otherwise, because some of the core modules are not yet multithreading aware.
With regard to the link you provided under "no usable object-model" - I would have to agree with the PHP developers on that one. self::$var or self::CONSTANT is supposed to bind statically. That is how self is supposed to work and actually what the bug reporter was trying to achieve is impossible in C# too AFAIK.
Namespaces are upcoming in PHP6 (again, AFAIK).
The rest of the observations are correct to the best of my knowledge and make PHP quite a horrible choice for all projects beyond a certain complexity threshold. Personally I wouldn't think of Perl as a substitute, even though I understand in many respects Perl is superior to PHP, I just find it difficult to make the right choices when it comes to picking a module for a given job. I think Perl suffers from a lot of duplicated effort, there is no concerted effort to establish a de facto framework of modules.
PHP is dying a slow death in my eyes because of all the inconsistencies. The object orientation feels more like a patch or a workaround rather than a core architectural choice. The total mess created by the ton of functions in the core is not going to go anywhere with the devs maintaining backwards compatibility indefinitely.
Sure, l7filter works well for what it was designed to do.
However, it won't allow you to mark traffic coming from, say, firefox and konqueror separately, as both are http traffic.
That's where owner matching comes into play. It's the right tool for the right job, really...