They are a corporation. This makes them evil but also utterly predictable, because their sole obligation, the purpose in everything they do, is to make money. If we make it so they make more money by embracing open source, then they will do it, never mind that they've been fighting it for years.
As someone else said, slashdot's CMS (if such it is) is open source, there's a link to the system right there on the sidebar, and it's used by other sites, IIRC plastic.com runs on it or at least did.
I don't see it as being an asshole. He's said the project will stay free and he's committed to finishing it, other than that what could be better for someone who likes.net than to work on the official implementation?
By overwriting them with your code, obviously. It's a little harder since you have to know there is a function a certain distance above you that will be called after where you are, but far from impossible.
No problem with it being a non-cvs system, but they have to release it. The kernel people made their repository publicly accessible. Anything internal apple development is using when they modify the source, has to be released to safari users.
Not necessarily. Source alone is not always the preferred form for making modifications. If they modify it based in part on the cvs logs or similar, those need to be released too.
No, it's not a gift horse, because it's LGPL rather than BSD to stop this kind of thing. Stripping the comments on the code you distribute is violation of the license, you have to distribute the original form you use to modify it. Sounds to me like refusing to give a CVS log could be too.
You have to provide the source in the preferred form for making modifications. If the cvs log is something you'd use when modifying the code - and it seems like it would be, since you need it to do regression tests - then it's part of the source and needs to be distributed.
Mm, but one thing everyone seems to be missing is that they have to offer the source to anyone they've distributed binaries to in the preferred form for making modifications. The article says apple wanted an NDA to view the cvs history, which though it may not seem like it is a part of the source - it's used for understanding and making modifications to the program, developers wouldn't work as well without it. So they have no right to insist on this to people who have a (legit) copy of safari. I'd be surprised if there wasn't at least one apple customer willing to insist on it.
There's no way they're making all those different changes together. It would collapse. Whether it's CVS and diff is unimportant, what matters is that they release the source in the form they use it. Which they don't seem to be doing.
The problem is those are just for his patches, and the codebase he's working on is significantly diverged. We need all the apple developers who made the code his patches depend on to do the same thing.
No, the license says they have to release the changes *in the preferred form for making modifications*. They have to release the code in the nicest way for modifying. Which could mean a complete separate patchset with cvs logs of what did what rather than simply the safari source. And anything the apple developers would want while working on it, such as comments about what each patch does, has to be distributed.
Nope. You have to make available the source "in the preferred form for making modifications". If that's the cvs tree, you have to give the people you distribute binaries to the cvs tree.
Yeah. Just like it's pretty fucking simple to shut up and let yourself be taxed without representation, hand over the jews you were sheltering, etc. It doesn't matter that there was an easier way to do it, he has the right to do what he did.
I don't know, but the post I was replying to implied the source was being released in a form that made it difficult to merge it back into khtml, and that apple can do that. Which they can't.
My parent claimed they were releasing the source in a way that made it deliberately difficult for their patches to be merged back into khtml. If they are doing that, it's completely against the spirit of open source, and in violation of their license.
The US used to have this thing called "fair use" where copying that would otherwise be infringing for the purpose of parody was legal. Wonder what happened to that.
WTF? It's included in Debian and every major GNU/Linux distribution, there are just as many that don't include gnome as don't include kde. The licensing issue was worked out years ago. What are you talking about?
Wait for Qt4 and kde4, then there'll be a native (no cygwin) release of kde for windows. As it is you can use konqueror on windows fine, and performance isn't too bad.
If it's because of how azureus works, how come every java application is like that? The only text editor that even approaches the delay of jedit on startup is eric3, and that takes less than half the time. (a text editor needs a splash screen because of its load time!). Yaggui is less featureful than other giFT frontends like apollon or giftui, yet takes longer to load and has horrible latency when clicking applications. I have yet to see a gui java program that performs acceptably on my system.
No, they can't. They have to release the source "in the preferred form for making modifications". If they're making it hard to use the code, they're in breach of the license.
Slashdot Mirror
They are a corporation. This makes them evil but also utterly predictable, because their sole obligation, the purpose in everything they do, is to make money. If we make it so they make more money by embracing open source, then they will do it, never mind that they've been fighting it for years.
As someone else said, slashdot's CMS (if such it is) is open source, there's a link to the system right there on the sidebar, and it's used by other sites, IIRC plastic.com runs on it or at least did.
I don't see it as being an asshole. He's said the project will stay free and he's committed to finishing it, other than that what could be better for someone who likes .net than to work on the official implementation?
By overwriting them with your code, obviously. It's a little harder since you have to know there is a function a certain distance above you that will be called after where you are, but far from impossible.
No problem with it being a non-cvs system, but they have to release it. The kernel people made their repository publicly accessible. Anything internal apple development is using when they modify the source, has to be released to safari users.
Not necessarily. Source alone is not always the preferred form for making modifications. If they modify it based in part on the cvs logs or similar, those need to be released too.
No, it's not a gift horse, because it's LGPL rather than BSD to stop this kind of thing. Stripping the comments on the code you distribute is violation of the license, you have to distribute the original form you use to modify it. Sounds to me like refusing to give a CVS log could be too.
They should look at what they themselves use when modifying the program. I'd be willing to bet they use the cvs logs themselves.
You have to provide the source in the preferred form for making modifications. If the cvs log is something you'd use when modifying the code - and it seems like it would be, since you need it to do regression tests - then it's part of the source and needs to be distributed.
Mm, but one thing everyone seems to be missing is that they have to offer the source to anyone they've distributed binaries to in the preferred form for making modifications. The article says apple wanted an NDA to view the cvs history, which though it may not seem like it is a part of the source - it's used for understanding and making modifications to the program, developers wouldn't work as well without it. So they have no right to insist on this to people who have a (legit) copy of safari. I'd be surprised if there wasn't at least one apple customer willing to insist on it.
There's no way they're making all those different changes together. It would collapse. Whether it's CVS and diff is unimportant, what matters is that they release the source in the form they use it. Which they don't seem to be doing.
The problem is those are just for his patches, and the codebase he's working on is significantly diverged. We need all the apple developers who made the code his patches depend on to do the same thing.
No, the license says they have to release the changes *in the preferred form for making modifications*. They have to release the code in the nicest way for modifying. Which could mean a complete separate patchset with cvs logs of what did what rather than simply the safari source. And anything the apple developers would want while working on it, such as comments about what each patch does, has to be distributed.
Nope. You have to make available the source "in the preferred form for making modifications". If that's the cvs tree, you have to give the people you distribute binaries to the cvs tree.
Yeah. Just like it's pretty fucking simple to shut up and let yourself be taxed without representation, hand over the jews you were sheltering, etc. It doesn't matter that there was an easier way to do it, he has the right to do what he did.
The extra M seems to be just because people wanted a new name. What's the real, non-graphical difference between Everquest and British Legends?
I don't know, but the post I was replying to implied the source was being released in a form that made it difficult to merge it back into khtml, and that apple can do that. Which they can't.
My parent claimed they were releasing the source in a way that made it deliberately difficult for their patches to be merged back into khtml. If they are doing that, it's completely against the spirit of open source, and in violation of their license.
The US used to have this thing called "fair use" where copying that would otherwise be infringing for the purpose of parody was legal. Wonder what happened to that.
No, it should be a smiley face.
Try VLC? Not included, but has a working mac port
WTF? It's included in Debian and every major GNU/Linux distribution, there are just as many that don't include gnome as don't include kde. The licensing issue was worked out years ago. What are you talking about?
Wait for Qt4 and kde4, then there'll be a native (no cygwin) release of kde for windows. As it is you can use konqueror on windows fine, and performance isn't too bad.
If it's because of how azureus works, how come every java application is like that? The only text editor that even approaches the delay of jedit on startup is eric3, and that takes less than half the time. (a text editor needs a splash screen because of its load time!). Yaggui is less featureful than other giFT frontends like apollon or giftui, yet takes longer to load and has horrible latency when clicking applications. I have yet to see a gui java program that performs acceptably on my system.
No, they can't. They have to release the source "in the preferred form for making modifications". If they're making it hard to use the code, they're in breach of the license.