How the Swedish Pirate Party Platform Backfires on Free Software

Who cares about code that is five years old? At the rate software is changing you'd be lucky if there is anything in code that old still being used. This law would NOT place the current code under public license. Only the code that is five years old.

Maybe this is an issue for some, but probably not many.

The *vast* majority of code that makes your system work is over five year sold, in fact (assuming you're using Linux).

That is an interesting question... how do you measure the age of a code base that is constantly evolving? Using the SVN history to determine the age of each single line?

I am not a lawyer, but I guess the whole work is considered as new as the newest line in it. So if you configure your repository such that it will not allow checkouts of code older than, say, one year, and the gpl is adapted to allow refusal to provide ancient code, most of the problem goes away. It would force the proprietary software companies to regularly check out and archive copies of anything that might become useful to them in four years...

Of course, all the features that one would really like to copy into proprietary software tend to be much less than five years old. Re-fixing the security bugs found over five years also doesn't sound attractive. So it all doesn't seem to be a very big deal to me.

@Sander: I'm not a kernel programmer and haven't gone looking at the source very much but I would be surprised if there are many routines that have not been touched in some way over five years. Bug fixes, rewrites to improve efficiency, replacement of sections with newer functions. All it takes is one little change and it's not old any more.

Also, what is "the code"? As in the case of the kernel, is a little part of it? Or the whole? Change one function and the whole is no longer old.

The interpretations could be endless.

Quoting:Using the SVN history to determine the age of each single line?

~$ svn blame

Copyright is only updated on the modified segments of code, not the complete, work when it is changed.

I'm pretty sure this was revealed as part of the SCO case. In any event, that's how lawyers seem to interpret it in my experience. You'll also note that a lot of Free Software project who rip code from other projects include the original copyright notice for those functions or segments of code.

That said, I think this is a serious issue (which is why I submitted it). Stallman brings up some good points.

KS -

Not exactly.

Whenever you change a piece of code, you create a derived work, and that work is new and wonderful in it's own self. The individual parts may fall out of copyrightm but the whole will remain.