The Modernization of Emacs: terminology buffer and keybinding

[dan at telent.net]

Ken Tilton wrote:
> Kenny happened to solve the traveling 
> salesman problem and protein-folding and passed the fricking Turing test 
> by using add-42 wherever he needed 42 added to a number, and  RMS wants 
> credit and ownership and control of it all. 

That might be what RMS wants (or not, I've never asked him), but it 
doesn't follow from the licence.  What follows from the licence is that 
you have to distribute the derived work as GPL _or not at all_.  I 
practice - if not in marketing terms - that's no more a land grab than a 
proprietary licence saying "you can't use this to add your own numbers 
to 42 at all and if you do we'll eat your brains".

The other consideration is that, and notwitshtanding any text to the 
contrary in the GPL, it's not actually up to the copyright holder to 
define what "derived work" means: it's for the court to decide that. 
Now, I don't want to imply that courts are rational animals that can be 
relied on to understand all the issues in technical cases like this (ha, 
I slay myself) but really, if there's a reasonable concern that an 
implementation of the major advances in computer science you describe 
are legally derivative of someone's function that adds 42 to its 
argument, your legal system is fucked.  Redo from start.


-dan