GNU License warning:

Ignacious via Digitalmars-d digitalmars-d at puremagic.com
Thu Jan 12 18:25:03 PST 2017


On Friday, 13 January 2017 at 01:27:02 UTC, Chris Wright wrote:
> You offer an API and someone decides to build on it using the 
> GPL -- no trouble there; your work is not a derivative of 
> theirs, so their copyright cannot place restrictions on your 
> work.

That makes no sense(it's obvious by the definition of derivative 
so you are not saying anything meaningful/useful). Obviously if 
you build an independent work you are free to chose a license and 
no one building any work off of yours or not can cause you 
problems.

> You build against an open standard and the only implementation 
> is GPL -- your work is a derivative of the standard, not 
> necessarily the GPL'd work.

That depends. The standard could be GPL's too. Anything can be 
copyrighted and licensed how the creator wants as long as it is 
legal. In any case, that says nothing about a single work.

> You build against an open standard with an MIT licensed 
> implementation and someone else builds a GPL implementation -- 
> no trouble there; your work is not a derivative of theirs, so 
> their copyright cannot place restrictions on your work.

You haven't really said anything relevant to the post.

The issue is with how the GPL defines proper use of pre-existing 
works. The ultimately point is that they arbitrarily decide how a 
work uses another based on "fork and exec" and "library". My 
point is that those are ultimately artificial because whether we 
call a function/app through a library or through a command line, 
they are effectively the same(the difference being 
performance/convenience, which is the whole point of loading a 
library vs using the command line).

They admit this in the gpl FAQ(if you read it you will see) but 
the fact that they still create arbitrary division suggests the 
license is somewhat meaningless/incompetent.

Licenses should be more specific in their terminology and their 
behaviors and effects rather than using arbitrary divisions.

Also, while not proof, the fact that the majority of donations to 
the foundation go to the lawyers(if true) also suggest that it is 
somewhat of a scam(at the very least, something is fishy).








More information about the Digitalmars-d mailing list