[dmd-internals] DMD copyright assignment

[Jonathan M Davis via dmd-internals]

This didn't show up before for some reason, so I'm sending it again. Sorry if
anyone gets it twice.

On Wednesday, June 25, 2014 09:04:36 Walter Bright via dmd-internals wrote:
> 8. Dealing with this is not something that I want to do. I just want to
> write code. But who else is going to? Somebody has to step up and do their
> best to make these sorts of decisions, even when not everyone agrees. For
> better or worse, the ball's in my court, and I need to deal with it the
> best I can. I believe that copyright assignment for major contributors to
> the DMD compiler code is the most practical and pragmatic solution for us.
> It protects you, I, and the other contributors to ensure our work will not
> get discarded for unfortunate legal reasons. So I respectfully ask for your
> indulgence on this.

I confess that before this discussion I had no idea that by contributing to
dmd, contributors transferred copyright to digital mars, and I do think that
the concerns about boost are a bit paranoid given how simple and permissive
the license is (though it certainly is theoretically possible that something
could go wrong with it eventually), but given the nature of dmd, I don't think
that the requirement that copyright assignment be transferred is unreasonable.
And it's always been this way, so it's not like Walter is asking that the
copyright for already contributed code be transferred.  So, while I probably
wouldn't make the same choice myself, I support this decision.

I think that it would be a very different question if druntime or Phobos had
this requirement, given how much more modular they are and how they cover a
lot of different areas instead of just a compiler. It's not only easier to
work around licensing issues with Phobos due to is modularity, but I also
think that we're a lot more likely to run into problems with folks not wanting
to contribute when you're talking about them adding some piece of
functionality to the standard library that they may have and use elsewhere as
opposed to fixing a bug in the compiler, or even adding a new feature to it.
Most folks don't have their own compilers. But it doesn't sound like the
question of assigning copyright for druntime or Phobos is even on the table,
so while I think that the concerns over copyright assignment are more relevant
there, they're also a non-issue at this point.

>From the sounds of it, all that Walter is requesting is that we maintain the
status quo. And I don't think that that's unreasonable.

- Jonathan M Davis