GPL-licensed binary plugins

[Mikhail Gusarov]

"Iain Barker" <iain at member.fsf.org> writes:

> It sounds likely, yes. But is not necessarily so, these 'grey area'
> situations are notoriously difficult, it very much depends how the
> individual implementation was done.

> I think some further investigation would be prudent before jumping to
> any conclusion publicly.

Authors have released SDK for their device, including API for the
application mentioned. This API tightly binds the plugins to the
application (even not mentioning that it is C++ API), it is not publicly
available one, but application-specific.

> However, if the main program loads the plugin into memory and calls
> its functions directly (say for example, a dynamically linked library
> in the conventional sense), then most likely they are sharing intimate
> details of their sata structures etc; the use of these internal
> semantics is what would constitute a 'whole based on the copyrighted
> work', and therefore copyright derivation and the GPL requirements for
> open source come into play.

Seems to be a case here.

> As I said, it's a grey area and not 100% clear for your particular
> case. You can certainly try to approach the parties involved, and see
> if they are willing to comply without any further action.

Vendor quickly provided the sourcecode for GPL-licensed toolchain they
published when I asked for it, so I'm going to ask them to publish other
GPL-covered code.

If this won't work I'm going to ask gpl-violations to help with this
case. Actually there is much more GPL-covered code without corresponding
sources in their firmware/SDK than just mentioned application and
plugins, but this was a gray area, so I decided to ask for help.

Again, thanks for information.

-- 
JID: dottedmag at jabber.dottedmag.net