[R-gui] RE: [R] The hidden costs of GPL software?

John W. Eaton wrote:

On 17-Nov-2004, Philippe Grosjean <phgrosjean at sciviews.org> wrote:

| - There is no possibility to make a commercial GUI for R (thanks to
| the GPL),

This is false.  Please don't confuse "commercial" (Red Hat
and SuSE GNU/Linux distributions are commercial software)
with "proprietary".

jwe

Ooops! Sorry, and thank you for correcting me. I mean "proprietary", of
course.
Best,

And it isn't obvious that it is true even if you mean proprietary. A GUI 
that ran R just by sending commands to stdin and getting results from 
stdout could clearly be proprietary without violating the GPL.  The 
question of exactly what level of closer integration is allowed would get 
complicated and I won't speculate.

 	-thomas

John W. Eaton wrote:
 

On 17-Nov-2004, Philippe Grosjean <phgrosjean at sciviews.org> wrote:

| - There is no possibility to make a commercial GUI for R (thanks to 
| the GPL),

This is false.  Please don't confuse "commercial" (Red Hat 
and SuSE GNU/Linux distributions are commercial software) 
with "proprietary".

jwe
   

Ooops! Sorry, and thank you for correcting me. I mean "proprietary", of
course.
Best,

Philippe Grosjean
 

This thread has gone to be centered around the GUI of R and what it is 
good and bad.

However, is the above statement correct? To me it seems like there is a 
fully working R-proxy dll for windows and other ways to interface 
against R that only binds to LGPL components. You can build completely 
proprietary packages and front-ends to R without having to make sources 
available, as long as you distribute changes to R itself as source.

In my opinion anyone can be to R what S+ is to S. Can any developer 
comment on this?

    Best regards,

    David