LCDproc development and user support list

Text archives Help

[Lcdproc] GPL and loadable module

Chronological Thread 
  • From: n9ubh AT (David Douthitt)
  • Subject: [Lcdproc] GPL and loadable module
  • Date: Mon Dec 10 05:14:01 2001

On 12/7/01 at 4:02 PM, David GLAUDE Mailing
<dglaudemailing AT>

> RMS "is" the Free Software Fondation so maybe we should
> listen to him.

Richard Stallman was the founder of the Free Software Foundation and
the creator of Emacs. He left the MIT lab when the hackers started
leaving to work for corporations like BBN and Symbolics producing
proprietary code.

RMS has alienated a LOT of the Open Source movement. He's actually
gotten into heated arguments at more than one Open Source convention
with big Open Source players. Me, I think that I'm more in line with
Eric Raymond - who wrote fetchmail, and the missive "The Cathedral and
the Bazaar" which instrumental in convincing Netscape to release the
code to Navigator.

> What Linus might be relevant to him but not to Linux,
> because Linux contain copyright of many.

Linux is actually trademarked by Linus, after a fiasco in New Jersey
when someone trademarked the name and sued. They lost.

> As far as I am concearn, we should not accept non GPLed
> source code into our CVS tree. It should be clear that any
> patch send to us is implicitly GPL. Anything else is
> searching for trouble.

I disagree - I would think that any NEW driver code could be licensed
under an OSI approved license. OSI is the Open Source Initiative.

However, under a single code tree there should be a single license -
LCDd should be under one license. As I said before, the drivers under
the current model are unique - but I'd think that any GPL-compatible
license (such as the current BSD or MIT/X License) should be okay.

With dynamic modules, I agree with Linus - if some LCD manufacturer
releases a binary module, so be it. Why not? I'm with Linus on this
one :)

> In fact many MP3 player (and other) might love it to
> bypass LCDd since they don't need IP/TCP/socket anymore
> and don't care about multiple driver.

That calls for #ifdef's and the ability to stub all of that out.

Of course, you don't need more than one driver if you want - and the
system already is set up to connect to instead of a visible
network socket.
David Douthitt
UNIX Systems Administrator
HP-UX, Unixware, Linux
n9ubh AT

Archive powered by MHonArc 2.6.18.

Top of page