Subject: External Deployment Re: OVPL & "Otherwise Make Available"
From: Ernest Prabhakar <>
Date: Fri, 26 Aug 2005 10:59:09 -0700

Hi Chuck,

On Aug 26, 2005, at 10:49 AM, Chuck Swiger wrote:

> Ernest Prabhakar wrote:
>> On Aug 26, 2005, at 7:32 AM, Lawrence Rosen wrote:
>>> Nope. External deployment occurs in the OSL when the software is   
>>> "used by anyone other than You," not when it merely delivers  
>>> email to those  people.
>> I wonder if the term "external deployment" is a much clearer  
>> synonym  for "distribution" that might help resolve this ambiguity.
> Is making the output of a program publicly available "external  
> deployment"?
> Is making a program available for remote execution "external  
> deployment"?

Rather than attempt to answer that explicitly, I would simply point  
out that other OSI approved licenses have used such a term (e.g.,  
Larry's "Open Software License"), so the OVPL could merely inherit  
that pre-existing community understanding:
> 5. External Deployment. The term "External Deployment" means the  
> use or distribution of the Original Work or Derivative Works in any  
> way such that the Original Work or Derivative Works may be used by  
> anyone other than You, whether the Original Work or Derivative  
> Works are distributed to those persons or made available as an  
> application intended for use over a computer network

If the OVPL's various obligations triggered off a definition of  
"external deployment" like this rather than the more ill-defined  
"distribution", I suspect this issue would go away.

-- Ernie P.

> I would not expect free software to claim ownership or impose  
> obligations onto the output of a program, unless such output  
> includes significant distinctive content from the software itself.   
> And even if it does, software such as GCC and flex/bison explicitly  
> disclaim that the GPL would apply to the output of the compiler  
> toolchain, even if the program includes the flex skeleton or  
> certain low-level GCC code interfaces (floating point math ops,  
> libiberty.a?).
> However, if I "telnet smtp" and send a message  
> directly by hand, I am running and using the SMTP software.  The  
> classic network architecture would have had a specific sendmail  
> process dedicated just to me, which was invoked from inetd to  
> service my specific connection, and which would go away after I am  
> done and close the connection.
> The organization which operates that mail server has not  
> redistributed either the source or the binary of their SMTP server  
> to me, I am simply running it remotely and interacting with the  
> output it generates.
> -- 
> -Chuck