Which License should I pick?

[Chuck Swiger]

On Dec 3, 2003, at 4:21 PM, Scott Long wrote:
> Hello, everybody,

Hi, Scott--

> [ ... ]
> I have briefly skimmed the list of licenses at
> http://opensource.org/licenses/ and the BSD license looks like it
> fulfills my conditions. I'm posting to this list to see if my
> interpretation of the license is correct. I'm also open to suggestions
> for other licenses on the list of OSI accepted licenses.

Given the criteria you've mentioned, the BSD license is probably the 
best choice, as you yourself have concluded; the MIT license is another 
good choice.  Almost any other license would be significantly longer.

> SECOND LINE OF QUESTIONS:
>
> The project itself performs actions on the in-core binary images of
> running processes. It is capable of saving snapshots of the address
> spaces of running processes to disk. I would like to have
> clarification whether or not such snapshots would be considered
> "derived works" of the programs they came from.

Probably, yes, unless your program specificly disclaimed rights with 
regard to the output being generated by the program.  For instance, the 
GPL states:

"The act of running the Program is not restricted, and the output from 
the Program
is covered only if its contents constitute a work based on the Program 
(independent of having been made by running the Program).  Whether that 
is true depends on what the Program does."

> Would it be legal to distribute one of these binary snapshots if it 
> came from a running program which is licensed under a more restrictive 
> license such as the
> GPL?

If it was legal to distribute the original program under the GPL (for 
example), it would be legal to distribute a binary snapshot, but that 
snapshot would also be covered by the GPL.  In specific, the Emacs 
editor unexec()s itself after loading a bunch of LISP code, and the 
binary image produced by this is still GPL'ed.

> This brings up a related question, which although it doesn't directly
> pertain to my project, is equally important: most UNIX-like operating
> systems are capable of creating core files of crashed processes. Are
> these core files "derived works" of the programs which produced them?

Yes, or perhaps a core file would be considered a direct copy of the 
original work, not a derived work.

> After all, they contain portions of the address space of the processes
> in execution -- they contain binary code. If they are derived works,
> it would appear that all the various crash-reporting applications
> which send core files to a central location for analysis are actually
> breaking the law!

Sending a complete core file of a copyrighted program might present an 
issue, but that's not a concern for Open Source software.  Also, most 
crash-reporting systems (I'm thinking about Talkback in 
Netscape/Mozilla, the "Dr. Watson" crash reporting in Windows, or 
Apple's bug reporting system in OS X) actually only send a stack 
backtrace and library/DLL version information, not a complete memory 
dump.

-- 
-Chuck

--
license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3