[License-discuss] comprehensiveness (or not) of the OSI-approved list

Rick Moen rick at linuxmafia.com
Thu May 23 18:33:01 UTC 2019


Quoting John Cowan (cowan at ccil.org):

> The deep [patent] threat comes from third parties, which is a risk
> that neither ther the licensor nor the licensee can reasonably
> mitigate.

Sure.

> Okay, but if we accept that patent infringement makes a piece of
> software not open source, we are in this position:  if someone asks
> "Is program X open source?" our only replies are "Definitely not"
> and "Maybe."  A classification like that isn't very useful.

Which is a compelling reason not to be quite that pessimistic.  
But consider a worst-case example, a new video codec design and example
open source-licensed implementation.  The hypothetical author is no
neophyte and does as much patent research as can be reasonably justified
absent being backed by a deep-pocketed large organisation with a high
degree of resolve.  No patent landmine is found.  The happy day comes
and version 0.91 emerges, the first public beta, and everyone rejoices
because it's good (let's say) and it's open source (e.g., as to licence
terms and source access).

I _would_ join everyone in calling that open source -- and would
continue to do so in this scenario until two days after v. 2.1 emerges,
two years later, when MPEG-LA or Qualcomm suddenly says 'Please pay
exhorbitant royalties for our cherished if little known patent.'  I
would now, if asked (and if credible people say the infringement claim  
has teeth), say the example codec implementation is _not_ open source
(within reach of that patent), even though I asserted yesterday that it
was -- because now it is known to fail OSD#1, and yesterday it was
believed not to.

This course of events is unfortunate, but, I continue to say it's
reality.  It can happen to any codebase.  For parser code, say, the
danger is not significant, but remains non-zero -- like stepping on a
non-figurative landmine in Central Park.  With video codecs, it's more
like the risk of stepping on non-figurative landmine in Western Sahara:
_very_ non-trivial.

Does that mean a new example video codec is automatically at most
_maybe_ open source?  I wouldn't say so.  I'd believe the licensing
terms -- but IMO there's always an invisible asterisk saying that sudden
discovery of unsuspected patent problems could badly impair recipients'
rights, and this cannot be otherwise dealt with because that's what
'unsupected' means.

OSI doesn't need to volunteer to make sure everyone knows this, nor do
users of the phrase 'open source'; not their job.  


> It's more meaningful IMO to say that such a program *is* open source
> and accept that not all open-source software is usable everywhere.

Quibble:  I'm almost in agreement, but I'd say that such a program is
open source except in places where (during a patent's runtime) it isn't.

The problem with your formulation is that it would have you say that a
program validly issued under an OSI-Certified licence is open source
even if it totally fails OSD#1 on account of royalty requirements.
I would suggest that's not the soil you want to build on.

> (BTW, I've decided to start omitting the "-- " line because so many
> mail clients, notably Gmail, suppress it and everything after it.)

Understandable, but my own parsing of that situation is that if users
choose an MUA that unconditionally strips everything below a son-of-RFC
1036 (aka RFC 1849[1]) .signature delimiter, then they have made a valid if
life-impoverishing choice and that their request to be denied enjoyment
should be honoured.

Your way, you are putting people in a position where they cannot 
turn off .signature display in the conventional way, even mindfully.  


[1] RFC 1849 proclaims this standard 'not now appropriate for use in
current implementations.  I respectfully differ.  (However, on further
examination, I see that the remark concerns other parts of son-of-RFC
1036 that are more-evidently obsolete and problematic.)

-- 
Cheers,              "I am a member of a civilization (IAAMOAC).  Step back
Rick Moen            from anger.  Study how awful our ancestors had it, yet
rick at linuxmafia.com  they struggled to get you here.  Repay them by appreciating
McQ! (4x80)          the civilization you inherited."           -- David Brin



More information about the License-discuss mailing list