Re: Draft HTML5 licensing survey

On 2011-04-26 20:07, Sam Ruby wrote:
> On 04/26/2011 01:55 PM, James Graham wrote:
>> What is the reason for excluding options upfront? It seems bizzare to
>> have a poll where options known to have support are prefiltered. It
>> seems particularly strange where the discussion has indicated that the
>> options being removed might represent common ground for at least some of
>> the involved parties.
>
> It is equally as bizarre to have a poll that explicitly includes an
> option over which absolutely nobody has advocated that option.

I'm not sure how you're defining advocacy in this case, since the MIT 
licence has been suggested many times as a suitable licence over the 
years of discussing this issue.

As to what should be included in this particular survey, that depends on 
exactly what we wish to achieve at this time, should none of the 3 W3C 
PSIG options be successful.

It's clear that providing one or more additional options for those of us 
who are opposed to those 3 licences, so that we may say so, would be 
very useful.

Personally, an alternative licence that is a well established, 
non-copyleft free software licence that is widely compatible with other 
licences and permits forking would be suitable.  This would more or less 
include 2/3-clause BSD, MIT or CC0.

But at this stage, there are two questions to answer here:

1. Should the W3C reconsider their position on the use of free software 
licences that allow forking?
2. If they do reconsider, which free software licence can we agree on?

At this stage, answering the first question seems more important than 
answering the second.


Thus, there are 2 approaches we could take:

Approach 1: Present the survey as effectively a choice between 1) one of 
the 3 PSIG licences (listed independently), and 2) other free software 
licences (listed as one option).

Approach 2: List all suggested licences as independent options.


The first approach would only address the first question.  It would 
basically be an option that called for a well established, non-copyleft 
free software licence that does allow forking, but without being 
specific about which one, and noting that further discussion would be 
required if and after the W3C agrees to even consider such licences.

The survey should note support of this option indicates general support 
for the idea of using a free-software licence, rather than equal support 
for all possible alternatives.  (i.e. Supporting this option doesn't 
necessarily mean you support both CC0 and MIT equally.)  Respondents can 
always clarify their answer in the comment field if necessary.

The advantage of this approach is that those of us who support the use 
of a free-software licence, but don't have a strong preference for 
either one, can say so without needing to be so specific.  This also 
lets the debate about exactly which free software licence continue at a 
later time, once the W3C even agrees to consider such licences.


The second approach would list the 3 PSIG licences as well as the other 
suggested alternatives.  This should include MIT, 3-clause BSD and CC0.

This approach has the advantage of presenting all options equally, 
including those that the PSIG has not put up for consideration, but at 
the risk of excluding other potential options that may not have been 
considered yet.

If we take this second approach, the reason that each of MIT, 3-clause 
BSD and CC0 should be listed is because:

1. Each of those licences has been suggested in the past, either
    formally or informally, as a suitable alternative.

2. They are all well-established, non-copyleft, widely compatible free
    software licences that meet all previously discussed use cases,
    including forking.

3. We know that there are some organisations who would probably support
    either MIT and/or BSD, but which do not support CC0 (notably Apple).
    So listing CC0 as the only alternative to the 3 PSIG licences,
    simply because it is the only one that has been formally proposed,
    would be unfair.

4. While BSD and MIT are substantially similar, 3-clause BSD also has
    the no-endorsement clause which may influence the preferences of
    some respondents, who may prefer one over the other.  (I don't think
    2-clause BSD and MIT have any substantial difference from a legal
    perspective)

Finally, if it helps, and if this second approach is taken, I am 
personally advocating for both MIT and 3-clause BSD to be considered and 
included in the poll.  (I am not formally representing Opera here.)

-- 
Lachlan Hunt - Opera Software
http://lachy.id.au/
http://www.opera.com/

Received on Wednesday, 27 April 2011 21:06:32 UTC