LGPL Survey Results

[Andrew Eddie]

Well, the survey period is over and I've tallied the results:

http://www.slideshare.net/AndrewEddie/joomla-platform-lgpl-survey-15825187

We got 113 responses. That number should give us a 5-10% margin of error on the results (as far as I can tell, yes, we have a statistically representative sample).

The overall result is 64% in favour of allowing the Joomla Platform to be licensed under the LGPL.  I've included a breakdown on a few other metrics:

For people that have signed the CLA, the approval level is 71%.

For people that have added Platform code, the approval level is 81%.

For people that have added CMS code, the approval level is 72%.

For people that have written a Joomla extension, the approval level is 68%.

26 people indicated they had contributed code to the Platform. Github suggests the total number of contributors is about 100. 

I don't think there are any surprises in the results except there are more people writing non-CMS Platform apps than I expected (which is a good thing).

The Platform Maintainers and PLT still have to review the results but I think it's reasonable to assume, given we have close to a two-thirds majority or better on a number of fronts, we'll be recommending that OSM investigate whether the license change is possible/feasible.

If there is interest, I can make the raw data available.

Thanks to everyone who participated.

Feel free to discuss the results.

Regards,
Andrew Eddie

[brian teeman]

Reading the full report at slideshre I think that 30% who "strongly disagree" is a very considerable number

[Amy Stephen]

Those are compelling numbers from those contributing code, esp. the platform code which is of relevance here. That is much higher than I expected. It's rare to see that kind of agreement.

[Donald Gilbert]

That's good news. Let's see what OSM thinks about those numbers and hopefully we get a positive outcome. :) (totally unbiased right there haha)

[Radek Suski]

We are working all possibilities through since the idea come lately out.

I'm (personally) supporting this move but please consider that there are many legal and formal aspects we have to work out.

Regards,
Radek

[Paul Orwig]

Hi all,

Thanks Andrew for initiating this discussion and the survey, and thanks to everyone who has shared their thoughts about this subject in this and the threads on the other development lists. I have read every post in those threads, and I have pointed OSM board members to these threads as well with the request that they also follow along. And as Radek mentioned, we have also been having some internal discussions about some of the issues to potentially consider for more than a few weeks.

From Andrew's earlier email today, it sounds like the next step is for the Platform maintainers and/or PLT to discuss this a bit further and then decide if they want to formally ask OSM to consider this change. If that happens, then OSM will want to take some time to get some legal guidance before we discuss things further. I have expressed to OSM that I think if things do get to this point, then our discussion should be public.

Thanks again,

paul

[Andrew Eddie]

On 2 January 2013 23:16, brian teeman <joom...@googlemail.com> wrote:

Reading the full report at slideshre I think that 30% who "strongly disagree" is a very considerable number

It depends how you define "very considerable". It's "just a number" - 30 respondents (26.5% actually - just over a quarter which, to me, doesn't qualify for "very considerable") vs 41 that strongly approve.

Of those:

21 have said they have developed a CMS extension (70% of that group)

8 have said they have signed the JCA (27% of that group)

5 have said they contributed Platform code (17% of that group)

11 have said they contributed CMS code (37% of that group)

Personally I would not have been happy with much less than a two-thirds majority (if it was 50/50 I would have let it go), which is an accepted "norm" for most decisions like this. My interpretation is that amongst contributors, there is better than two-thirds support.

Another observation is that the people that signed the JCA should all be fully supporting the LGPL change - because that's what they signed up for. However, this is not the case and it means that a) people didn't really read or understand it, or b) the understood it but hoped it would never be enacted. So we have some work to do there in terms of community education.

That said, for those that have read the JCA, there is a little clause about how the license of the Platform interacts with the Joomla CMS. What we are talking about is allowing the Joomla Platform to be used standalone under the LGPL (in hindsight, I wished we'd allowed MIT, but that option is lost now). However, when the Platform is used in combination with the CMS, the license behaves as if it is the GPL. So as far as the CMS side of the project goes, everything is still GPL. If you are rolling a command line application, or a daemon, etc, the license behaves as LGPL.

I have been under that assumption that everyone understood that's what the JCA allows us to do. The survey results highlight that this assumption is faulty. So in terms of the people who raised the comment about marketing the GPL to extension developers, that will not change. The CMS must remain GPL through and through.

I am surprised nobody has mentioned how the gender and region results. I think those are something we need to work on for the platform. The familiarity with Github was, to me, very high which is encouraging.

Regards,

Andrew Eddie

[Donald Gilbert]

What were the gender / region results?

[brian teeman]

On Wednesday, 2 January 2013 22:33:36 UTC, Donald Gilbert wrote:

What were the gender / region results?

slides 4 and 5 http://www.slideshare.net/AndrewEddie/joomla-platform-lgpl-survey-15825187 

[brian teeman]

On Wednesday, 2 January 2013 22:29:44 UTC, Andrew Eddie wrote:

On 2 January 2013 23:16, brian teeman <joom...@googlemail.com> wrote:

Reading the full report at slideshre I think that 30% who "strongly disagree" is a very considerable number

It depends how you define "very considerable". It's "just a number" - 30 respondents (26.5% actually - just over a quarter which, to me, doesn't qualify for "very considerable") vs 41 that strongly approve.

Yes its just a number but a much higher one than I would have expected especially for the reasons you highlighted below

[Donald Gilbert]

Thanks Brian. :) I guess I should actually visit the links people post. Haha

[Andrew Eddie]

On 3 January 2013 09:52, brian teeman <joom...@googlemail.com> wrote:

Yes its just a number but a much higher one than I would have expected especially for the reasons you highlighted below

Okay.

Regards,

Andrew Eddie 

[Elin Waring]

If you don't have a random sample margin of error is meaningless.   

Elin

[Andrew Eddie]

On 3 January 2013 10:31, Elin Waring <elin....@gmail.com> wrote:

If you don't have a random sample margin of error is meaningless.   

Thanks for the clarification.

Regards,

Andrew Eddie

[Amy Stephen]

On Wednesday, January 2, 2013 4:29:44 PM UTC-6, Andrew Eddie wrote:

I am surprised nobody has mentioned how the gender and region results. I think those are something we need to work on for the platform. The familiarity with Github was, to me, very high which is encouraging.

Regards,

Andrew Eddie

I did not mention it but I did notice it. I appreciate it that you raised it as a goal for the platform. Definitely a worth goal. Thanks!

[Alonzo Turner]

So 57% of respondents say that they are "Developers", but fewer than 11 respondents out of 113 responses ( < 10% ) actually write software. How exactly does that work?

[Andrew Eddie]

On 3 January 2013 11:41, Amy Stephen <amyst...@gmail.com> wrote:

I did not mention it but I did notice it. I appreciate it that you raised it as a goal for the platform. Definitely a worth goal. Thanks!

That was just a personal observation. Nothing official about it.

Regards,

Andrew Eddie

[Andrew Eddie]

On 3 January 2013 11:45, Alonzo Turner <alonzo...@subtextproductions.com> wrote:

So 57% of respondents say that they are "Developers", but fewer than 11 respondents out of 113 responses ( < 10% ) actually write software. How exactly does that work?

I think the "Your role" question was badly designed on my part (that became obvious during the first days of collecting results). I'd just ignore the results of that question.

Regards,

Andrew Eddie

[Andrew Eddie]

Just FYI

For those that missed the news (http://developer.joomla.org/news/551-plt-meeting-notes-february-2013.html), the PLT has officially asked OSM to look into the license change.

Regards,
Andrew Eddie
http://learn.theartofjoomla.com - free tutorials and videos on Joomla development