[Qt-interest] Difference between commercial and open sourcerelease of Qt 4.7.4

[BRM]

>From: Scott Aron Bloom <Scott.Bloom at onshorecs.com>

>To: QT Interest List <qt-interest at trolltech.com>
>Sent: Monday, October 3, 2011 11:16 AM
>Subject: Re: [Qt-interest] Difference between commercial and open sourcerelease of Qt 4.7.4
>
>>> But now we have some kind of mini-fork with qt-opensource and 
>>> qt-digia which isn't using the same source and unavailable as 
>>> standard distribution package.
>> 
>> Now? All distributions have done a "mini-fork" as you call it. All of 
>> them apply patches.
>
>The more interesting question would be, in the case of Digia at least:
>are those changes ever merged back into the "main" Qt source tree? Or do
>they have their own closed-source tree and constantly pulling in the
>changes from "main Qt" behind locked (commercial) doors?
>
>Cheers, Oliver
>_______________________________________________
>
>Shouldn't all their changes be "public", and any one of us could "merge"
>them in by hand to the git repository right?
>
>Isnt that the whole point of Opensource?
>


That is a common misconception in the FLOSS community.
Even the GPL does not guarantee that the community itself will receive the changes made by someone downstream of a pure GPL project.
It only guarantees that the recipient of the GPL'd project will receive the changes; that recipient may then push the changes back to the original project, but there is no guarantee of it.
This is, after all, how forks of a project work - the forked version doesn't necessarily push the changes back to the original project; but their receipients may.

A corporate entity could do the same thing - a private fork - and only pass those changes on to whomever they distribute their private fork to.
They will be in full compliance with the GPL.

Of course in this case, Digia benefits from the Commercial licensing portion of the license as they could do something private without providing the changes on.
However, the commercial license does provide a copy of the source, so that would be relatively hard to do.


$0.02


Ben

P.S And yes, this theory has been vetted on gplviolations.org; though I probably explained it better here.