My last remark was too snarky and out of context, I am sorry. You want both projects to merge again with OO being the 'core' and LO the 'user interface'? Sounds intriguing, but are both code bases still compatible enough for that? Would this not be more effort than gain?
What is obvious is that the AOO project cannot support, at
the present time, being an end-user focused effort. I would
suggest we focus on not being one, but instead being a
framework or library that can be consumed by actual
end-user implementations.
OpenOffice and LibreOffice are already "merged". In fact in the past years, after the fork, LibreOffice developers regularly tracked OpenOffice source code repository and merged all the useful changes they could find, by either picking the original code or re-implementing it their own way.
In addition to that they also completely overhauled and simplified the build system, removed tons of dead code and translated German comments which where there since StarOffice, replaced Java components with equivalent ones written in C/C++, are in the process of switching from GTK2 to GTK3 (mandatory to get Wayland support in Linux), etc. Doing so they made the LibreOffice source code base sustainable on the long term and lowered the difficulty entry level for new contributors, while apparently OpenOffice buildbots are not even able to rebuild their software since last year...
The only useful asset which remains to OpenOffice is the trademark.
"In fact in the past years, after the fork, LibreOffice developers regularly tracked OpenOffice source code repository and merged all the useful changes they could find, by either picking the original code or re-implementing it their own way."
And therein lies the problem... AOO was simply consumed. There was no quid-pro-quo where these were then donated back to AOO. Of course, the ALv2 does not require that, and so LO was perfectly within their right to not donate back. But it also seems somewhat "shady", and not the kind of behavior one would expect of a fellow FOSS project, but more like a corporate FOSS bottom feeder.
So development was basically all single stream... what was useful was used but nothing was given back.
I wish people would recall this when they mention how "arrogant" Apache is, or how Apache acted in bad faith and stuff like that or that Apache wanted to "control everything". Certainly if a single, unified OO eco-system was important to LO/TDF, they had opportunities to help make that happen. Let's at least be honest here.
LibreOffice exists because Sun wasn't accepting useful contributions to OpenOffice or putting much effort into maintaining it, and then Sun was acquired by Oracle, an open-source-hostile (to say the least; they killed OpenSolaris and sued Google for reimplementing Java!) company. So they had to fork.
Of course, because everyone then ran away to LibreOffice, OpenOffice basically died. And, as expected, Oracle hardly cared for it, stopped work on it entirely and handed it off to the Apache Foundation to die.
Really, they should've transferred the trademark and copyright to the Document Foundation instead, or at least coöperated with them. But that wouldn't have been spiteful enough of the community. In fact:
> Oracle was invited to become a member of The Document Foundation. However, Oracle demanded that all members of the OpenOffice.org Community Council involved with The Document Foundation step down from the OOo Community Council, claiming a conflict of interest.
> It was originally hoped that the LibreOffice name would be provisional, as Oracle was invited to become a member of The Document Foundation. However, Oracle rejected requests to donate the OpenOffice.org brand to the project.
Why should LO have an interest to switch to an OO core/library? Just to gain the OO name back and be present on the website? Tough deal. Also, I still think the code bases are too different by now [1] and there is little to gain.
Without at at least one 'consumer'/UI (which you do not want or cannot provide) there is little point in developing an OO core/library, or is it? Which leads to my snarky comment about end-users and their role...
The Open Office eco-system is larger, by far, than LO and AOO. Maybe LO may not benefit at all from a permissive licensed core (although it certainly has benefited from it, as noted elsewhere), but certainly other OO implementations might.
We have seen, and even the FSF admits it, that open standards do best with permissive licensed implementations. So if you want wide adoption of open standards, a permissive license is likely better.
The world is not just LO or AOO. After all, the enemy, so-to-speak, is this proprietary s/w called MSO...
In what parallel universe do you get that conclusion from that thread?