Oracle proposes donating OpenOffice.org to Apache Software Foundation

We are beginning to see the strategic use of software foundations, not as
doting and caring parents, but as separate and incompatible fiefs used to divide -- and hopefully -- conquer. Oracle has pioneered this, first with
their proposed donation of their fork of the Hudson codebase (after
forcing the authors of the code to rename through trademark threats) to
Eclipse -- a foundation whose procedural and legal overhead is of no appeal
to the Hudson authors. The intentional result is the "appearance" of
distance by Oracle, but the proposed stewardship of the code primarily
by Oracle's partners, while keeping the authors at a distance and making
them appear as the bad guys (they are apparently partnering with SPI,
a better match to their style). Donating code that you do not morally
own (and relicensing it to boot -- from MIT to EPL) should incite the
contempt of every free software partisan.

In this case, Oracle actually does morally own a significant chunk of
the Open/Libreoffice codebase, and it is available under
free software, though non-copyleft licenses. They do have the
moral right to transition significant blocks (though not all) of the
codebase to an Apache license. But once again, Oracle is using
foundations eager to expand mindshare and codebases as pawns in this
game of kings to position themselves against existing communities.

Apache lessens itself by being Oracle's pawn -- one cannot say unwitting,
because Apache has had bitter lessons of late in the faithlessness and
arrogance of Oracle. Oracle should not be praised for this move, since
it is clearly in bad faith. IBM should also be condemned for encouraging
an adaptation that intentionally leads communities divided. Their move,
clearly, is to defend against any potential future evolution of
Libreoffice that includes copyleft. The communities be damned.

I have very mixed feelings about this.

My first instinct is that LibreOffice stepped up, provided a huge service to the community, and essential won against Oracle in the market. We should all just get behind LibreOffice and leave OpenOffice.org to die on the vine.

It seems like LibreOffice has some real momentum. They have made it really easy for people to contribute and people are actually pitching in. They seem to have unlocked a groundswell of contributor potential that OpenOffice.org just never mustered. I even got myself on the contributors list.

I do not agree that LibreOffice could not compete with Oracle, the ASF, and IBM. I think that even IBM was contributing to LibreOffice before Oracle made this move.

That said, I do actually prefer the Apache license overall. I also understand the value of having commercial backers like IBM. So, if LibreOffice would have been APL2.0 to begin with, that would have been great.

The whole idea of Oracle and IBM taking what little leverage they have left in this situation to essentially spit in the face of LibreOffice also just rubs me the wrong way.

that which can be easily explained by incompetence. Before coming up with conspiracy theories, consider the possibility that Oracle failed at managing OOo and is now trying to fix things. So they're handing the code over to a group of people capable of doing things properly (hopefully).

I expect such a "donation" won't come without strings...namely one which prohibits the ASF from turning around and donating OO to the Document Foundation, on the off chance that ASF principles have both a sense of justice and a sense of humor.

I had slept over this news (actually, the announcement / response from TDF), and after giving the whole thing some thought, feel that the right thing for ASF to do would be to turn down the donation.

There is already TDF which looks after the LibreOffice / OO.o project and is doing a remarkable good job of it. (just look at the start up time for LibreOffice over OO.o - first impression for a normal user).

There may be other issues - like licensing, which Oracle (or whoever is pushing Oracle to do this) wants to skirt around. But right now, this is an attempt at (a) washing its hands off OO.o (b) insulting TDF's efforts in improving the project's performance and usability.

Probably, ASF may accept the code base without (or refusing) relicensing, but as others have pointed out, one cannot give what one does not have.

My expectation is that Oracle doesn't care about any of this. OO.o isn't why they bought Sun, it's not something they care about, and it has been just generating bad press for them. They'd probably prefer to just never talk about it again. IBM, however, wants to be able to make a proprietary version, can't start from the LGPL base, but would like to stay in sync as much as possible with it. Oracle doesn't want to piss IBM off more than necessary, particularly since they're competitors and Oracle gets a lot of anti-trust attention. I think IBM wants the ASF version to be something that LO will merge regularly and to which people will contribute the work they'd like in all office software, and Symphony and LO can diverge less fundamentally.

FWIW, RMS's canonical example of a piece of software which should be under a non-copyleft license is libpng, because it implements a free standard which competes with a proprietary format. Surely, therefore, there should be an implementation of the semantics of ODF under a permissive license, and it should be used by all open-source software that uses ODF as well as any proprietary software vendors who are not trying to damage the format. So it would make sense to end up with OO.o under a permissive license having the standard ODF semantics engine and an unmaintained UI; Symphony replacing the UI with a proprietary one; and LO replacing the UI with an LGPL one.