OSNews

In the business world, this move is called divesting.

We just have to be happy he didnt let it die.

Well I'd say we should be happy there are enough developers who won't let it die. Larry couldn't have affected that, especially after the fork.

Actually, I am a little annoyed that they did not.

LibreOffice has really taken off. The weaker OpenOffice.org is the better. Now LibreOffice is hampered by the OpenOffice.org anchor around it's neck.

I think this "donation" was done out of spite on Oracle's part and to advantage IBM (who use OO.org as the basis for Symphony). It has nothing to do with the community or the users.

IMHO.

Edited 2011-06-06 14:46 UTC

No. Not if they aren't making money from it.

Even if they could make money off of it, it wouldn't be enough for them to be interested in it. It just doesn't fit their overall business strategy.

As much as I loathe Larry Ellison and Oracle, he's no fool when it comes to sound business choices. Pity it took this long for Sun's assets to end up in competent hands.

It is true that it was hard to figure out exactly what Sun was up to in their later years. They went on a massive buying spree of buying up smaller companies, including ones that couldn't possibly benefit them financially, such as StarOffice. The bought up Cobalt Server appliances, tried to sell their server appliances for a little while, and then just discontinued them. They bought up Virtual Box, which was another one of those "why?" moments. Virtual Box is useful for desktop virtualization, but it's not exactly enterprise level server virtualization type stuff.

Sun ended up going massively into debt by going on this buying spree, and the companies it was buying simply didn't seem to provide a lot of benefit to Sun's overall business strategy.

Their whole "open source" strategy was baffling from the beginning. It seems like they wanted the buzzword cred but had no clue how to make the process work.

at least they had one.

Virtualbox would have been an interesting buy if Sun had continued the bare metal project Innotek had going and expanded it.

Everything Sun did could have been interesting if they had any vision about where to go with it.

No. Not if they aren't making money from it. "

Libreoffice and Openoffice both can operate as services providing document conversion. This is a feature of value to Oracle. Now its not worth fighting with the community and ending up poorer quality document conversion in other products Oracle sells.

So yes their is a reason for Oracle to remain interested. The OpenOffice program does allow Oracle to provide customers with better products.

When you start pushing people, don't be surprised if they fall.

Oh I can tell exactly what they gained.

First, they gained a permanent seat on the JCP (Java Community Process Board), which gives them more infuence over the future of Java. And Oracle's entire enterprise application stack is built on Java.

Second, they gained the Solaris operating system, which has long been the "gold standard" for running high transaction / high availability Oracle databases.

Third, they gained control of Sun's hardware business.

What they hope to do with all this is simple. They want to become a full solution provider for all your data needs. They will probably build "database appliances" with zero maintenance and such. Just shove your data into it, pull it back out, and you are good to go.

Edited 2011-06-01 21:38 UTC

Don't forget they also got MySQL, and probably more importantly the customer base from Sun as well. The patent portfolio is also nothing to be sneezed at

Edited 2011-06-01 21:52 UTC

True. MySQL might be important to their small business strategy. Small business don't run Oracle after all. Too expensive, and also largely too complicated for their needs.

They gained also the right to sue Google for Dalvik that is similar to Java virtual machine for mobiles phones.

Why would they need to merge back?

There's nothing inherently wrong with a fork.

Two communities developing two almost exactly identical pieces of software instead of just one; makes both communities weaker and makes progress slower.

[Citation Needed]

But seriously.. I don't think that's necessarily true. I think that's a notion that people come up with because it makes them angry that there are two "competing" projects.

With proper source control, sharing patches between the projects could be a no-brainer.

Time and again, it's been proven that competition is healthy.

Edited 2011-06-02 04:15 UTC

That's usual in GNU/Linux/FOSS world.

How many widget toolkits are there? How many desktop environments? How many distros? How many pieces of software that do the same thing? People like for sure reinventing the wheel.

And how is it different from proprietary?

There's more than one proprietary DB implementation, in games more than one FPS, MMORPG, many raster graphics programs, C++ compilers, operating systems, Media players, etc.

Add to that there is a lot of dead wood lying around everywhere. the code in OpenOffice will be picked upon for LibreOffice. LibreOffice will be the fork of choice until some other fork comes along and Libre becomes dead wood.

That's usual in GNU/Linux/FOSS world.

How many widget toolkits are there? How many desktop environments? How many distros? How many pieces of software that do the same thing? People like for sure reinventing the wheel. "

That's not a valid comparison: there is only toolkit with the Qt API and toolkit with the GTK+ API. There's OpenMotif and there is/was lesstif but that was not a fork.

There some different office packages already but, while there are some specific variants of OpenOffice, the only real fork is Libreoffice. This will cause a lot of repeated efforts until the projects diverge (which is admittedly not that bad). The *real* downside, of course, is that 100 people were laid off by Oracle after libreoffice forked.

The license may become an interesting difference though. I would see why IBM is more interested in keeping the code under a non-copyleft license, and others may follow.

All in all, I like the forking idea: there was no advantage in giving the code to libreoffice since they already have it with the license they want and they are not asking for code attribution so there was nothing to gain by giving the code to libreoffice.

That's usual in GNU/Linux/FOSS world.

How many widget toolkits are there? How many desktop environments? How many distros? How many pieces of software that do the same thing? People like for sure reinventing the wheel. "

That's oversimplifying. Forks happen for a reason or they die. If that reason is gone then a fork is pointless. This is different from competing projects that do not share a direct lineage to each other. Often those projects have different goals in the first place, although they generally overlap in some places.

The Apache 2.0 license is a liberal non-copyleft license. AFAIK this means that both open source code and closed source code can be contributed to an Apache 2.0 license project.

LibreOffice is licensed under LGPL v3. This copyleft license means that no closed-source components can be accepted.

So from now, open source code contributed to ASF OpenOffice can be adopted (and re-licensed as LGPL v3) by LibreOffice, and LGPL v3 code contibuted to LibreOffice can be incorporated into ASF OpenOffice (but it must remain LGPL v3, and copyright attribution must remain with the original authors).

I don't think the corporates (Oracle and IBM) want the latter to occur. I think they want the ability to make all or prat of ASF OpenOffice closed source. I'm pretty sure they don't want any LGPL v3 copyleft code where the copyrights belong to individuals.

Therefore, IMO, no re-merge is likely to be accepted by the ASF OpenOffice crowd.

ASF policy is not to incorporate LGPLv3 source as LGPLv3 isn't AL 2.0 compatible ( http://www.apache.org/legal/resolved.html ). The same holds for closed source code; it would need to be under a compatible (ie: liberal) license.

LibreOffice can incorporate AL 2.0 code. ASF OpenOffice can't incorporate LibreOffice LGPLv3 code.

Note that the copyrights belong to individuals on Apache projects too. Any issues for users (the 'corporates') are likely to be around the LGPLv3 licensing; both the copyleft issues and the patent terms.