The big news on the open source beat last week was the move of OpenOffice to Apache.
It was a very big deal, although I guess in the Great Game not such a big deal. After all, Steve Jobs is strutting on a stage right now, merging his iOS and OS X Lion platforms (a rather obvious move), while Microsoft executives are fighting to respond by merging Windows with XBox, and Nokia with Windows.
But in open source it was a very big deal, and for me it was a poignant one.
I kept wondering what the folks at LibreOffice thought of this. They're like the wife of divorce whose spouse gets a trophy wife. She's gone on with her life, with a big update, and while her friends can talk all they want about Oracle being a prick, the fact is that OpenOffice — not LibreOffice — has the money and now OpenOffice — not LibreOffice — is going to have the talent, too.
I also know what it's like to lose a job. It hurts. Someone takes what you've built, gives it to someone else, and they get the benefit of all your hard work while you're left in the cold.
I get it.
But we're not talking here about jobs or wives. We are talking about a software project. And to paraphrase what I wrote last week, it's The Code That Counts.
Open source productivity has some steep challenges in front of it.
- How will it deal with the Cloud?
- What about Apple, and the interface changes it brings?
- How about mobile? What's OpenOffice for the iPad going look like?
- Microsoft is still there.
If you want your code to be competitive, it can't sit still. It has to progress. The new features at LibreOffice are handsome, but in the end they're backward-looking. (Word Perfect? Really?) They are useful to some, but they should be after-thoughts.
To compete you move forward.
Apache has some big brains, and while I know LibreOffice does too, big brains think best when they're brainstorming together. That's all I'm saying here.
I know. The Apache and GPL licenses are different. You can't just go back-and-forth between a copyleft and a non-copyleft code base. Maybe that was part of Oracle's intent, though — to break up development in this way so that it could never go forward, just get lost in a pissing match between two groups that should be on the same side, not enemies.
For now, some back-channel communication is needed. Send an e-mail. Have a chat. Share a coffee. Then appoint someone on each side as liaison, and continue the dialog, participating fully in each others' wiki, and reducing duplication of features as much as possible, finally doing some sort of feature exchange so the code base progresses.
What Oracle has put asunder can be put back together, but it will take diplomacy, patience, tact, and a willingness to give a little on both sides. Oracle is betting these two groups can't do that, that the Trophy Wife and the ex-wife can never get along.
Why not prove them wrong?
Word Perfect was a good little program, and it least it didn’t try to tell you how to spell your words and that the phrases in your bibliography were fragments. And when something went wrong, there was a Reveal Codes that showed you what it was.
Word is too automated.
Word Perfect was a good little program, and it least it didn’t try to tell you how to spell your words and that the phrases in your bibliography were fragments. And when something went wrong, there was a Reveal Codes that showed you what it was.
Word is too automated.
I don’t mean to dump on WordPerfect as software. I still do some things using WordStar control codes like ctrl-c to copy, ctrl-v to past and ctrl-x to erase.
But software development needs to look forward, not backward. The focus needs to be over the horizon, or the program dies.
As WordPerfect died.
I don’t mean to dump on WordPerfect as software. I still do some things using WordStar control codes like ctrl-c to copy, ctrl-v to past and ctrl-x to erase.
But software development needs to look forward, not backward. The focus needs to be over the horizon, or the program dies.
As WordPerfect died.
Hi Dana, you point to the features of the older 3.3 version, and not to those of the newer 3.4 version. Some features, I agree, were definitely minor, but were kept away from OOo because of the Copyright Agreement, and have been integrated just because they have been available for years.
You should look at code cleaning and other development activities at LibreOffice as a necessary step for making the software leaner and faster, in order to prepare the next wave of new features.
OOo has accumulated a technical debt over the years, which we have started to recover. Because of the AL all this work cannot be integrated back into OOo. Of course you may think that AL is better than LGPL/MPL, but you know that discussions around free software licenses may be endless (and I don’t want to foster them).
ASF governance is rather clear about money, and thus I don’t think that OOo has got the money now. ASF projects are based on individual contributions, even when people works at corporations.
Anyway, we have started to talk, at least to show that Oracle has forgot some 30.000 files from OOo code base, something that not a single ASF signed committer did notice. This should make it clear to everyone that getting to know OOo code base is not a trivial task to be overlooked.
Hi Dana, you point to the features of the older 3.3 version, and not to those of the newer 3.4 version. Some features, I agree, were definitely minor, but were kept away from OOo because of the Copyright Agreement, and have been integrated just because they have been available for years.
You should look at code cleaning and other development activities at LibreOffice as a necessary step for making the software leaner and faster, in order to prepare the next wave of new features.
OOo has accumulated a technical debt over the years, which we have started to recover. Because of the AL all this work cannot be integrated back into OOo. Of course you may think that AL is better than LGPL/MPL, but you know that discussions around free software licenses may be endless (and I don’t want to foster them).
ASF governance is rather clear about money, and thus I don’t think that OOo has got the money now. ASF projects are based on individual contributions, even when people works at corporations.
Anyway, we have started to talk, at least to show that Oracle has forgot some 30.000 files from OOo code base, something that not a single ASF signed committer did notice. This should make it clear to everyone that getting to know OOo code base is not a trivial task to be overlooked.
Of course, we have not started to talk with ASF to show them that we are better, as could be understood from my previous comment.
Once the incubation project starts and ASF gets all the OOo code from Oracle, we will start working on common grounds, like the ODF file format.
Until then, we will only be able to help on minor issues like the one I have pointed out on the previous comment, thanks to the fact that we know OOo code better than ASF (Oracle core developers, with individual exceptions, are not going to contribute).
Of course, we have not started to talk with ASF to show them that we are better, as could be understood from my previous comment.
Once the incubation project starts and ASF gets all the OOo code from Oracle, we will start working on common grounds, like the ODF file format.
Until then, we will only be able to help on minor issues like the one I have pointed out on the previous comment, thanks to the fact that we know OOo code better than ASF (Oracle core developers, with individual exceptions, are not going to contribute).
I am thrilled to see that y’all recognize the issues involved here and are starting to work them through.
What critics always underestimate in open source is the kind of collaborative attitude I see you taking, and that I’ve seen Apache take. If this were proprietary code, this would not happen. Most money would go to lawyers, and marketers, and PR, not the code base.
Thanks to open source, infrastructure can be shared and more efforts on both sides go into the code, which is what matters.
Good luck, and do stay in touch.
I am thrilled to see that y’all recognize the issues involved here and are starting to work them through.
What critics always underestimate in open source is the kind of collaborative attitude I see you taking, and that I’ve seen Apache take. If this were proprietary code, this would not happen. Most money would go to lawyers, and marketers, and PR, not the code base.
Thanks to open source, infrastructure can be shared and more efforts on both sides go into the code, which is what matters.
Good luck, and do stay in touch.
simply click the up coming article
Dana Blankenhorn: What is LibreOffice Supposed to Do?