Arnaud's Blog

Opinions on open source, standards, and other things

LibreOffice should declare victory and rejoin OpenOffice

When OpenOffice went to the Apache Software Foundation I started writing a post about this topic that I never got to finish and publish.

The post from my colleague Rob Weir on Ending the Symphony Fork prompted me to post this now though.

I should say that I no longer have anything to do with what IBM does with ODF and anything related. I’ve changed position within IBM in the Fall of 2010 and now focuses on other things such as Linked Data which I may talk about in some other post.

In fact, I’m now so out of touch with the folks involved with ODF that I only learned about OpenOffice going to the Apache Software Foundation when the news hit the wire. I had no knowledge of what was going on and have no insights as to what led to it. Similarly, I only learned after the fact about IBM deciding to merge Symphony with OpenOffice.

So, if anyone wants to blame me for speaking as a person on IBM payroll, I’m not even going to bother responding. This is entirely my personal opinion and I’ve not even talked about it to anyone within IBM.

But let me say quite bluntly what Rob is only hinting at: It’s time for LibreOffice to rejoin OpenOffice.

LibreOffice started from a sizable portion of the community being tired of Oracle’s control and apparent lack of interest in making it a more open community. I certainly understand that. But now that this problem is solved, what does anyone have to gain from keeping the fork alive?? Seriously.

While forks in the open source world can be a tremendous way of shaking things up, they can also be very damaging. In this case, I think it’s a waste of resources and energy to keep this going. Instead of competing with each other the LibreOffice and OpenOffice communities should get together to fight their common and real competitor.

I know a certain level of competition can be healthy but I’m tired of seeing open source communities fight with each other to their own loss.

I know the fork was painful and people still hold a lot of angst against one another but they need to get over that. They need to realize they would do themselves and everyone else a real service by putting all this behind them and uniting. LibreOffice should declare victory and join forces!


February 3, 2012 - Posted by | opensource, standards | , , ,


  1. It is funny to read this, because my thinking when I read that Oracle was donating the code (and that it was announced by IBM) was that they should have have donated it to The Document Foundation for LibreOffice. But given that IBM never liked the LGPL licensing of the OpenOffice code base and that a deal was stroke with Sun for Symphony, it was just bound to happen.

    The Document Foundation has already welcomed IBM to join. It hasn’t happened. Seriously.

    Comment by Hub | February 3, 2012 | Reply

    • What is funny is for the people who forked to expect the community they left behind to make the effort to join them and act offended when it doesn’t.

      Comment by Arnaud Le Hors | February 4, 2012 | Reply

      • Given the number of community members from the former project who are actually contributing to AOO is negligible, what you are actually sayings there is that it’s time for IBM to stop fighting the community and join in with LibreOffice instead of faking it using Apache’s generous services.

        Comment by Contrarian | February 4, 2012

      • Again, I have no info as to what is behind IBM decision but I imagine that sticking with the original project and preserving the brand name OpenOffice played a major role, along with the fact that Symphony is based on OpenOffice and not LibreOffice.

        Comment by Arnaud Le Hors | February 4, 2012

      • See the Ohloh data and weep.

        Comment by Contrarian | February 4, 2012

      • Weep? You’ve evidently missed my point entirely.

        What I’m saying is that LibreOffice started as a fork of OpenOffice because it was under the control of Oracle. Now that it no longer is, the very reason LibreOffice was started in the first place is gone. So, why not merge back? The fact that more people are working on LibreOffice today makes no difference. I still think it would benefit everyone if everyone joined forces under the original brand name.

        Your comment is unfortunately a good demonstration of the kind of destructive attitude that plagues open source. If I were to weep, I’d be weeping at the fact that your bragging about LibreOffice having more developers than OpenOffice shows that you’re focusing on the wrong competition.

        Comment by Arnaud Le Hors | February 4, 2012

      • This is preposterous.

        The one who left the community behind are Oracle when they decided to abandon the project and IBM when they pushed to get it relicensed so they could still fork it as Symphony without requirement to give it back to the community. Previously they just stroke a deal with Sun to do the same. At that point I wonder why they just didn’t take the code base, but this is probably the most hypocritic part of the story – but obviously, and the IBM have acknowledged it, they have no paid developers working on that code base. Just paid evangelists.

        LibreOffice was embraced by the community. The folks behind TDF made sure the license stayed friendly for the community, that the barrier of entry for the community to contribute was lower, etc. The hard decision to go from the patch set that is go-oo to LibreOffice was the result of several years of frustration with Sun/Oracle in stewardship of the project. Disagreements coming from various sources, including (but not limited to) the copyright assignment that was required, for which IBM with Symphony was the example as to why it was not a good idea.

        Also, read @keithcu post carefully (linked below).

        Comment by Hub | February 4, 2012

      • Doesn’t the fact that IBM is merging Symphony with OpenOffice contradict your speculations?
        In the process IBM is giving back to the community the work result of a whole group of developers that have worked on Symphony. You should check your assertions.

        Comment by Arnaud Le Hors | February 4, 2012

      • I understand you completely. But I also understand that, as an outsider who has no history with the old community, let alone the new reality, you are completely misrepresenting the facts. Both the Apache OO project _and_ LibreOffice are different things from the old community, which is now pure history and cannot be “rejoined”. “Rejoin” is just spin that Rob Weir has been putting about to harm TDF as they go from strength to strength and negate each of his criticisms of them.

        The truth is, the people working at Apache Open Office are a few Sun employees hired by IBM, a few Symphony engineers IBM has drafted in from China and a handful of people from two groups; those who were either in poor standing with the old community and are thus feel unwelcome at LibreOffice, or people who have an ideological bias towards Apache or against copyleft and see an opportunity they can exploit. The point of those graphs was to show you that the new community at Apache, whose only claim to being the “real” community is that Oracle gave Apache the trademarks, would be a tiny fraction of the community even if they were actually historically part of it.

        It is insulting nonsense to talk about the hundreds of people who joined LibreOffice “rejoining” this new, rigged community IBM is running. Whoever put you up to this blog posting inside IBM did you and your reputation a disservice.

        Comment by Contrarian | February 5, 2012

      • I thought I was pretty clear on the fact that my post wasn’t related to my employer in any way. The fact that you’re ignoring that statement and choose instead to promote some big idea of a conspiracy from IBM makes everything else you say much less interesting unfortunately.

        The truth is my post comes from the very simple fact that I think the forks of OpenOffice are detrimental to the adoption of ODF. ODF is the only thing I really care about. I spent a significant amount of my time working on promoting ODF and fighting the OOXML substandard. So I care about ODF.

        I don’t have any such history with OpenOffice, Symphony, or LibreOffice. I never contributed to any of them other than by promoting ODF, and I don’t fundamentally care which one wins among them. My point is that the fight among them is detrimental to the main goal: promoting ODF. And focusing on competing against one another is distracting them from competing against their real competitor: MS Office.

        The world would just be better with only one ODF office suite into which all interested parties put their energy.

        Of course, you’ll certainly say that people involved in OpenOffice should simply surrender and join LibreOffice but the reason I think LibreOffice should join OpenOffice instead is simply based on the fact that OpenOffice is the root of LibreOffice, not the other way around.

        I realize that LibreOffice’s choice of LGPL makes merging back much harder but that’s only LibreOffice’s doing.

        Comment by Arnaud Le Hors | February 7, 2012

      • > The world would just be better with only one ODF office suite into which all interested parties put their energy.

        I don’t agree. IBM’s needs for its customers are so different to the needs of non-corporate users that it’s probably smart for them to have their own fork, and that’s confirmed by the fact they intend to do just that by bundling a great deal of extra software with AOO in an “IBM Edition”. That also means IBM is unlikely to be happy with any other commercial party getting involved at Apache unless they share IBM’s outlook and will play by IBM’s rules under Rob Weir’s direction. I guess that’s why all the other commercial players have joined in with LibreOffice.

        > Of course, you’ll certainly say that people involved in OpenOffice should simply surrender and join LibreOffice

        Not at all. This “surrender” idea is unhelpful and divisive. If IBM wants to continue the fork it has started at Apache with Oracle’s help, that’s just fine. As long as they don’t pretend to be The Only One That Matters. That’s why your blog posting is so troubling.

        > but the reason I think LibreOffice should join OpenOffice instead is simply based on the fact that OpenOffice
        > is the root of LibreOffice, not the other way around.

        No. Sun’s project ended when Oracle bought Sun. There is no sense at all in which Apache Open Office is the “root” of LibreOffice; rather, both have as their “root” the code Oracle inherited from Sun and abandoned. They are peer projects with a common ancestor.

        Here’s the timeline for you:
        * Oracle bought Sun and froze development of
        * The non-Sun community waited for a bit and then a large subset of them created a fork, LibreOffice.
        * The rest did nothing apart from complain they hadn’t been given thrones, as the Ohloh statistics show.
        * Then IBM persuaded Oracle not to just walk away but on the way out to change the license on the copyrights (which Oracle retained) to Apache License v2 and donate the trademarks to Apache.
        * IBM then started a completely new project at Apache which was named Apache Open Office so that outsiders like you would think there was continuity. IBM’s project manager Rob Weir has been a master of spin on this one and has shown us all why standards committees fear him!
        * IBM hired a small percentage of the old Sun staff, and told a few of its staff in China to show up.
        * A few people with grudges against LibreOffice founders joined in (although mostly not working on the code), and a few people with grudges against the LGPL also joined in.

        AOO is thus a derivative of OO.o that’s a rich mix of IBM product strategy, historic grudges in the old OO.o community and ideological grudges that can never heal. The only sense in which its continuous with OO.o is in having a group of people drawn to it who want to tell developers what to do. I doubt that LibreOffice would even want that brew added to their project!

        I understand your confusion, though. Some of the AOO project members have gone out of their way to make things confusing for outsiders. I hope is clear for you now.

        Comment by Contrarian | February 7, 2012

      • We’ll just have to agree to disagree on what would be best I guess. What I can say is that I see the same history repeating itself again and again. I went through this with Unix, with all the various vendors creating their own variant to only help Microsoft Windows prevail. When Linux came out, it seemed somehow the industry had another chance at it but it didn’t last and all hopes were quickly killed by an avalanche of competing distributions, again to Microsoft’s benefit. Now I see it happen with OpenOffice.

        There always are people with “good” reasons to fork and compete to the detriment of the whole.

        PS: It’s pretty funny to me that anyone would “fear Rob”. It shows how things can get distorted on the Internet.

        Comment by Arnaud Le Hors | February 7, 2012

  2. It is clear the author of this article is not aware of the many mistakes that were made in creating the Apache OO fork.

    Here is something I wrote at the time:

    Comment by keithcu | February 4, 2012 | Reply

    • LibreOffice is the fork.
      Quite frankly, I don’t care to know all the details and argue over every claim but going through your post I can see quite a few that are baseless. For one thing, when I joined IBM in 1999 I was part of a department of about 12 engineers who worked full time on an open source project (Xerces), there was no holding off as to how much we contributed.

      Comment by Arnaud Le Hors | February 4, 2012 | Reply

      • Perhaps some parts of IBM are more supportive of free software than other parts. It is hard to generalize about such a large company and perhaps I was being cynical. However, if the “open source” evangelists make such big mistakes, what hope is there for the average IBMer to be doing the right thing? Symphony had a proprietary license for many years. What does that say about IBM? Their proposal should have included plans about Symphony before they started incubation. Their plan ignored Symphony and LibreOffice.

        You said I made many baseless claims, but you only pointed out one offhand exaggeration.

        It is possible for the communities to rejoin. First, the AOO team should switch to the LGPL license. Once that is done, the next steps will go faster.

        In the meanwhile, the LibreOffice community is young but doing just fine, and was doing fine before AOO. If the hacker and contributor community just ignores AOO and works in LibreOffice, which is fully up and running anyway, the wasted manpower will be minimized.

        Comment by KeithCu | February 4, 2012

      • Even Rob Weir agrees that Apache OO is a new project:

        Comment by Contrarian | February 5, 2012

      • That Appache OpenOffice is a new project is undeniable. That it means it’s a fork is clearly not. What do you see when you go to

        Comment by Arnaud Le Hors | February 6, 2012

  3. I wonder if the two communities could find a common space for a dialogue on what is going on and evaluate if there are the conditions to rejoin.

    Comment by Paolo Scattoni | February 4, 2012 | Reply

  4. The LibreOffice code base should be migrated to Apache OO immediately. Any delay is a waste of resources.

    There has been no objective response provided by LibreOffice with respect to the Apache OO gift (and they appear equally as quiet regarding the Symphony merger). Based on KeithCu’s response to Rob Weir, It would appear their wits – while profitable in their initial context – have outgrown them ( If individuals are not willing to consider the options, why should they be taken seriously by the open source community? Distributors – Red Hat / SUSE / Ubuntu – should be asking questions.

    Whatever happened to honour – forks are not plastic toys – software development costs money, and so do monopolies. And in this respect, Microsoft may as well be paying LibreOffice…

    Comment by richardbrucebaxter | February 8, 2012 | Reply

    • @RichardBruceBaxter: You are almost correct: the Apache OO team should join the LibreOffice team and abandon the Apache license.

      There has been plenty of responses by LibreOffice to the Apache OO project in June, on multiple mailing lists, blog postings, articles in the news, etc. I wrote a summary of all the reasons I had heard in a link you can find above.

      The person who is wasting time and money is Rob. He should have made a plan to join with LibreOffice right when he started. Lots of people complained about his plan, but he didn’t bother to come up with a different and better one.

      You are clueless if you think that the distros should be concerned about LibreOffice not joining forces with Apache. If it didn’t make sense for these teams to work together when they got started, what makes it a good idea now? The Linux community is happy with LibreOffice, and inside LibreOffice people are not complaining about a reason to fork.

      Comment by keithcu | February 8, 2012 | Reply

      • keithcu, this appears to be a very sided view of the world. “Go my way and everything will be fine.” 🙂

        Is the whole contention really just about the license? If it is, then there is probably no hope because, as your posts seem to demonstrate, people tend to be religious about it.

        Comment by Arnaud Le Hors | February 8, 2012

      • @keithcu:

        There are great corporations willing to donate their time, money, and source code to develop products where needed to combat the inevitable consequences of monopolistic behaviour – market inefficiency. is one such donation, and so therefore is LibreOffice. Although acknowledged corporate interest must be economical, their support continues unabated. is one of a few open source products which has not been specifically a community development, and it never will be. The corporate world, whose work happens to rely upon the integrity of a common set of cross-platform, cross-application file formats, needs support, and for support to exist it helps greatly to have licenses which do not deter companies from investing in R&D.

        Just because the developers of or its derivatives should earn equal respect for their work, it doesn’t mean that they should write-off the work done for them. The users of this software respect both the code created by corporations, and the code created by the community – and they therefore continue to respect the brand name.

        If anyone seriously believes or its derivatives can take on a monopoly whose behaviour consists of fast tracking ISO standards to convey a sense of interoperability while not conforming to them, by a divided effort, then let the world know. And if this division remains for the purpose of asserting license restrictions, then perhaps even the FSF should shed a tear.

        Regarding happiness, the desktop Linux community at present does not consist of corporations, nor the mainstream populace. This reason for this happens to relate directly to the very issue being discussed. If the Linux community wishes to capture this new market, then it needs act quickly and decisively. If it wishes to maintain its exclusivity, then I would highly recommend division on this front.

        Yet to reiterate – if not economic, this demands a human response – we must respect the code.

        Comment by richardbrucebaxter | February 9, 2012

      • It isn’t all about the license. That is just one that is simple to state. If Rob changes to L/GPL, then it demonstrates his seriousness about joining with the LibreOffice community.

        This is not just about the license because it is also about a person who ignores the community and the feedback to his plans. Why should a person who makes such big mistakes be taken seriously with future plans?


        Rob should have joined up with the LibreOffice community last June. LibreOffice is not turning down anyone’s help. He is rebuilding everything LibreOffice just did. He is trying to get people to work in the Apache project instead of the LibreOffice project. Rob did all of this.

        Please read this as you clearly don’t understand the recent history and get other basic facts wrong:

        Comment by KeithCu | February 10, 2012

      • It isn’t all about the license. […] If Rob changes to L/GPL, then it demonstrates his seriousness about joining with the LibreOffice community.

        So, it’s not all about the license but if only Rob accepted the license you chose that would be a good start? And that’s not one sided? 🙂

        And, talking about getting basic facts wrong, I don’t know what role you think Rob has at IBM but unless he’s been promoted several levels to an executive position while I was sleeping he doesn’t nearly have the power you think he has. 🙂

        Comment by Arnaud Le Hors | February 13, 2012

      • Pardon Keith, I didn’t know who I was speaking to – which is a great honour at that. Please accept my sincere apologies. It is obvious that you have thought about this thoroughly. SW was a great book (v1.028) – well done; the dream is solid. Note I think that the case for office suite development is perhaps a critical exception. This is the binding force of the software wars. With a stable, professional, and interoperable open source office suite, people (companies) will be able to comfortably choose something other than Microsoft Office. In turn, they will have the ability to choose something other than Windows/MacOS, and third party app developers / peripherals manufacturers will start seriously targeting and deploying for Linux. One might even dare to call this a free market.

        Regarding other basic facts however, all I know is that one of the only flagship features of LibreOffice 3.3 (SVG import) did not work, and something is therefore seriously wrong with their QA. clearly had the insight not to retain let alone promote features that did not work. I tested it again yesterday, thinking that I could have been underestimating their capabilities, but to no avail ( – in this respect, LibreOffice 3.3.0+ appears to be in the same state as pre-3.3.0 (

        The arguments you listed on your website regarding Apache seem very reasonable, but I am not yet convinced. In a sense, the question I ask myself is; how are we any better than the model we are trying to replace if we cannot respect people/corporations who try to help us?

        Comment by richardbrucebaxter | February 15, 2012

  5. […] LibreOffice should declare victory and rejoin OpenOffice While forks in the open source world can be a tremendous way of shaking things up, they can also be very damaging. In this case, I think it’s a waste of resources and energy to keep this going. Instead of competing with each other the LibreOffice and OpenOffice communities should get together to fight their common and real competitor. […]

    Pingback by Links 15/2/2012: New Fedora Leader, Debian Lenny Support Ending | Techrights | February 15, 2012 | Reply

  6. [This comment was edited for removal of a personal attack against a third party]

    @Arnaud: It isn’t all about the license, but it is an important one that is easy to state. No point discussing the littler issues like Subversion versus Git until that one is resolved. I do believe that the LGPL license is better than the Apache license in general and for this codebase: ( It is possible to build things like a proprietary Watson on top of LibreOffice with the LGPL. [personal attack deleted]

    Glad you liked my book! I have spent a lot of time researching the industry. Unfortunately, writing everything down doesn’t make anything happen, and new mistakes are made — like this Apache OpenOffice fork.

    I agree that the office suite is a hard problem, but it can be solved like all the others. As it gets more users, it will get more companies supporting it. LibreOffice is good enough for easily 90% of people today. That is a good start. LibreOffice is taking off: The problem has been that the community has been stifled by Sun / Oracle.

    Regarding SVG, often times new features don’t work the first version. In the kernel and other places, new code is marked “experimental” and disabled by default. Unfortunately LibreOffice doesn’t have this infrastructure. It is not a lack of concern for quality, it is a lack of people.

    Corporations can contribute to LibreOffice. Many are:

    Comment by KeithCu | February 16, 2012 | Reply

    • KeithCu,
      I have had this blog up for several years and while there has been some heated discussions in the comment section before I’ve never felt the need to reject or edit submitted comments before. I typically accept all comments, whether I agree with the opinion expressed in them or not because I think debate is good. However, reading your comment made me feel too uncomfortable to accept it. Rather than deleting your comment altogether I resorted to editing off the offending section. I would appreciate if you could keep the tone respectful and stay away from personal attacks.
      Thank you.

      Comment by Arnaud Le Hors | February 16, 2012 | Reply

    • You have your reasons to work on a fork. Fine. But keeping on calling Apache OpenOffice a fork doesn’t make it so.

      I understand your position would appear more tenable if we really had two forks living in parallel but as I stated before: that Appache OpenOffice is a new project is undeniable. That it means it’s a fork is clearly not.

      Again, what do you see when you go to ?
      This is OpenOffice, the one and only, now hosted by Apache.

      Comment by Arnaud Le Hors | February 16, 2012 | Reply

      • If you consider the Sun / Oracle line of code dead because all the people were laid off, that makes this new Apache project a fork. Anyway, it is just a shorthand.

        No problem about deleting my comment. I mostly wanted you to read it 😉

        Comment by KeithCu | February 16, 2012

      • You’re saying “the Apache OpenOffice fork” is a shorthand for “OpenOffice”? 🙂
        I’m pretty sure none of the people involved in the Apache Xerces project I was part of 12 years ago are still involved. All the developers have changed. That doesn’t make it a fork. So, even if that were the case here that would still not make it a fork.

        Comment by Arnaud Le Hors | February 16, 2012

  7. As a school-based “evangelist” for open source software, I could not agree more.

    Having to now promote “LibreOffice” instead of “OpenOffice” makes our job harder.

    It makes open source software seem flaky and unreliable.

    This feeds one of their biggest fears — that after learning one system, they will have to learn another.

    They are still recovering from the trauma of switching from WordPerfect to Word! 😉

    Comment by | July 2, 2012 | Reply

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: