The Right Way and the Wrong Way to Produce a FLOSS Office Suite

“Sun’s open-sourcing of StarOffice Suite is the single largest open-source software contribution in GPL history and it adds a key application suite to the open source portfolio,” said Marco Boerries, vice president and general manager of webtop and application software at Sun Microsystems, Inc. “This announcement will leverage StarOffice Suite’s role as the leading productivity suite for multiple platforms and will further increase our momentum as the only competitive alternative to other office suites. Sun will also continue to develop and aggressively market future versions of StarOffice Suite, which will be based on”
“Since innovation happens in many places, making the source code for StarOffice Suite available will enable the enormous community of developers to bring their expertise and energy to improve and expand the reach of this full-featured software suite. But is much more than just source code–it also includes APIs and XML-file formats. This will allow developers to easily work across platforms and have a huge impact,” said Boerries.


That worked, more or less, until Oracle tried to stomp all over FLOSS developers by contribution agreements and stacking the management of the organization. Then LibreOffice forked and thrived, doing it the right way inviting widest participation. Now, Oracle, seeing that their methods were feeble by comparison, attempts to thwart LibreOffice by converting the licence to ASL. I am also shocked that IBM, who derived their own product from, does not see clearly enough the benefits of FLOSS development that they support ASL over GPL along with Oracle.

I suppose IBM feels that keeping its own product closed source/non-free is in the best interests of IBM but I think they are short-sighted. The reason Linux, ODF, Android, and so many other open technologies work so well is not because of the individual contributors but because the FLOSS community is the right way to produce software. Any non-free implementation is liable to be more limited, buggy and frail than the FLOSS equivalent.

Attempting to put under ASL is nonsensical in the long run. Either or both of these things will be true:

  1. It will cost IBM more to develop.
  2. LibreOffice will eventually become a superior product.

Oracle and IBM’s handiwork will be to delay the inevitable triumph of LibreOffice. You cannot do better than FLOSS, no matter how much you try because the world is a bigger organization and can produce its own software. I don’t see LibreOffice and getting back together now. The die is cast.

Bob Sutor wrote, “In some sense, the value of a standard is proportional to the number of people who use it. An Apache implementation of a standard means that software, be it open source or proprietary, can start using the standard quickly and reliably. An Apache implementation of a standard immediately increases the value of the standard.” That assumes that more people will use FLOSS+proprietary than merely FLOSS. That is not in evidence as is widely used on that other OS and there are billions of people about to enter the Age of IT and are extremely sensitive to cost. FLOSS has explosive growth behind it as well as ahead of it. Oracle dropped the development of as soon as the FLOSSists left camp after all.

see Mark Webbink’s summary of the situation on GROKLAW – CLA Redux – The Donation of OpenOffice to the Apache Foundation

Read some of the comments. Many different views are evident.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology. Bookmark the permalink.

9 Responses to The Right Way and the Wrong Way to Produce a FLOSS Office Suite

  1. Depends on what you mean by largest. Say it’s megabytes of code:

    Package: linux-image-2.6.32-5-686
    Priority: optional
    Section: kernel
    Installed-Size: 76296

    Priority: optional
    Section: editors
    Installed-Size: 24256

    Priority: optional
    Section: editors
    Installed-Size: 2580

    Priority: optional
    Section: editors
    Installed-Size: 18592

    Priority: optional
    Section: editors
    Installed-Size: 3968

    Priority: optional
    Section: editors
    Installed-Size: 121724

    So, indeed, has more code.

  2. Dann says:

    “Sun’s open-sourcing of StarOffice Suite is the single largest open-source software contribution in GPL history and it adds a key application suite to the open source portfolio,”


    The linux kernel and/or GCC/G++ compiler are the single largest open-source software contributions in GPL history.

    Do these PR guys even know what they are supposed to be talking about?

  3. That’s true unless IBM imports something more current.

  4. Contrarian says:

    I found this at

    “Symphony is available for Linux, Windows, and Mac OS X. It is based on Eclipse Rich Client Platform from IBM Lotus Expeditor for its shell and 3 for the core office suite code. version 1.1.4 was dual-licensed under both the GNU Lesser General Public License and Sun’s own SISSL, which allowed for entities to change the code without releasing their changes. Therefore, IBM does not have to release the source code of Symphony”

    It seems to me that IBM wouldn’t care one way or another in so far as their own business is concerned, since they can do what they want anyway.

  5. has been growing like topsy. Many folks are running it on that other OS and almost everyone using GNU/Linux uses or LibreOffice. I don’t see any change in that. The fact that IBM is backing the move to ASL is all the proof I need that IBM is making tons of money from the codebase. IBM does not want that cash-cow to dry up.

  6. Contrarian says:

    I think this is all a tempest in a teapot. I cannot imagine any company making an entry into the office software business with a commercial product that is based on OpenOffice code with improvements or not. OpenOffice has a rather poor acceptance rate, not because it is a poor performer but because it cannot compete with the decades of user exposure to the Microsoft brand and the almost universal use of MSO in commercial situations.

    Differences between OO and MSO are just fly specs and pepper sorts of things and no amount of innovation is going to change that today. If some company actually found a need to make some useful code modifications to OO, they would be able to use them themselves in any case, GPL or ASL, since they would not be distributing if they did not feel like it. The ASL might actually encourage distribution since a company could do it with no cost of meeting the GPL requirements and no fear of lawsuits.

    When you distribute something with GPL coverage, even as a mere reseller, you expose yourself to the harrassment sort of lawsuits being brought by the FSF against companies like Best Buy who are handling the devices with the Busy Box utilities embedded.

    The spirit of FOSS is met with the ASL and anyone who is motivated to defy that spirit has plenty of ways to get around the GPL. The ASL just adds another freedom, that is the freedom to not participate, to the rest.

  7. The ASL does not require modifications that are distributed to be revealed in open source. That is, the source can be taken closed, which I find strange for a FLOSS licence. Why even talk about open source when it will not be provided? An organization receiving source code at no cost should be willing to reveal source code derived from it, IMHO. IF IBM or whomever wanted to hide their source code, they should write it all by themselves. Arguments that no one is harmed by hiding source code are false because the original authours are wanting their contributions to be public for their own reasons, perhaps to gain feedback from users or to gain notice in the industry.

    The ASL and Apache web server are a better fit because the product provides a standard service on the wire and is not likely to be customized. is a complex app with a complex UI with plenty of opportunity to customize. It makes no sense to allow contributed code to be hidden. It makes more sense to reveal source code for and indeed to send changes upstream so that the feed includes them for the future, reducing regressions, and allowing others to tweak the tweaks making them even better. Plugins can be allowed by using LGPL but the main source code should not be hidden.

  8. Eric Yeoh says:

    @Ray – essentially it means you can take the code, do whatever you want and not contribute back. While I treasure the immense investments IBM has made over the years to FLOSS, I cannot help but feel that IBM is just trying to harness free labour to create Symphony which a very pretty product with low compatibility with MSO docs.

  9. Ray says:

    The Apache License is a open-source/free software license, and is used in many Apache products like Apache web server. What’s the problem with it?

Leave a Reply