Post-Dated Musings


I think that much of the point of an open source project should be the creation and distribution of source code from which others can create competitive derivations, with, one might hope, interesting new software, which they can then contribute back to the community, or perhaps more accurately, the commons.

I'm not sure why I haven't posted much since the end of October but it was coincident with my discovering that people I actually knew, like my family--Hi Mom!--were reading my blog. Of course, there were also, at the same time, a lot of things going on at OpenOffice.org, and I had to catch up on work I had not been able really to do while travelling. So continuing what is for me either a self indulgence or a work of marketing and public debate (with a public that turns out not to be public at all) was put in suspension. But it's not that I didn't think of topics.

So here are some of the things I didn't publish on but wish I had, at the time:

The OpenDocument Summit, held jointly by Sun and IBM early November. I was invited to this in my capacity as OpenOffice.org representative, and it was interesting. The summit was held at IBM's Westchester headquarters which are beautiful, and beautifully sited. If you've seen the original Rollerball, the one made in the 70s, you might get a sense of the wealth in evidence there and of the quiet merging of corporate and estate wealth into a Wordsworthian poetics of nature and property, a no end of naturalized poetic property in the service of a landscape ideal. But this is so often the case with landscaped beauties, and the only difference the today brings to yesterday's vision is that instead of it being a family's wealth it's a corporation's. (Arguably, that's an improvement: you can always get hired by and move up, supposedly, in the corporation; the wealth can become yours to enjoy, too. But we all know that's a limited truth. The divide between those who have a lot and those who have some is now vast and pretty much unbridgeable, even if one were to win the lottery.)

The summit had a lot of the theatrical about it. This is not a criticism. Theatre is necessary, especially where systems of power like open source are concerned, and necessary for forming consensus. Theatrical events create the environment for the forging of consensus, involve the players in a space that allows for the staged performance of actions that will later be re-performed in more "real" circumstances: theatre is necessary. (In fact, as a digression, for all its insistence on the real value of code and the real meaning of community as real conversation about code, FOSS is and always has been concerned with the theatrics of its own doings, with the consciousness of its consciousness.) Was the theatre effective? Yes: in the four or so months since, we have publicly worked through arguments about the ODF and publicly identified ourselves as supporters of the ODF. Liaisons were formed, the issues clarified. And for what? A file format. But it's an important issue, for, as I recently described in an interview for a Danish IT magazine, it's really about creating and communicating information without the fear that what you produce, what you exchange, will be incomprehensible to your interlocutor.

(So, what is the OpenDocument format? As I've written before, and will write again, it's the open-standard file format that OpenOffice.org uses. It's maintained by the office technical committee of the OASIS standardization body, which is a corporate and public industry group set up to create and maintain ecommerce standards. What does it mean for the OpenDocument format, which everyone more or less loosely refers to as the ODF, to be an open standard? It means that any one can use it, including proprietary companies, such as IBM's Workplace. Corel had been inching toward adopting it but has seemingly decided to keep with its proprietary standard, I suppose because it really likes being in a tiny niche.)

Prior to going into the summit, I knew, of course, that Workplace will use the ODF in its next iteration. That much is common knowledge. I also know, because I had been shown this a while ago, that IBM had forked OpenOffice.org when it was licensed under both the SISSL and LGPL and that it uses OOo in Workplace. (With 2.0 Sun dropped the SISSL, thus foiling the plans of more than one company, including IBM. SISSL works like the BSD and is thus friendly to proprietary companies wishing to modify code without publishing those modifications. In itself that may not be bad, as the license in effect promotes the fast growth of a market, so arguably open-source projects using the SISSL also benefit from the increased market. This at least was probably behind the reasoning for using the SISSL to begin with: create a large market for OOo and StarOffice. It was surely incidental that Microsoft would lose some market share as a result of that spread. Claims to date of OOo having more than 7 percent of the market for office suites I generally laugh at. But perhaps I shouldn't.) The existence of the fork was confirmed to me by IBM's relevant executive, as was the probable existence of really nifty technology and features (sigh). But as anyone who uses OOo or who follows OOo knows, IBM is not a contributor at all to OpenOffice.org. It is not because, they tell us, of some technological issues and because, it seems, of the license/copyright holder combination. Sun holds copyright to the OpenOffice.org code and this, in conjunction with the LGPL, which forces the publication of modifications to the code, evidently prevents IBM from contributing code to the project and thus community. it doesn't prevent other companies, such as, Google or Intel or Novell, but it prevents IBM. Novell is a also a competitor of Sun, as is Red Hat (remember Scott's harsh words about Red Hat?).

I think that much of the point of an open source project should be the creation and distribution of source code from which others can create competitive derivations, with, one might hope, interesting new software, which they can then contribute back to the community, or perhaps more accurately, the commons. The production of the source should not be held hostage to the invidious politics of corporate antagonism--otherwise, why bother with open source? As mentioned, the continuing contributions by Intel, Novell, Propylon, Red Hat, and now Google, to name but a few, suggest that at least for now a separate IP-holding entity is not necessary: others have not found it impossible to contribute. What is necessary is ensuring that developers are able to find information on how to build OOo, how to contribute to OOo, how to work with the project; and that there is transparency both in how the "leadership" is established and in how it makes policy. We've done a lot of this and are doing more. But as I have expressed in many presentations all over the world, we are by no means there yet. More can be done to make it easier for developers and other contributors to contribute; and we need greater transparency in governance broadly understood: how code and other contributions are submitted, evaluated, acted upon.

With the full development of the Extensions project which Laurent Godard is leading (I'm sure he would appreciate assistance), I expect to see even more interesting developments and just maybe a departure from the static banality of Microsoft's Office. I've long held that OpenOffice.org should be really a source project issuing useful binaries but whose strength really lies in what others can do with the code: add to it, extend it, shape it. The Extensions project promises to start realizing that potential by giving the large community a space and the tools for contributions.

But what about the case of those who would still like to contribute to OOo but for one reason or another cannot or refuse? In a recent interview, I invited IBM to contribute to OOo or at least to publish its code, so that the community (however understood) might benefit. (Okay, my phrasing in the article could have been better.) Shortly after, an IBM executive contacted me, and we had a friendly discussion about how OOo and iBM can work together, even if IBM further declines signing the Joint Copyright Assignment (JCA), which jointly assigns copyright to Sun and the original copyright holder. My guess is that IBM is not going to sign the JCA, though I strongly urge it to and note that it is a joint copyright assignment, not a plain copyright assignment: IBM and any other entity can retain copyright. But let's say they do not sign the form. How could IBM (or any other similarly disposed company contribute to OOo?

#1 publish code, either on its website or on a 3rd-party's (more on this later). The point is to publish the code so that the relevant OOo community benefits.
#2 donate cash or equipment to the project
#3 create documentation or donate documentation its written
#4 help out with mentoring and with presenting on OOo to developers at developer conferences
#5 help out with marketing

Now, let's look at some of these points...

# 1, publishing code on a 3rd-party site. The idea for this (a version of which I proposed to the OOo community in November), would be that an independent site could hold code that would otherwise be unable to be held by OOo. We've had versions of this before: OOoDocs, for instance, holds GPL documents; that's why it was created, in part. Such a site could hold code as a complement to OOo. What's more, the site could promote not only OOo but the ODF, and mainly to endusers; developers could be thus channelled to OpenOffice.org. The site would thus focus more on providing endusers with the site they need and want while also offering a harbour for code that could not otherwise be accepted by the OpenOffice.org project, say because of licensing reasons. Other entities could create binaries of the code held by this foundation and the OOo code and distribute them. This site would by no means compete with OpenOffice.org but would rather complement it.

# 2, donating cash/equipment to OpenOffice.org: seems pretty straightforward: we need both. Sun has been very generous to OOo, as has CollabNet. But we really need a staff to manage the enduser sections and a staff to help with the developer ones, too. To give an illuminating example, Mozilla has a staff of at least five, I believe, to manage the work of being a huge project with incessant demands. We do much the same work with volunteers and with two paid and overworked persons. Of course, accepting funds would be easier if we had a US account (which I'm working on setting up), but even here there are solutions.

# 4, help out with mentoring and presenting OOo to developers. OOo is at a crux right now where we need, really need, to have developers present on OOo at conferences and other events to other developers. The problem is we have a shortage of available, willing, and capable developers who can travel to various places. I can certainly present on how to contribute--I did this last month (March 2006), in Copenhagen, at the LinuxForum event there, and it was very successful--but one would rather have a developer who can look over code directly and concretely and thoroughly speak to developer interests and issues. As much of OOo's growth is anticipated to be in Brazil, India, China, where IBM is quite strong, I think it would be terrific if developers there from IBM (for instance) could speak to local audiences at regional events.

#5, marketing... Well, I think that could be an easy way to start. One could focus at first on the ODF--this is happening--but also move to foregrounding OOo as a developer and development space. I'm not sure that is done now. IBM could also help in other ways, such as funding marketing events. Right now, it is participating in at least two ODF promotion groups, the ODF Adoption TC (OASIS) and the ODF Alliance (I'm a member of both, too), and that certainly helps, but it mainly helps the ODF not OOo, as is obvious.

These are all just possibilities. I think it is time we consider a serious site or entity as described above--something that could serve endusers, promote OOo and ODF, be a harbour for some code, and also earn funds to support the mother project, OpenOffice.org.


Posted: Mon - March 27, 2006 at 05:51 PM          

©