Community Open Source as the Raw Material of Computing Utility Providers

It’s April 2nd, so the Apache Software Foundation’s 2010 April Fools’ joke is over. Here is why I liked it a lot. It represents a hypothetical: What if the ASF and its projects could be bought? Or, if not bought, then put under control or strong influence of corporate interests like in traditional open source consortia? It would put the very software infrastructure we take for granted under partisan control and there is no guarantee that those partisan or corporate interests would be in the interest of the public good.

This is not to say all open source is or should be a public good; much is not and probably never should be. There needs to be ample room for commercially motivated innovation using an open source approach. But I’d rather not have software at the core of the Internet that is owned or controlled in any way by a particular interest group or corporation. The availability of community open source that everyone can build on has sped up innovation on the net to an unprecedented scale. Few software startups these days choose a closed source or commercial open source stack.

I like the utilities metaphor. If software/platform/infrastructure-as-a-service providers are the new utilities of the Internet age, then you don’t want anyone to control major parts of these providers’ raw materials, that is the software they are using for their services. If anyone was able to restrict the access of computing utility providers to their materials, *-as-a-service could be controlled more easily, barriers to market entry would arise, and innovation would slow down. The community open source provided by the ASF, the A in LAMP (and much more), is a major part of those raw materials, and so I’m glad the April Fools’ joke was just that, a joke, but with depth to it.

3 Replies to “Community Open Source as the Raw Material of Computing Utility Providers”

  1. then you don’t want anyone to control major parts of these providers’ raw materials

    Saw your Tweet earlier about the Apache April 1st takeover, but alas their blogs are actually down now (unbelievable — 502 Proxy on an Apache site. Oh irony), so I’m going out on a limb here without reading the entire blog post.
    As I recall it, a low monetary value was attached to the supposed takeover. Is this perhaps only half in jest? i.e. non-copyleft FLOSS has a different way of being captured than their copyleft counterpart. Practically nobody owns the IP for the former, so commercial companies are free to poach but disinterested from taking ownership (unless there is the possibility of buying up the entire brain trust without annoying outside developers and users). With the latter, on the other hand, copyright assignments (cf. MySQL et al.) mean that large-sum takeovers are more likely.
    Compare and contrast: PostgreSQL and MySQL, Apache and JBoss, LPRng and CUPS (all are reasonable to compare as they are all open source practically from the start, unlike, say, Java).
    Even with copyleft, one could, of course, fork, as is the case with MySQL, but then one is held hostage to either the current license (if the GPL is applied without the “or any later version” provision) or to the FSF (if the provision applies. This is normally less likely in commercialized OSS though, I suppose because copyright assignment means the IP owners can always relicense at their convenience, not the FSF and their users!)

  2. Hi Michel, it’s “a $1.5B cash deal” according to the blog entry, so I don’t think they attach a low value to it (though some may argue it should be much higher!).
    You are of course right—an open source license gives some protection (but not all). Trademarks and patents can come to bite you. The trick is to be the one organization that holds all the rights (not necessarily ownership but at least grants and permissions). However, a “new owner” of the ASF could significantly derail the projects, and in the disruption of moving to a fork and clarifying IP rights and reconstituting communities, a lot of damage could be done.
    When I wrote the above piece, I was more wondering about the nature of community open source in the age of the Internet. As I said, I’m playing with the metaphor of it being the raw materials for computing utilities and startups alike. How would innovation suffer, if you could only build your stack from commercial open source? Or only from closed source products?

Leave a Reply