Playing to Win in the New Software Market
Published 12:32, 17 September 09
That's the title of an important report just issued that “represents advice tendered to the European Commission by external experts”. It's not official EU policy, then, but it represents input into the process of framing it, and it's well worth reading.
One striking – and heart-warming – feature is how significantly open source features throughout. Or, more accurately, how it is now simply taken for granted as a major part of the software market:
Models are converging as companies become increasingly active outside of their traditional parameters. So-called ‘commercial’ software providers are developing and offering open source programs and components, while companies typically considered open source are offering commercial software for sale. Further, the two are actively working together to produce software. For example, Red Hat, the best-known open source producer also ships some commercial software products that run on top of its open source platform. IBM offers integrated open source and commercial solutions and even Microsoft, typically viewed as the archetypal commercial software vendor, actually produces products along a spectrum of software models.
Some change regarding the direction of software model choices can be observed. Many companies have tended towards OSS, in particular mixed model companies moving to a fully OSS software model to leverage the full benefits of the model for their strategic growth. Established large companies tend to use OSS components for diversification and to reduce cost in areas which do not determine the differentiator for their business model. For new market entrants the choice of software model is between proprietary, mixed, or fully OSS, a decision intertwined with considerations of development and business model requirements.
The competition between software models has generally led to more choice for the users and has impacted on the way in which the proprietary model is being applied. An example is the practice by some vendors to offer customers the full source code of a solution upon discontinuation of support for a particular software product. This seeks to emulate some of the advantages of OSS for customers of companies using the proprietary model, and while the effectiveness of this particular offer depends upon the particular terms, the example demonstrates the positive impact that software model competition can provide.
Just over six months ago, I analysed a leaked early version of this, which was fascinating for the insight it gave into the manoeuvring going on by the different factions within that group. For alongside obvious supporters of free software, like the FSFE, there was also that well-known friend of Microsoft, the Association for Competitive Technology, and one of the biggest chums of software patents in Europe, SAP.
In the final version, too, both stick out like sore thumbs for their frequently hostile stance to free software. For example, we have little comments like this:
ACT disagrees with the content of paragraph 3.3.8 and argues that RAND standards are functionally compatible with open source, even if they cannot be implemented in GPL code.
RAND – Reasonable and Non-Discriminatory – is a licensing scheme that is fair to everyone, except, of course GPL'd software, since it can mandate equal but non-zero licensing costs that GPL'd software cannot accommodate. So ACT is saying that RAND is "functionally compatible" with open source - which presumably means can be used alongside it - but can't actually be used *in* code under the GPL – which is most of it. So, for all practical purposes, RAND is incompatible with most open source code.
SAP, meanwhile, tries to muddy the waters when it objects to the idea that Europe should give preference to open source code:
Due to the mixed model growth, software vendors are combining open source with closed source, and as a consequence, the line between open source and closed source increasingly blurs. Therefore, any preferences or mandates favouring open source may be harmful for all software vendors including most open source vendors.
For example, if an open source vendor monetizes its open source contributions by selling closed source add-on components and closed source enterprise editions, such a vendor will be discriminated or excluded during such public tenders. This is particularly true when the closed source “enterprise editions” have been productized under a different brand name and thus are not recognized as an open source product anymore. Thus, even though it might sound paradoxal, preferences or mandates for open source may harm open source, because they reduce the opportunities for the contributing open source vendors to get a return on their open source contributions. Therefore, open source preferences or mandates could be counter productive in growing the European software industry.
What this conveniently glosses over is the fact that open source companies that offer closed-source products, or have closed-source add-ons to their open source products, *always* have purely open source versions – otherwise they would simply be a purveyor of closed-source products, and not open source outfits at all. So the only thing that “preference or mandates favouring open source” would do is to give preference to the open source products that open source companies offer, and not their closed-sources ones...which was the whole point.
Strangely, SAP also “disagrees with the following part of this proposition”, even though the proposals seem sensible and non-controversial (a company *against* transparency of the judiciary?):
On top of providing Compliance statements this body could have the following
goals from which Open Source will strongly benefit
- push for ex-ante disclosure on patents
- call for transparency of the judiciary in charge of software IPR rulings
- promote acknowledgement and full integration of alternative IPR modes aside the RAND types by Standards Development Organisation, research projects, public procurement, and public/private European entities delivering IPR-related assets.
- promote alignment of e-procurement processes to ensure the risk of vendor lock-in is evaluated and part of the decision criteria.
- push for Systematic “prior art” research on open source projects as a step of new patent analysis
But in some ways, the best wilful misunderstanding is the following, from ACT again:
On the other hand, Europe has to be wary not to fund OSS loosely under schemes that would continue to result or even increase the problem of third countries being the ultimate beneficiaries.
For a broad range of innovations resulting from labour-intensive and costly research and development, proprietary or mixed models are and will continue to be more akin to contribute to Europe’s competitiveness.
Leaving aside the deeper issue that our mothers always told us that it was good to share, this comment suggest a fundamental inability to grasp the genius of the open source development methodology. The point is, you *want* other people to benefit from your work, to use it, to adapt it – and then, thanks to the miracle of reciprocity built into open source licences – to share it back with you. It's called “not re-inventing the wheel” and “building on the work of others”: it is a demonstrably more efficient way of developing code, and to worry that “third countries” might actually become “ultimate beneficiaries” gets things exactly back to front.
These and other comments of ACT and SAP in the report must call into question whether there is any point in including them in these specialised working groups on open source. If they could make useful comments, based on a proper understanding of the field that they are considering, then they would be welcome. But if their participation is essentially limited to pushing their own antithetical agendas - “Playing to Win” in the narrowest sense - I think it would be better if they just read the final report produced by people with a more helpful attitude.