Microsoft throws $1m open-source party
A-list guests needed
Industry organizations and foundations are like parties - it's who you don't invite and who shows up that's really important.
IBM and Microsoft once tried to usurp Sun Microsystems' role over in web services by launching the Web Services Interoperability (WS-I) organization pointedly without the Java daddy. Humiliated, Sun had to claw its way in later through the regular voting process.
Further back, IBM created the Eclipse Foundation literally to overshadow and steal - yes - Sun's community stewardship of Java.
Now, Microsoft has unveiled an organization it says will help open source work with commercial software organizations - the CodePlex Foundation. Microsoft is also the Foundation's sole funder, having donated $1m - a number the company will review annually.
The problem with open source today is that existing Foundations target only specific projects, according to the group's site. It wants to address the "full spectrum" of projects, by sharing best practices and to increase participation in open-source community projects.
Why is Microsoft acting? "We saw a great opportunity to drive change," the site says.
As ever, though, it'll be who participates in the group as much as how this particular group answers some major questions that'll determine whether the Foundation moves beyond just a Microsoft talking-shop and achieves some results.
Microsoft wants - and needs - to engage with open source developers and projects for the Foundation to succeed. There are a couple of hurdles en route to winning them, though.
First, there's the name, uncomfortably close to Microsoft's CodePlex forge that it created in May 2006 to host open-source projects. CodePlex is dominated by Microsoft-centric code.
It will make uncomfortable and confusing reading for independents and big vendors to read on the CodePlex Foundation's site that the Foundation is an "extension" of the CodePlex "brand", yet the Foundation will be independent from the CodePlex site. The sharing of the name along with Microsoft's initial funding means people will need to take Microsoft on trust, and that's not something Microsoft enjoys a great deal of in open source.
Microsoft has published the Foundation's bylaws in an attempt at visibility, but it's difficult to see how anybody in the community will see this as anything other than an extension of the Microsoft site, especially given Microsoft's funding and the fact that CodePlex is largely full of projects for .NET and Windows.
Further questioning its independence, there's the subject of the Foundation's inaugural leadership: four of the seven inaugural board members are from Microsoft, with the interim president being Microsoft's senior director of platform strategy Sam Ramji.
Of the remaining three, one is Miguel de Icaza leading the open-source implementations of Microsoft's .NET and Silverlight at Novell, with the Mono and Moonlight projects. Although de Icaza is a decidedly independent thinker, he has taken tons of flak from diehards and refuseniks for Novell's patent pact with Microsoft. This will create further acceptance problems for the Foundation.
The other interim board members come from a .NET project or have a background in community organizations.
To the Foundation's strength, Ramji has gained some credibly among open-source and Linux leaders for his ambassadorial work. Also, he only has a temporary role: he will not only serve only for the next 100 days, but is also leaving Microsoft - for personal reasons.
Incidentally, this might actually serve to weaken the Foundation inside the Microsoft organization, as Ramji has been one of open-source's biggest champions inside the company and was clearly closely associated with the Foundation.
More stumbling blocks
The whole board, meanwhile, will also be replaced during the next 100 days.
All eyes will be on who will succeed Ramji, de Icaza, and the others, and their backgrounds and affiliations
The subject of who is in control is as important as the Foundation's's supposed goals. The group's stated aim is to work with open source projects. The board will pick projects, but on what basis is unclear.
The fact that projects have to be "picked" by anybody will be rejected by independents in the community.
Backs will likely be further put up by some of the claims on the site: that many existing foundations target particular projects. For example, Mozilla might be best known by most people for Firefox, but - as you can see - it stewards plenty more projects.
Other potential stumbling blocks are the license question and the fact the Foundation will accept code donations - to what end is not clear given that the group says it does not wish to compete with existing open-source industry foundations but work with them.
Ramji said Thursday that the Foundation would be license-agnostic, but there is a contradiction: the group will default to BSD yet it will also support any license, he said. The circumstances under how the default is made are not clear. Further, this Microsoft-funded body wants those donating code to do so using a royalty-free license, something opponents will throw back at Microsoft, which is selective in the licenses it picks for its own code.
Grass-roots developers and projects are one thing that could inhibit the Foundation's growth and it's ability to become a force by hitting some kind of critical mass.
A bigger problem is attracting the power brokers that could give it some clout and help drive best practices - companies like IBM, Oracle, and Red Hat that have the money, history, staff and influence to make a real difference in open source and Linux. Like the WS-I and Eclipse, this latest organization is notable for who's not on board: IBM, Oracle, and Red Hat. It would be hard for any Linux or open source foundation to achieve credibly or reach success without their blessing.
Furthermore, these organizations know this and are unlikely to bless an open source effort coming from Microsoft, as they will have ceded the leadership role to Microsoft. Ramji Thursday said the group hoped to attract funding and membership from "other parts of the industry" to gain credibility and dilute the influence of Microsoft.
Time to hold back
Therein lies the catch-22, though: those who matter know this, and are doing very well driving open source and Linux ecosystems around their own stacks. They have no need to join a Microsoft effort, an effort that will take longer to succeed or reach a critical mass without them.
The CodePlex Foundation hits a number of memes in open source: the need for improved best practices and co-ordination within and between projects, and the ability for open source to run well on Windows and Windows to work well with open source code.
Microsoft, meanwhile, has a notoriously rocky relationship with open source. Sections of the organization have tried to repair the damage, but Microsoft has only ever engaged tactically with open source and never joined an independent or accepted open source organizations.
Microsoft needs to join other peoples' parties first before it can convince those who really matter to attend its own. ®