Sakai Newport 2007: The State of the Union

Six months ago, following the Amsterdam conference, I was highly encouraged by some of the signs of progress I was seeing in the Sakai community. In an exchange with a commenter on that post, I wrote,

The question I’m trying to answer in this blog post is, given these sorts of concerns, how much progress are we likely to see in the next 12-24 months? This turns out to be a difficult question to answer. How does one gauge the health and vitality of any community? My experience has been that an open source project can change direction and take off like a rocket, drop off a cliff, or drift aimlessly, and that this change of direction can happen with little advance warning to even careful outside observers. The leading indicators of a change in direction or velocity tend to be relatively intangible, having to do with the alchemy of the particular inter-personal and inter-institutional relationships that drive the project. They are subtle enough that you often won’t pick them up if you don’t hang around and observe over a significant period of time. For example, if I hadn’t been at the Atlanta conference six months ago, my impression of the Amsterdam conference probably would have been different (and closer to yours).

My observation as somebody who is not quite an outsider but not quite an insider to the community is that there have been subtle but extremely important shifts in the community dynamics. If I am right, then we should start seeing more tangible indicators of this change within the next 6-12 months and real progress within the next 12-24 months.

So, six months later, did I see “tangible indicators” of change? The answer is unequivocally yes.

Let me start by assessing the same dimensions of progress that I assessed last time:

• Commitment to transparency: Nearly all Foundation Board interactions happen on the public lists now. (The main exception is conversations with the Executive Director on a small number of financial decisions are made, particularly in cases where private information such as employee salaries is discussed.) No decisions are made at the conferences; instead, consensus proposals from conference meetings are raised on the listservs so that non-attendee community members can have input. There were also some conversations about improving and clarifying the formal community decision-making processes, although I missed those conversations. (I assume there will be something posted soon, consistent with the policy I just mentioned.) Since openness doesn’t help if you can’t actually find the information that’s been put in the open, efforts are being made to improve accessibility to information as well. A revamp of the public site is under development, and a reorganization of the listservs is under consideration.
• Increasing sense of ownership from non-developer stakeholders: There was a significant increase in attendance from teaching faculty. I don’t have a real count, but I’m guessing it was about 10-15% of attendees. Perhaps the biggest news was that new Board member Josh Barron organized a two-hour session devoted to planning more faculty involvement and focus on teaching and learning. (Again, I had to miss about 75% of this session. I’ll post details of the outcome when I know them.) The Fluid team is having direct impact on several projects and indirect impact on a few more. In general, they are doing an excellent job of cementing the role of UX designer in the development process. Also, there seemed to be more vocal participation from the system administrators, those unsung heroes that keep the trains running on time.
• Openness to change: This is hard to quantify, but it was very clear to me that the community has taken a strong pragmatic turn. One good example is in JSR170 adoption. There was nothing but support for the notion of (over time) cutting out a big chunk of Sakai code in favor of direct use of Apache Jackrabbit. I don’t think this could have happened two years ago. Also, Sakai will be participating in next April’s JA-SIG conference, which promises to be something of a love-in for higher education open source and standards groups.
• Commitment to usability: See my earlier comments about Fluid’s growing (and very welcome) influence on Sakai development. I’m starting to see independent tool development projects actually reaching out to the Fluid team for help. That’s a very significant cultural shift. Also, usability is one of the three priorities of the Sakai Foundation for 2008. I’m told that there will be modest improvements in usability in the 2.5 release (due out Real Soon Now) and significant improvements in the two releases that will follow.
• Commitment to release quality: In addition to usability, the other Sakai Foundation priorities for 2008 are reliability and scalability. Foundation resources are being redirected toward QA, and conversations I’ve heard regarding development priorities of contributing institutions have focused heavily on improving quality through better testing and architectural refactoring.
• Commitment to standards support: I’ve mentioned JSR170 repeatedly. Watch this one over the next 18 months, because I think it will be a leading indicator of how this standards support will improve the value of the platform to the end users. Oracle and Unicon are working on support for IMS Enterprise Service v2. (At some point in the not-too-distant future, I’ll have a lot more to say on this subject.) Andrew Petro apparently demonstrated a proof-of-concept for swapping in JA-SIG’s CAS (Central Authentication Server) to replace Sakai’s (somewhat problematic) own authentication capabilities. There is also interest in evolving the component manager in the direction of the OSGi standard. I know that a lot of this is probably meaningless to non-technical readers, but I do believe it will all have significant long-term impact on the real value of Sakai as a teaching and collaboration platform.

I would call all of these “tangible indicators” of progress. Let me add a few more:

• More adoptees coming: While there were Sakai shoppers at the Amsterdam and Atlanta conferences, my impression was that many of them were just kicking the tires. In contrast, I heard a lot more serious talk about adoption in the imminent future at this conference. Expect to see announcements over the next 12 months, and expect a fair few of these to be international.
• More support vendors: There were at least three new vendors providing some form of Sakai support. These new entrants do not seem to be coming at the expense of the existing support vendors. A healthy commercial ecosystem is beginning to develop.
• A trend toward humility: Several times during the conference, I had the somewhat dislocating experience of actually being more optimistic about Sakai’s future than the long-time community members I was talking to. The increased number of production deployments in the past year (and the problems that came with them) has clearly pushed the community to get real in a hurry.
• Increased awareness of the coming of the post-LMS: Several community members expressed an understanding that we are moving beyond the LMS box and a desire to see Sakai support these new trends. The idea even got a brief reference in Michael Korcuska’s talk about the direction of the Foundation.

I think we are on schedule for my original prediction that Sakai will show “real progress” in its released product within 12-24 months of the Amsterdam conference. In particular, the release after this coming one, which some in the community are beginning to refer to as “Sakai 3.0”, represents several steps down the road of an ambitious refactoring of Sakai with all the right priorities in mind. If the community pulls it off, I think Sakai will look very different–and better–going forward from 2009.