Planning for Sakai 2.10 and beyond

I believe it is time for the Sakai community to embark on a proper planning and management process for the Sakai CLE, to take us well into the future. Not one that is based on existing Jira’s that are left hanging, not one that is based on just fixing a bunch of bugs and calling that a release, but one that completely realigns the focus of the CLE and brings the product back into the spotlight within the LMS market.

The Sakai Technical Coordination Committee (TCC) was formed in 2010 because there was no suitable technical or functional leadership for Sakai. We took control of the technical aspects, and have so far done an excellent job in getting bugs fixed, features implemented and releases cut.

However, I believe that at this point in time, what the CLE needs is to regroup and refocus on what it is what we want to achieve. Focus groups, steering committees, surveys, evangelists and visionaries, connecting with the people that have the skills, that make the decisions and that have the funding. In my personal opinion, we lack direction, the product is becoming stale and we are at risk of institutions and organisations leaving the community altogether.

As Michael Feldstein writes, the Sakai project is at a critical inflection point. We need to seize the opportunity to re-evaluate the LMS/CLE market and see what it is that teachers and students want from an environment, and how we can deliver that.

None of this will be easy. We don’t have dedicated funding to hire developers, nor the group of people as I described above, to discuss where the product needs to go. The community source model works, previous years of the Sakai project has proven that, but it falls down when the product gets stale and momentum is lost. It cannot always be left to volunteers to pick up the pieces.

What we do have though, is a solid product that is in use by millions of people every day and a thriving community of implementing institutions and commercial partners, all of who want the best for their students and clients. This is the perfect time to reflect on our product and start re-imagining what needs to be done.

Advertisements

12 thoughts on “Planning for Sakai 2.10 and beyond

  1. Steve, I am trying to understand what problem you are trying to solve. The OAE is at an inflection point but the CLE is not at an inflection point. The CLE has excellent governance and roadmaps and activities that are well-aligned with its available resources. The CLE has a strong and vibrant adopter community. I think that with Sakai 2.9 we will see a product that can hold its own with *any * LMS in the marketplace. The CLE does not need “Focus groups, steering committees, surveys, evangelists and visionaries, connecting with the people that have the skills,…” – all of these ideas have been tried and all they do is sap energy from the resources we do have and lead to finding ourselves once again paralyzed and devoid of energy with no apparent way out.

    The current CLE governance and Samigo governance is open to all that want to truly contribute and willing to listen to any opinion from any stakeholder. The CLE efforts have an 18-month roadmap already and are moving slowly and deliberately. Throwing in a bunch of non-resourced brilliant ideas with no additional resources will only cause frustration.

    We are *finally* after nearly ten years working on some of the critical core issues in the Sakai CLE code base to make it more solid and scalable. We can and should address usability issues along the way – but we need to be careful to put our energy into things that we can actually accomplish with the resources we have.

    Having a focus group where we ask 50 people what would you like to see in the Sakai CLE would be a complete waste of time – the resulting list of things would be impossible to achieve. If on the other hand we took 50 long-term teacher and student users of Sakai and got them into a room and asked them to tell us the top-10 within-tool usability problems – we might find a set fo things that could make the CLE a better product *within* the recourse constraints we have.

    History has shown that it is *not harmless* to ask folks a wide-open question about what they want and then when the resulting requests are impossibly difficult and so they never get done. Be careful what questions you ask…

  2. Thanks for the comment Chuck. The problem we need to address is that Sakai is losing ground to other LMSs and we need to re-establish ourselves. Perhaps inflection point is not the correct analogy, plateau is perhaps a better word. I’m not seeing the drive and vision that people once had.

    People DO find it difficult to get started in development, they DO find it difficult to get an install off the ground and more and more code and awesome new features are being locked away in local/vendor repositories and not contributed back to the community. These are huge barriers to adoption and to progress.

    The CLE does have a fantastic community, and that is why I devote so much of my energy into this project, but there is no CLE roadmap beyond the current release. At the conference we talked about issues that should be addressed for 2.10, but the large majority of them are technical issues that we, as developers, would like fixed. The closest thing is the wish list that came out of the post conference meetings here:
    https://confluence.sakaiproject.org/display/TCC/TCC+Atlanta+2012+Conference+Meetings

    This isn’t what I am talking about.

    A crucial part of what I wrote is this:
    “We need to … see what it is that teachers and students want from an environment, and how we can deliver that.”

    We need to ask the right people, the people we want using our products, what they want from this system – included in that can be their top ten annoyances/usability problems. That sort of constructive feedback is SO valuable, but it doesn’t happen often enough. Developers rarely know what users want unless they ask. The survey that the TCC sent out earlier in the year asking what version of Sakai people are using/will be upgrading to was a perfect example of engaging the community in a simple and unobtrusive manner, yet yielded vitally important information to help plan releases. We need more of those.

    I still firmly believe that we need some form of management and guidance to drive the requirements gathering process and move the CLE forward. A lack of management is just as bad as being over managed in my opinion. The TCC does a great job at handling the low level technical details, but it takes a cautious approach which needs to change. And there needs to be a separate structure which guides the top level requirements, and ideally, provides resources to get them implemented.

    Going forward, what I would really love to see is a period of consultation, requirements gathering and community engagement, an extended period of active development to get new features in and old code removed/rewritten, a version bump (reclaim the 3 – CLE 3), incremental minor releases of 2.8 and 2.9 to keep adopters happy, to go mad on test fests and file/fix as many jiras as possible, and then to unleash a re-energised, stable and scalable CLE with a beautiful and consistent UI.

    • David Roldán Martínez says:

      I strongly agree with Steve. CLE is at a critical point. CLE as it is now cannot compete against other LMS. Though releases has a well-established roadmap I cannot see a product plan focused on what the market (teachers, researchers, students and so on) are demanding.
      Of course, changing this won’t be easy because it involves that we must change our minds, paying attention to other criteria that, until now, seem to be in a secondary place. We must focus community efforts in improving existing tools instead of developing a completely new XXX2. This is a waste of time, resources, efforts, ilusions…and, to top it off, it makes CLE to be less competitive. If we want to keep growing, we must think in global, we have to give SCA arguments to sell Sakai…
      But..no panic. We still have time. The ground question is if community really wants to react and conquer the market.

  3. Steve, Its funny – I agree 100% as to *what* you are saying should happen. But I disagree 100% with *how* you propose to do it. First you are implying that the TCC is incapable of doing anything other than low-level technical tweaks and that some *other* group needs to be constituted to help bring some “big thinking” to the CLE.

    We already tried that once:

    https://confluence.sakaiproject.org/display/MGT/Sakai+Product+Council

    This was a *complete* failure and its only tangible output was hot air, confusion, and delay. It was a perfect example of trying to give power to people who had no skin in the game but had lots of opinions. Opinions without resources are worth exactly nothing – and if you give such a group de-jure authority over anything – they become a drag on scarce resources.

    History has shown that these “hot air” forums only cause pain.

    The TCC can and will work through the kinds of issues needed to get through 2.9, 2.10 and perhaps even once we are well into 2.10 contemplate reclaiming the “3” – lets just not get ahead of ourselves and lets not delegate such an important decision to a panel of “non-contributors”. Talk is cheap.

    • By its very nature and the T in the name, the TCC should stay oriented to technical decisions. The things that need to be discussed going forward are not technical though, they are pedagogical, green fields problems. Sure, people within the TCC are capable of this, but the TCC as a whole should not be muddied by this task.

      Profile2 went through the Product Council process and what came out the other end was a better application. The Product Council failed for various reasons, but I don’t think it is fair to write off the whole review process as it applies to software. We absolutely need people in the community that are contributing, to be the ones that are the ‘managers’.

      • Steve – the “T” in TCC was there because it was formed while the ill-fate Product Council was still in existence so that the TCC could function “under the radar”. In reality the TCC was always intended (and has always functioned) more like an Apache PMC. We don’t need two groups – we need to expand the remit of the one group.

  4. smkeesle says:

    My impression of the Sakai community has always been one that sought to scratch its own itch. The CLE is what it is because of the vision (and limitations) of the community that funded, coordinated, designed, implemented and tested the product. Chuck’s comments reflect a pragmatic approach to continuing the development of a tool for the people that use it. The Sakai community is “big enough” to continue to sustain itself.

    However, it seems that the Sakai community has often had a hard time attracting new contributors to the development of the core product. Isn’t that true? Something about this post makes me wonder if some in the community (like Steve) see a longer term problem that needs to be addressed. Steve says, “we are at risk of institutions and organisations leaving the community altogether”. That is significant statement since the value proposition of community-sourced projects is “shared risk”. When someone leaves, this means everyone else is left taking on a little more. At some point, that becomes a challenge for the decision maker who needs to compare their ROI in Sakai versus the alternatives.

    So, I agree that Sakai isn’t at that inflection point TODAY (if the trend inferred by the data presented on Ohloh is accurate: http://www.ohloh.net/p/sakai). But if there is dissatisfaction with the core product, then a tipping point may occur at some point. What leading indicators does Steve see that make him predict this risk being realized?

    • Zach Thomas says:

      I can think of a leading indicator: look at http://www.stanford.edu/online/courses/ There are no fewer than three platforms in use at Stanford for delivering online courses, and none of them is based on Sakai technology. Now, there can be many reasons for this, some political, some technical, but there is a tidal wave happening in ed tech right now, and we (longtime Sakai contributors) are not surfing the wave.

      • IU is another: http://next.iu.edu

        My fear is that institutions that have withdrawn from OAE or even those that are evaluating the market, are unlikely to choose the CLE in its current state, especially given the bad press it has been given by both those pushing Sakai3 since 2009, and the recent withdrawal announcements. The CLE needs to be a compelling and beautiful product, which is what my post is all about making happen. Incremental minor improvements are too slow to effect any major change which is what we are currently experiencing.

  5. No doubt IU and Stanford’s ed tech leadership are piloting all sorts of products as experiments. Due diligence is always a good practice. You best expect it even if your product is really good! There is always a niche that won’t be served by your “enterprise” platform and you need to be responsive…Is there something else that predicts the “brain drain” of the community?

  6. The issue Steve is raising is really orthogonal to governance. The question is whether the development team is finding and solving the problems that are most important to current and future users and other stakeholders (like CIOs, for example). In order to do that, you have to “get out of the building,” as Steve Blank puts it. You have to talk to a lot of folks. It doesn’t mean you have to do what they say, but you do have to *listen* to what they say. Ohloh is not going to tell you if the features in 2.9 or 2.10 are the ones school stakeholders are going to be passionate about. Doing that legwork is just part of being a good engineer in the era of the consumer web. There are lots of different ways to do it, most of which do not require changes in governance. But one way or another, it has to be done if CLE is going to remain a viable project. By the time you start seeing schools leaving the platform in numbers, it will already be too late. You have to anticipate the needs of the people you are building the software *for.* That requires…you know…talking to a lot of them in a fairly thoughtful and systematic way. That’s what I take Steve to be suggesting.

  7. If the community would at least be able to set a direction (eg UI style guide, preferred front-end technology) that we can all agree on then all those institutions scratching their own itches would together scratch the same itch in a more effective way. Our resources are working on our local needs and I think those needs can be aligned much better with the community. Currently there doesn’t seem much to align with 😦

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s