How “open-source” is Sakai?

Yesterday I had the privilege of heading down to the University of Virginia and catching up with Steve Stedman and crew who are strongly considering piloting a WordPress Multi-user installation for their ITC group. We talked a bit over a Mexican lunch about Content Management Systems in general (Patrick turned them on to Drupal, and I even got to talk Typo3 with the other Steve). It was extremely enjoyable and one of the topics that came up is UVa’s impending campus-wide move to Sakai, which I believe will be happening slowly over the next couple of years (the switch will be complete sometime in 2009). They have a pilot in the works called Collab, and in the afternoon show-and-tell session Patrick and I got to take a look at the inner-workings of Sakai for the first time.

It was pretty eye-opening in many regards. First and foremost, I have to be honest and say that I was pretty underwhelmed with Sakai. Steve and company did an excellent job skinning their install and, in fact, it immediately looks like a very inviting environment. But once inside, it seemed more like a labyrinth than a manageable tool for authoring and sharing resources quickly and easily online. I was surprised to find that Sakai, much like many of the proprietary course-management systems, fell prey to building their own applications such as the wiki, the blog, chat, forums, etc. None of them seem particularly stellar, and in fact many of them are admittedly sub-standard–specifically the blog. With all the thought, time, and money put into Sakai, you would think they would have realized that pulling the best of breed applications into this course management tool would be the way to go.

In fact, some of the UVa folks were excited about the feature that embeds websites using iframes into one’s own Sakai space (I believe they referred to it as links). While this may have some value, I think the integration of these sites is a bit tenuous at best. All that is really happening is that you are embedding another web page into Sakai -so you can’t really “run” Moodle or WordPress in Sakai, you can simply allow users to link them into the workspace. Making Sakai work as a kind of “centralized browser” of sorts. The problem with this is accessibility and the user interface. As we went through the demo of Sakai, the problems of navigating through the frames became readily apparent. Steve pointed out the issues here with the frames dis-orientating the user and how the system itself often has a series of “false homes” that can trap the user in a loop of sorts.

More than anything else, however, I was extremely disappointed with the limited RSS capabilities. You would think that an open CMS would have the RSS flowing like wine, and folks could have the option to hook in to one another’s content making for a community of rich syndication much like the feed-based architecture that Jon Udell discussed recently here. However, nothing doing from what I have seen. The one RSS feed I was able to see was for the wiki, and it seemed to have problems distinguishing between particular project pages. Making it effectively useless.

On top of this, the URI (or URL) for a particular user’s space is a long chain of undecipherable characters, making navigating to a site or sharing a link cumbersome–effectively hiding even open pages within a course. So radically different from the beauty of dynamic sub domains in WPMu, which in many ways may make all the difference for ease of navigation and the true feeling of each user having their own space on the web.

Now, truth be told, all these issues should be fixed, if not immediately, then at least eventually because the source code is open and folks can share extensions, work-arounds, etc. However, it may be open-source in theory and name, but given that it is primarily written in Java, I’m not sure how quickly Sakai will see any of these improvements. Perhaps for big, rich institutions like UVA putting Java developers on the scene can allow them to incorporate these improvements for the general community. Put even big, wealthy schools have definitive budgets, and a crack Java developer is both expensive and increasingly rare, making the open source ideology ring somewhat hollow. For at its root, open source, at least for me, has become so deeply associated with the ease of contributing to a code base.

Steve brought up several excellent points during the conversation, but I will focus on one in particular. Steve suggested that one of the possible dangers of a system like Sakai is that it is quickly becoming the flag-ship open source course-management system for many big research institutions, but with the community of contributors defined so narrowly to Java developers, at what point might open-source solutions suffer as an alternative for high-level administrators throughout education if Sakai doesn’t live up to its promise? Which in my mind is a definite possibility. A host of research universities threw millions of dollars at an idea of an open source application, but they did it in a traditional academic way which put the idea and the perceived needs before the actual community and real needs. Now, I firmly believe that the theoretical and abstract conceptualizations of academia is the manna of so many great things, but I am not so certain that this is the case for open-source course-management systems. In fact, I think projects like Drupal, WordPress, MediaWiki, etc., capture a vibrant community because the details are being imagined and re-worked as the applications evolve. The communal evolution doesn’t precede these systems existence, it defines it -a key difference from the organizing logic of a system like Sakai as I see it.

But, I am sure I’m overlooking, simplifying, or mis-representing several important elements of Sakai. I would love to hear about all the functionality that I have not seen yet, but I’m still a bit scared there is no there there.

This entry was posted in Uncategorized and tagged , , , . Bookmark the permalink.

23 Responses to How “open-source” is Sakai?

  1. Brian says:

    Those are disappointing reports you are making — especially the stuff about limited RSS support. (BTW, I hadn’t read that Udell bit on syndication architecture — thanks!) I’ve been a bit worried that SAKAI was going to represent some effort to build an open source version of an Enterprise BlackWeb product, and while your summary makes it sound better than that, it also sounds like it might replicate some of the same defects.

    I’m reminded of something George Siemens said at a symposium on distributed tool strategies: that schools should be in the business of managing data flows rather than in supporting an end to end user experience. We can only dream what might result if the energy going into the campus-wide LMS’s would go into creating flexible and easy to use “syndication buses” or to addressing pragmatic instructor challenges to using the “small pieces” approach — things like student management tools, gradebooks etc. And what about providing the service of institutional archiving and data backups to mitigate the risks of using third party tools?

  2. Sakai is absolutely Open Source. It may not be as accessible to developers as a PHP system, but even requiring Java ninjas is far better than requiring a proprietary commercial LMS.

    I’ve installed Sakai and Moodle. Moodle takes maybe 10 minutes to install, and works out of the box as a pretty full LMS. Sakai (when I installed it, admittedly several months ago) took much longer to install, and didn’t “do” anything out of the box. But it’s not intended to be a drop-in LMS, it’s intended to be a platform for development.

    The openness of the community is a separate issue from the openness of the code. Yes the code for Sakai is Open and available, but you need to be a coder with a strong background in a few areas before you can really contribute to the code, so the community of developers is smaller than with Moodle. But, an Institution might get more mileage out of Sakai than with Moodle, depending on the skillset and requirements of the Institution.

    Sakai may not be my cup of tea (hell – I don’t even DRINK tea), but I’d much rather my university used it than Blackboard. I’d love to see them use Moodle, too. And any of a number of great open source solutions that are available now.

  3. almost forgot – the crappy URLs of resources in an LMS is pretty much rampant. Even Moodle generates craptastic parameterized URLS.

    however, an LMS is at one level an agent of DRM, controlling access to resources according to an individual’s status (in a course, within a lesson, etc…) so having readable and world-addressable URLs isn’t a priority. Having a clean URL that returns a “you must be logged in to view that page” isn’t much better than having a crappy internally consistent URL…

  4. jimgroom says:


    “Syndication buses” -I love it, what I a beautiful way to think about how to manage the flow of information on a given campus and beyond. More than that, the notion that campuses should be concerned more with archiving the work being done in a variety of heterogenous spaces, rather than one monolithic one is exactly where the enrgy should be directed. Precise, articulate, and on the money as usual.

    I do know that Sakai is open source, and my question kind of plays with that notion a bit. And I am well aware that you have far more experience with both of these open source systems than I do, and I will defer to you in general on just about everything, as usual 🙂 However, I will needle you on one point, is the fact that these two applications (Moodle and Sakai) are open source change the fact that they generally reproduce the same errors and lack of imagination and creativity as BlackBoard, especially when the development communities of at least one will be so particular that they will tend to fuel the very specific needs of one university rather than many? In fact, Sakai has so far to go in order just to get up to speed on many of the essentials that I doubt they will be sharing significant features beyond the absolutely necessary anytime soon. And I wonder just how vibrant that community is. I dont think the fact that an application is open source makes in inherently valuable or better -where is the open source YouTube or Flickr or Google? And defending open source because certain applications have thrived does not convince me that all these open source tools are conceptualized and or developed equally.

    If I had a gun to my head, I would, quite frankly, stick with BlackBoard over either one of these for the basics, but now that they are phasing out BlackBoard basic, I think it is time to put an end to CMSs all together -they are a waste of time, energy and other valuable human resources.

  5. Open Source isn’t about features, or even about community. It’s about having a say in your own software destiny. Find a problem? You can fix it with open source. Want to extend it? sure thing. With commercial applications, you’re largely at the mercy of the company that builds it. And they may decide to differentially monetize their software right out from under you as well.

    I agree that the LMS model is a bit borked – but it’s not so useless that it doesn’t have a place. Some of the stuff we’re doing with Moodle would be difficult or just plain clumsy to do in blogs/wikis/etc… but works really well within the Moodle LMS.

    What I REALLY want to play with, but haven’t had time (yet), is the Moodle Drupal Mediawiki integration. Now there is a sick combo. And throw in stuff like BlogBridge FeedLibrary, WordPress µ, etc… for a fully flexible, extensible, distributed (and yet managed) set of resources. Sweet. And all of these Small/Midsized Pieces are open source (even if the BBFL license is a bit unorthodox).

  6. Pingback: OLDaily[中文版] » Blog Archive » 2007å¹´9月13æ—¥

  7. Colleen says:

    Although I agree with Darcy that open source is about having a say in your own destiny, I think the technical community has to remember that their destiny is intertwined with that of the customers.

    At my institution, most faculty hate Sakai, unless they’re using it only as a gradebook and to post their own content. Still, there’s ongoing warning that the institution is committed to Sakai as the next L/CMS and Blackboard’s days are numbered. Why? Because the CIO loves open source and the ‘control’ it provides.

    It’s buggy and blank and there’s not much there. When technical support folks talk about lack of control, they’re talking about their
    own, but they don’t use the product. Let’s turn the conversation to user control (instructor and learner) when putting together the score sheet.

    From the teaching side, I just want the buttons to work when I click them. With Sakai, they don’t – at least not consistently. Here’s the thing: we all drive different vehicles. I like a reliable Honda, but Blackboard is some kind of big, gas-guzzling beast of Motown make. Bothersome, but it still gets me to my doctor’s appt or the grocery store.

    Sakai is a Fiat. An adventure, and some love it, but…Fix it again, Tony! You might as well buy a bus pass if you want to get to your next appointment.

    I second the ‘syndication buses’ – but let’s make them small jitneys. If one lets you down, we could put a lot more zipping along the streets. Sorry, enough transportation metaphors. A lovely Friday afternoon and time for me to take a walk 😉

  8. jimgroom says:


    You have no idea how upset I was when I clicked on your name and it didn’t take me to your blog. “Fix it again, Tony!” I want more of that kind of commenting on a regular basis -right up my alley!

    It is interesting to hear how faculty are reacting to Sakai, from what I’ve seen (admittedly next to nothing) the usability could be a major issue for most faculty who are used to BlackBoard or some other relatively straight-forward system.

    But more to the point, at what point do we start with open source as a kind of base means of evaluation, then further examine what makes a good open source application. But perhaps we need to just admit that course management systems, in their current incarnation (open or closed source) seem extremely unlikely to fit the bill for so many emerging needs, desires, and possibilities. Syndication buses do make so much sense, and without an RSS driven system of content -I really think CMSs in general are failing to understand the power of the new web.

    Thanks for the comment, Colleen, I got a real kick out of it, and please come back and comment soon -or at least let me know where I can read more of your thoughts!

  9. colleen says:

    You’re very kind Jim, but in truth, I’m only funny on YOUR Blog. On my own neglected, scattered site I’m just cranky.

    But feel free to visit. It will feel much like visiting Second Life: quirky-interesting, but …eerily lonely as no one really visits there 😉

  10. Dan says:

    Indiana University moved to Sakai about 4 years ago as its university-wide CMS. Having heard great things about it, I was eager to give it a try. Prior to that we were using an in-house CMS that left a lot to be desired.

    After a brief test period, I pushed very hard not to retire that clunky in-house solution. Sakai was slow, confusing, and was a step down from other popular open-source elements out there. Not to mention, it had (has) one of the worst discussion forums I have ever used. Since then, the build at Indiana has added a lot of new, exciting features. Unfortunately, nearly all are sub-par.

    Being open-source, we imagined that it would be easy to do a little of our own development, but this is where university bureaucracies make open-source nearly meaningless. At that time, the process ran through a chain of requests, committees, and finally a masked developer somewhere in Michigan who might, if it fit in with the grand vision, look into finding an existing module or developing a new one to fit our needs.

    I’ve gone back to Sakai each year for the last 3 to test it out as an instructor. Each time I turn right back around and use Moodle on my own server. I might be one of the few who don’t want to use their own. Though it’s easy to use and provides a lot of freedom, it’s just another thing I’ve got to manage. I just want to simplify my life.

    I still hope that one day they’ll get it right. I’m hoping that one day, they’ll offer workshops on how to teach better instead of how to configure your class workspace.


  11. Pingback: Carnet de liens sur le e-Learning » Blog Archive » How “open-source” is Sakai?

  12. Pingback: Center for Teaching & Learning » Blog Archive » Sakai too academic?

  13. I’m one of the people working on the migration of CMS tools at UVa. I am not a programmer, but my research background is in Instructional Technology. Yes, we have a hard job ahead of us, especially in the area of integration of Sakai with teaching. But I find the possibilities offered by Sakai as an open source tool infinitely more interesting than anything offered by a closed-box commercial system–however smooth and dependable it may be.

    Two things I would add to this conversation: While I’m not sure about how easy it will be to integrate Sakai in teaching, Sakai has already proved to be a terrific tool for collaboration in a wide variety of ways (including research across institutions). In that capacity, my experience is that folks here have been very enthusiastic about the tool. Second, the description in the above review of Sakai’s LinkTool is not correct. LinkTool, which in my opinion is a critical feature of Sakai, is not a tool for simply embeding web content (there is another tool for that appropriately named Web Content). LinkTool is a webservices-based approach to hooking up with external applications. I dabble in ColdFusion and have found no problem integrating my tools for authentication, grade submissions, etc. with Sakai. In fact, we have also already tested using LinkTool to hook in, not only to Moodle course sites, but even to specific tools in Moodle (which, I admit, are in some cases much better than their counterparts in Sakai). We are hoping to do the same with WordPress, MediaWiki, .net tools, etc.

    How all this will turn out, I don’t know. But I know it will be exciting, and we will be hearing from many new voices.

  14. jimgroom says:


    Thanks for the clarification about LinkTool, I was hoping the iframed integration was not the extent of it. How exactly is Sakai going to integrate these external tools using this tool? It was not clear to me and I would love some more information on this. I imagine this would be a key component of joining a number of the “smaller” best of breed tools. Is the integration of these tools through LinkTool something a Java programmer would have to create? –or is it out-of-the-box? –or are these extensions/modules already available?

    I guess I have a lot more to learn about the inner-working of Sakai, and I’m glad you responded here because I’m sure my impression of Sakai was far from entirely accurate. At the same time, I wonder if Sakai promises more robust possibilities for collaboration than a handful of smaller, more agile tools that might be shaped into a larger, more heterogeneous learning network through which we can reflect and archive the mind of a university in a host of more flexible and portable formats.

  15. Jim,

    LinkTool was originally created by folks at Rutgers University who designed it to address the need to hook “legacy applications” into Sakai. It is still not part of the core tools of Sakai, but is available for those who want to turn it on. In order for your application to be hooked in, it has to be able to consume web services supplied by Sakai. We work with ColdFusion and PHP and both scripting languages have built-in web service capabilities. What we had to do is retro-fit our applications’ login pages to process the web service information from Sakai (user, role, etc.). LinkTool creates secure connections between Sakai and the application, so once the connection is made, we can trust the authenticated information about the user. Once your’ve got it to work, there is a lot more you can do to communicate other information between your application and Sakai.

    LinkTool works, but is not a finished product — I’m not sure it’s value is fully appreciated even by Sakai developers yet. But in my opinion, that’s what makes me forgive a lot of the surface weaknesses of the tool and see the possibilities. Your last comment is right on target: “I wonder if Sakai promises more robust possibilities for collaboration than a handful of smaller, more agile tools that might be shaped into a larger, more heterogeneous learning network through which we can reflect and archive the mind of a university in a host of more flexible and portable formats.”

  16. matt small says:

    This debate opens up a few of the big problems with Sakai:

    1) It is developed by research universities–and mostly by the top profs and their grad students at these universities.

    And naturally they use their own desires and dreams (in code!) for an online system as the primary use cases. But the majority of students in the US are not taught in small, highly independent groups. Maybe they should be, but that model does not scale.

    So for years now Sakai has been an application with great potential, and likely always will be, as it’s funders really have little understanding or interest in facilitating the day to day activity of a common teacher teaching hundreds to thousands of students.

    As such, Sakai largely replicates the functionality of Postnuke, Drupal, Xoops, etc. and as they do, fails when matched with Blackboard for meeting the needs of usual teachers.

    Of course the white elephant in the room has a name with two os, and must not be taken seriously as it is not java and all the profs at all the R1s _know_ that serious applications are written in java–the big grantors don’t give real money to PHP projects, period.

    And this is not about providing better teaching tools, improving student learning, or about making teachers lives easier, ok–it is about the ‘best’ faculty at the ‘best’ institutions getting funded to do work that interests them, people who are looking for a usable tool for common users at non-premier institutions to come from this process are going down the Huron with a theoretical paddle that has great potential to one day move water…

    But hey, “Together we stand, divided we fall…Goodbye Chenga” and all that.

  17. Pingback: Aeroplane Software: Sakai Consulting » Blog Archive » Sunlight is the Best Disinfectant

  18. Pingback: Don’t Call it A Blog! at bavatuesdays

  19. Pingback: University of Virginia and WPMu? Wahoo! at bavatuesdays

  20. Pingback: The Design of Openess » “I’m on a mission from Lamb”

  21. Pingback: Imagining Cyberinsfrastructure | bavatuesdays

  22. Pingback: Domain of One’s Own [visualized]]

  23. Pingback: Syndicating Myself

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.