The Future of Open Source in Higher Education

Thursday, September 27, at 12 noon, U.S. Eastern time

Brad Wheeler, chief information officer of the Indiana University system, is an active participant in several open-source software projects. He is also a professor of information systems at the Bloomington campus's Kelley School of Business, focusing on the use of computer networks to improve business processes.

A transcript of the chat follows.

Jeff Young (Moderator):
    Hello and welcome to our weekly Brown Bag discussion.

Today we're talking about a topic that continues to be a big one in higher education -- the future of open source (or community source, as I think Brad prefers to call it).

Our guest today is Brad Wheeler, chief information officer of Indiana University. He's been involved with open source projects for many years, and knows first-hand the promises, challenges, and pain of using such software in a college environment. Thanks for being here with us today Brad.

Brad Wheeler:
    I appreciate the Chronicle providing a forum for this timely topic. I've been involved in open source efforts for many projects, and I'll be happy to comment on these experiences and the road ahead.

Jeff Young (Moderator):
    Many of you probably know our format already, but here's the plan: for the next hour or so Brad will be taking your questions. We have a few good ones already in the hopper here (thanks for those) but please jump in if you haven't already... or send in a follow-up. Should be a good session. So without further ado, we'll get started with our first question.

Question from Herman A. Serrano, CMC:
    Considering all the pains facing the HE market what do you consider is the largest pain or concern when using open source or moving their proprietary system to open source?

Brad Wheeler:
    The largest pain is probably taking responsibility. It's much more familiar to be able to blame a vendor. When WE -- all of higher ed -- collectively own and improve our software over time, we are faced with the reality that we chose to deploy resources to fix/improve the software. We may still work with and hire commercial firms to implement, host, or even write tools, but ownership brings responsibility. It is a mindset change. In Scott Rosen's book, "Dreaming in Code," he makes very clear how difficult it is to build enterprise scale systems. It is very difficult for both companies and communities. There is no magic fairy dust that makes these challenges go away, and there are substantial risks in ANY path, but there are also many upsides to implementing and working with code owned by the higher ed community.

Question from Len Brin, Southern CT State University:
    It seems that open source solutions to needs such as course management or payroll management require more in-house labor where commercial solutions require more tech support or paid consulting. First, is that an accurate assessment? Second, if so, how do the bottom lines in terms of dollars spent compare?

Brad Wheeler:
    Yes, I think think your point is right. I call this the "n + 5" scenario. Take your TOTAL headcount supporting a legacy system (homegrown or otherwise). Switch to a vended system, then check total headcount again. The job titles will have changed (staff or rented consultants), but the overall number is likely larger. I think that you are right in pointing to differing blends of jobs required for different paths. Regarding the bottom line, you have to realistically assess near-term and longer-term costs. If a system fits better, it will cost less to implement. Corporate accounting systems are NOT a good fit for fund-based universities. I know some of the Kuali Financial potential adopters are viewing implementation costs as neutral or slightly favorable to KFS. They view longer term costs as very favorable in control of destiny and leverage of the higher ed community.

Question from Wytze Koopal, University of Twente (The Netherlands):
    I have a sense that opensource (or 'community source') is a model that really is getting big in the US. There is much more attention and appreciation for this model in the US, than in Europe, or The Netherlands. Could Brad comment on this: what could be causing this difference? And if so: is a community largely consisting of US institutions (like the Sakai community), a sustainable model?

Brad Wheeler:
    I've had many discussions on this and that side of the Atlantic regarding funding models. In the US, we look with some envy at the ability of many European nations to centrally coordinate investments for leverage. My friends there look with envy at the chaotic funding "market" of many disparate sources in US higher ed. I think it is the latter that gives motion to community source. In the US we are in NEED of a model to coordinate our funds, but we still want to retain that fierce individual control that is our heritage. Community Source provides a way to do so. Yes, I am very convinced of sustainability.

Question from Jim Farmer, Georgetown University:
    Do you expect consolidation of open source projects? Do you expect open source projects to be "purchased" by current software suppliers?

Brad Wheeler:
    The key word here is "projects." It takes a lot of community coordination to make these projects work. There are natural economies of scale/scope in some of the work for Quality Assurance Testing, Release Packaging, Member conferences, Code Distribution, Load Testing, etc. I think a lot of the "back office" functions can be consolidated for community good. I think a lot of projects can continue to function as a "front office" community and drive the code to meet there needs. Regarding the "purchase" question, no. What would they purchase? Kuali, Sakai, Open Source Portfolio are already under the open-open Educational Community License, so I don't think so.

Question from Shahron Williams van Rooij, George Mason University:
    Based on your work with Kuali and Sakai, what type of skill set is required of an institution's IT staff in order for the deployment and maintenance of enterprise-wide open source software systems to be successful?

Brad Wheeler:
    The collaborative "gene" goes a long way. Staff need to learn how to work effectively in a distributed community to influence things they need and how to ask for help. They have to learn how to contribute. They have to do all of this while DELIVERING for their local users. This is a mindset shift for some staff and institutions.

Question from Charlene, Desire2Learn:
    Do you see a particular type of institution or size of institution better able to go with Open Source than others and if so why?

Brad Wheeler:
    Not at all. Many very small schools are using Sakai with some commercial support at a great price point. They are also contributing directly to the community to improve the features and code...even if they don't have developers. The first Kuali Financial System implementation was at a small school.

Question from Linda Cantara Abbott, U of KY:
    One problem I've encountered with open source software is that at the same time local add-ons or adaptations are being designed and implemented, the open source community is continuously updating the original software, frequently leading to inconsistencies in the local adaptations (I'm thinking primarily of institutions that have developed their own front-end interfaces to Fedora). How can the HE community implement open source solutions without constantly reinventing the same wheels?

Brad Wheeler:
    Transparency is the key. Don't labor in private in a cave, but make the community aware of your work via the project's Wiki or listserv. That is the best way to keep local work coordinated. The other option is to skip releases of the core project code so that you have more time to adjust local enhancements.

Jeff Young (Moderator):
    Keep the great questions coming in ... we also invite readers to share their views or experiences about open source (briefly) by posting a comment.

Question from Len Brin, Southern CT State University:
    Besides course-management and financial systems, open-source desktop software such as the GNU/Linux operating system, OpenOffice.org, and the firefox browser is becoming increasingly sophisticated and stable. Do you see a place in higher education for these or other desktop applications?

Brad Wheeler:
    Yes, there is a place for higher ed, but these are not ours to lead. The digital humanities center at George Mason U. is doing some outstanding work on a Firefox add in that serves higher ed. I think that is the right role for us on these general purpose projects.

Question from Robert DeFranco - American College of Applied Science:
    Could you envision a graduate degree program in educational technology with a specialization in Moodle VLE and CMS?

Brad Wheeler:
    Absolutely. Each of these projects has a very large community and growing need. Skills that spanned educational technologies and effective learning would be great. Let me know when I can hire your first graduate.

Question from Robert DeFranco - American College of Applied Science:
    We tested Sakai for a short time and found it be clunky, unstable on a Linux platform and difficult to get any support for it without being a major supporting partner. We now use Moodle and love it so far. Is the Sakai project with all the money spent on it doomed?

Brad Wheeler:
    Its a very big world with lots of room for lots of options. Some people love Blackboard...some find it unusable, etc, etc. Sakai is running in production in many places as is Moodle, so I wouldn't overgeneralize from your experience. They do some of the same things, but their communities also have some varied objectives too. That's what's great about letting folks find the software and community that best fits their needs.

Question from Moodlerooms:
    How do you see the marriage of commercial companies and open communities evolving with the likes of joint R&D?

Brad Wheeler:
    Gartner recently reported from its Open Source Summit that Open Source would be very disruptive to software business models. Good capitalists (including me) always adapt to new opportunities. Red Has has thrived supporting Linux. Sugar is doing well with their CRM. There are new approaches to Blended models between the Cathedral and the Bazaar. I see joint R&D working out well for companies that get right-headed about the process. IBM and rSmart and many others are already demonstrating that.

Question from Jim Farmer, Georgetown University:
    What are the important lessons from the Kuali Project that every Chief Business Officer should know?

Brad Wheeler:
    Read them in the July Business Officer Magazine where I wrote about "Mitigating the Risks of Big Systems." The key point is there is no reality in the false game of "let's hide the IT risk." There are just different risk mitigation strategies and each institution needs to choose its path well.

Question from Ron Turner, Univ of Virginia:
    You mentioned that some commercial systems are not a good fit for the HE community. Can you compare cost and time to implementation of open source with commercial systems that seem to have appropriate functionality?

Brad Wheeler:
    I think commercial systems that had their genesis of design in higher ed are likely to fit fairly well versus systems that originated for other industries and are then repackaged to fit us. Again, the early indicators that I've seen are fairly comparable implementation costs, so I don't think that is where the real value of the choice is realized.

Question from Linda Cantara Abbott, U of KY:
    As a former IUB librarian, I know this is not a problem at IU, but for other institutions, how do you suggest we bring the administration on board to hire programmers to implement open source software solutions rather than pay large sums to commercial vendors?

Brad Wheeler:
    It is a great question. We marvel sometimes on the Kuali Board how institutions can easily write 6, 7, and 8 figure checks to vendors but community membership at $10K or 25K seems a big deal. Likewise, local developers can be a great bargain if they are a path to leverage the work of others. I see a shift in the higher admin ranks. The estimated $5B that higher ed spent on ERP is still of debatable value. The conversation is again open for assessment.

Question from Jeff Young:
    Are we entering a new period in the history of open source in higher ed when there's a critical mass of usage and acceptance? Or is it still early days? How would you characterized the current environment?

Brad Wheeler:
    Yes. The conversation has definitely changed. Open source application software for higher education has moved from "are they crazy" to a legitimate option for evaluation. We are seeing that across all projects -- Research Administration, Portfolio, LMS, Workflow, Financial...the conversation has shifted. The tone of the last Kuali Days was very different. There was much less information gathering and more serious evaluation.

Question from Charlene, Desire2Learn:
    As part of one of those "commercial vendors" do you see Open Source really taking off in the US and taking the lead in the market...in other words, do you see Open Source tumbling vendors like Blackboard who have a huge share of the market currently?

Brad Wheeler:
    I see the market going wherever higher ed wants it to. I've written elsewhere, "We don't have a money problem for IT in higher ed, we have a coordination problem." If vendors provide great value to collect our money, coordinate work, and sell us back a product at a good value, the commercial coordination is working well. When that breaks down...and it often does as higher ed markets get monopolized over time...then we need to look at coordination ourselves. I see a healthy collaboration between commercial and open source, but commercial controlling everything will someday likely run its course.

Question from Sherry, UGA:
    What do you think surprises people most when you talk with them about community source?

Brad Wheeler:
    That higher *CAN* actually collaborate. As an industry, we've tried and failed on various endeavors over the decades, so why would this one be any different. It is different because everyone has walk-away rights to the code. No one is holding them hostage to a particular path, yet everyone is incented to make collective efforts work for their own good. That balance is the key, and it changes perspective when folks really grasp it.

Brad Wheeler:
    These are excellent questions. So where do things go in 2-5 years? What are your thoughts for these communities and open source applications in higher ed?

Question from Jeff Young:
    What is the biggest challenge to open source at colleges at this point?

Brad Wheeler:
    The biggest challenge is not dabbling. Like many things, setting a strategy, executing it well, and extracting value from it is effective. Dabbling in the new (with attendant costs) and still paying for the old just keeps things in a state of flux with unfocused investments and value.

Question from Shahron Williams van Rooij, George Mason University:
    How would you define "healthy collaboration" between higher ed and commercial vendors?

Brad Wheeler:
    This requires changes of both the vendors and us. We are expensive to sell to, we are slow and long in making any decision, we wants lots of hand holding, etc....and all of this drives up the cost of serving us as a market. It is factored into the price. We need to learn to be better buyers of commercial services. Firms need to learn to sell us products and services of value without excessive attempts to lock us in. The switching costs for most apps in higher ed is very large, and it is cultivated by the commercial sector to make it hard to switch. That is not in our interest. Open source Intellectual Property neutralizes that and resets a competitive marketplace for buying services around common IP.

Question from Moodlerooms:
    What could community source and open source project members do to better encourage companies to invest resources in their efforts?

Brad Wheeler:
    I think companies are smart. They go where they believe they can make a profit and that is a very good thing. For many firms, they need to grow new capabilities of working in open source communities and wind down some expensive costs they have that are no longer needed with the community. I think open source participants need to help firms see how they can do this and still get profitably paid for their work. Some firms are leading on this transition now.

Comment from Robert DeFranco - American College of Applied Science:
    In response to your question, Open Source could displace commercial products in 5 years unless the commercial companies start working smarter. When our small college approached them they had no product with cost within our reach and they simply stopped returning our phone calls. So we did it ourselves, first with JESS (Jones Int'l giveaway software, then Sakai and now Moodle. The commercial market is controlled too tightly. Academia will continue to look for alternatives, and we see this happening now. I smell dead fish...

Question from Andrea Foster, Chronicle of Higher Education.:
    Are college presidents supportive enough of open-source software?

Brad Wheeler:
    No. Few know or understand, but some do. It is fair to say that mine does, and Jeff Lehman when he was Pres of Cornell saw the value. I think educating CFOs and other senior officers is likely the best investment in the near term. NACUBO, AACRAO, and others are working on this through a variety of conversations.

Question from Mike Zackrison, rSmart:
    There was a question about cost savings of open source. Here you can read about the savings Strathmore University realized by adopting Kuali Financials utilizing the help of a commercial partner instead of the alternatives they were considering (note the figures are quoted in Kenyan Shillings): http://www.bdafrica.com/index.php?option=com_content&task=view&id=1916&Itemid=5810

Brad Wheeler:
    Absolutely. I think the Kuali ERP applications have enormous potential to help higher ed save money and get the systems they really want. [Whisper...watch for a Kuali Student announcement later this fall, shhhhh]

Question from Josh Fischman of The Chronicle of Higher Education:
    You say the way to avoid "reinventing the wheel" in an open source project is to work transparently and alert the wider community to what you are doing. What are the best forums for doing that? Any in particular that you like?

Brad Wheeler:
    If it is work related to a particular project, DSpace, Moodle, Fedora, Sakai, etc., these projects have rich community spaces to signal your work. In Sakai, there are levels for a project that indicate it is simply a discussion, it is an untested "Contribution" or it is "Provisional" and in use by the community. If you have a great Dorm-Room-Assignment-Widget, most of those get discovered through informal conversations. I see lots of these valuable widgets ultimately finding a home with one of the larger communities.

Question from Moodlerooms:
    Institutions do not disclose their operating costs for either commercial or open source systems making it very difficult for other schools to make informed decisions. Do you see this changing?

Brad Wheeler:
    I hope so. I'm a pretty big believer in transparency. There is a LOT of information sharing with communities regarding their costs, but not much sharing with folks who may have made or are contemplating different decisions. Also, there is much variance in the abilities of sites to do implementation, thus a $500K expense at one institution might honestly only require $150K elsewhere. The key point is what is your strategy and why?

Question from Charlene, Desire2Llearn:
    Please excuse me if this question has an obvious answer...I just don't know it....is there a place/organization where commercial vendors and open source can collaborate and work together so that we can co-exist and higher education walks away with what it truly needs at a price it can afford?

Brad Wheeler:
    Yes. Sakai and Kuali both have Commercial Affiliate programs where commercial firms are voting members of the respective foundations. They often participate in the community with contributions and even code. Moodle has a well developed commercial engagement path. Dspace has engagement with HP and others, etc. Thus, I think this is a VERY important question and path. The key is that the desire and efforts must be real. Co-branding for marketing is not helpful, but real work is. Send me a note. John and I talk often.

Comment from Moodlerooms:
    Comment - We see the biggest element that is missing from the educational open source projects is the focus on operational characteristics such as security, manageability and performance. These things really dictate the tru cost of ownership.

Jeff Young (Moderator):
    Looks like we're out of time for today. Thanks to everyone for all the questions and comments.

And thanks to Brad Wheeler for all the views and information.

Brad Wheeler:
    Thank you Jeff, and thanks to the participants. For those who want to know more, I'd encourage to to attend a MoodleMoot, The Sakai Conference, Kuali Days, and other gatherings of our communities. There is nothing quite like rubbing shoulders with the people who are making all of this possible.