Wednesday, January 09, 2008
Following on the heels of IHE, there's a story in the Chronicle about several universities (Northwestern and Arizona State among them) giving up their own internal email systems in favor of Google's gmail. I'll admit, it strikes me as one of the best ideas I've heard in a long time.
If you haven't tried it, gmail is free, almost always up, and backed up with storage capacity you wouldn't believe. The actual physical servers are heaven-knows-where, which means they're separate from most campuses. (In the event of natural disaster, that's no small thing.) And it's free. Did I mention that?
Anybody who has ever tried to contact students via their campus email accounts knows that you're frequently better off relying on a carrier pigeon. Students rarely use their campus email, rendering the system as useless as it is expensive. They'd be likelier to use something they can take with them wherever they go, and that offers enough storage to use indefinitely.
I think I personally drove our previous IT guru to retirement with my constant nagging about 'open source' that and 'free' that. (See this post from 2005 as an example.) His responses started off generous-but-condescending -- “that's an interesting idea, but as you know, we don't have the staff to support it” -- and eventually became downright testy. But it struck me as a good idea then, and it strikes me as even more so now. In a time when we're shrinking the cadre of full-time faculty to save money, why the hell are we buying servers and paying staff for our own internal email system? Why not use gmail (or something similar) and use the savings to, I don't know, hire faculty?
Going farther, why the hell are we sending boatloads of cash to Microsoft for a gazillion Office licenses when AbiWord and OpenOffice are out there for free? (Google Docs shows promise, too.) For that matter, why not try Linux instead of Windows? Let Bill Gates absorb the hit, rather than my English department. He's better able to take it. And the time we save with fewer system crashes wouldn't be trivial.
And have you tried Blackboard/WebCT recently? Sheesh. I mean, Sakai and Moodle are just sitting there...
The only semi-persuasive argument I've heard for continuing to feed the Windows pig is that it's the “industry standard.” That's true, but circular. It's true until it abruptly isn't.
Usually, pushing for new technology involves spending more money. This is that rare case in which pushing for new technology would actually save money.
I'm not usually a big fan of 'outsourcing,' but I'd much rather outsource email and embrace open source – and apply the savings to just about anything else – than continue to outsource our teaching to adjuncts while paying ever-higher licensing fees to software monopolies.
Has your campus tried any of the open-source or free stuff out there? Has it worked? Is there a relevant downside I'm not seeing?
Oh, and I'm actually not the only faculty member hacking on AbiWord. There's also a physicist from Australia.
More to your point, I'm actually a bit ambivalent about using gmail itself, due to the limited namespace (my own gmail address is my third choice), but certainly something like "Google Apps for Your Domain" sounds like it ought to be a viable option, and I'd definitely support using open source software for other campus necessities. I sure don't run any Microsoft software on my computers (but again, atypical case).
Just one more point: people spend a lot of time and effort fighting Microsoft products, but they just kind of expect it to come with the territory. Open source software has different battles to fight; I'm not sure they're harder to fight, and the cost is lower.
I can think of one, with regards to Gmail: confidentiality.
As you know, Google stores the contents of your email for you. They also gain revenue by putting keyword-driven ads over on the right-hand panel. That means that there is no such thing as a confidential Gmail message.
In some cases, that's no big deal. When I teach, I send class communications to my students' campus accounts plus any other address they ask me to include. Class schedule changes aren't a high-security item, so who cares?
However, an email about a sensitive personnel issue or a possible lawsuit is another kettle of fish. If it goes through campus mail, at least the email is held on servers and managed by employees that are under the control of the institution. On Gmail, not so much.
I can think of several other benefits and potential costs for open source vs. proprietary---I am partially in the business, after all--but I've got a paper to work on at the moment. May post more later.
The reason we use Microsoft products is because if we didn't, a trivial number of our students (at the most) would be employable. We could produce the most open-source tech-savvy graduates in the state, but if they knew everything but MS products they wouldn't be getting the jobs that list "Microsoft Office experience required".
In my previous job we did actually bypass a few consecutive versions of Office, justifying that decision on the combination of no useful new features and increased infrastructure costs required to run the newer versions. That experiment was ended when the company owners realized we'd overlooked one critical feature - the ability to open documents in the only format clients were willing to use.
As a CC, our mission is to train and educate students (we still offer a number of technical programs) so they are better-equipped to succeed in the marketplace. So long as the employers want Windows and Office certifications and training, we need 1000+ licenses for the student machines so we can provide that training. After that, the cost of buying and supporting MS products in the offices is trivial, and the cost of running a completely seperate set of OS and applications is decidedly non-trivial.
Good, bad, or other - they're among the skills employers consider dealbreakers during the hiring process, so they're the skills we have to teach and thus the software we have to buy.
"As you know, Google stores the contents of your email for you. They also gain revenue by putting keyword-driven ads over on the right-hand panel. That means that there is no such thing as a confidential Gmail message."
Actually, the inclusion of automated advertisements does not, of necessity, indicate a violation of confidentiality. It is an automated, and algorithm driven process.
Be worried that they are storing all those emails. Be concerned that, when you log on to a browser with your google/gmail name they are then recording all your searches and web site visits. Just don't worry that the computer that matched an advert to your email is somehow "learning" something about you.
On the other hand, we run many of our systems through PeopleSoft...which is no bargain, financially or operationally.
There's no perfect way. unfortunately.
While that process wouldn't associate an individual user with a particular chunk of content, it can provide aggregate information. That would be especially dangerous if a college had negotiated some sort of way of distinguishing their email from other people using Gmail (email@example.com). I agree that it's less of a risk than the offsite storage, but it's still a risk.
A lowly history prf
I've also been a Banner admin. That biased me in a different way. I loved moodle. A lot. Banner (and PowerCampus)? Not so much.
I would much prefer to open-source the back end than the front. Let the students have their Office. But run the website on a LAMP stack, run the courses on Moodle (and not sakai, IMO). I'm watching the Kuali project because I think it offers the possibility to similarly move folks away from expensive server software. Ten thousand dollars buys many copies of Office and Windows. It buys far fewer infrastructure programs (often less than one). And since those are licensed, you get to pay again and again.
Email is a weird one. If it goes down, peoples' heads start to explode. I don't know why. So letting those go, even though they're the biggest pain ever, is tough for anybody (school or business). But it seems like if you're going to trust someone, google is as good a bet as anyone. Better the devil you know....
Bad. Plus, do you really want to have to send emails to firstname.lastname@example.org when you need to get in touch with a student?
Google email for education can be negotiated with no ads, and with .edu addresses.
Who honestly thinks their campus email system is backed by strong enough support to a) keep from getting hacked and b) withstand legal demands from the U.S. Government or courts. I think I'd rather have Google and it's money & lawyers protecting my privacy than the legal support that my campus could afford.
Also, amen to the "let's spend the saved money on hiring faculty."
The argument that colleges can't withstand demands of the government is sort of self-fulfilling. Colleges are well suited to stand up for student rights, educational rights and freedoms... at least better suited than either corporations like Google (whatever their slogans) or individual students or employees.
We have a whole lot of government and private interest coming together to monitor what's happening on college campuses for what is arguably private interests (that of the big copyright companies) as well as politicized ones (like encouraging surveillance and reporting of suspected undocumented immigrants). I don't think it should be a given that universities are going to roll over on all these issues, I think it should at least be up for debate.
these are cases where the laws are shifting, they are not constant and it's not a foregone conclusion. So giving up on resisting them just allows a precedent to be set against you.
Some colleges (and not always the most wealthy ones) have pushed back on these issues already. Before trusting Google to stand up for students privacy and free speech rights, it would be worth weighing the value of colleges taking that on (with attendant costs of keeping one's own email system, etc).
That's a problem with the particular email system, not college emails in general. Both of the college email addresses I've had recently have allowed me to log on from anywhere just as easily as Gmail.
Once I'm logged in, I'm not sure which I actually prefer. There are definitely little features (like the ability to open more than one message at a time) that Gmail doesn't have that my college email does.
It's an open question whether there is more pain in administering MS products or open-source ones, but effectively there are a lot more MS-trained technicians than open-sourcerers. Add to that the natural inertia of any functioning system, and it becomes difficult to make the change.
gmail - the big hairy question as already noted, is ownership of the data. Once using gmail, effective control has passed to Google.
At ASU faculty/staff accounts are still Microsoft exchange.
Student gmail accounts are still the old @asu.edu address. Most students use the account they are given, i.e. Jane.Doe@asu.edu, but it is possible to set up aliases.
Email at a public university is public record. It is absolutely a FERPA violation to put grade information in an email. Don't do it - regardless of who is hosting the email. My entire department's email was subpoenaed when we were on the old ASU run EMMA system so I don't really see that having google vs. a university system makes a difference. (Although you might be able to make a case that the google mail holds a higher expectation of privacy - but I doubt the university would fight these, most of the email subpoena's are for news stories not law suits, the university tends to just hand over the thousands of emails and say ok, find your needle).
- Why not migrate your email to an outsourced provider such as Google For Your Domain or Windows Live? (Let's ignore the Cloud Computing issues for now.)
- Why aren't people using more Free/Open Source Software for running campus IT services?
So to the first: I think the privacy issues have been well addressed here. I will add that Google's motto is "Do No Evil" but they also don't promise to expunge your data after you "delete" it. It's up to your lawyers to determine if that's acceptable risk, or not. Beyond risk, there's the issue of damage control: Does your upper administration have the intestinal fortitude to deal with your campus community during a large scale loss of service event from an outsourced provider? What happens when your Gmail isn't available for a day and you have no control over the situation? There are also level of service issues -- you'd be surprised how many people open Help Desk tickets demanding to know just what exactly happened to a random email someone sent them a month ago that they now can not find. Google doesn't offer that level service. For some that's a problem.
Outsourcing something as business critical as email is a very difficult decision, shouldn't be taken lightly, and should be well evaluated from all angles. Yes, there are a lot of advantages (cost savings being chief among them) but they are equaled by a similar list of risks. Risks are scary. Most bureaucrats don't get paid enough to deal with scary things. The status quo persists.
For the second question...
In I.T. years ago there was a wonderful saying "Nobody ever got fired for buying IBM." fast forward 20 years and it's "Nobody ever got fired for buying Microsoft." So, above and beyond the "everyone else uses Microsoft, so we should too" and the "incredible inertia" claims there are issues of business processes, in-house knowledge and skills, initial cost of ownership, and return on investment.
For a lot of places, the cost savings of Moodle, Sakai, Cyrus Mail, or OpenLDAP (versus, say, Blackboard, WebCT, Exchange, and eDirectory) are irrelevant when calculated against the cost of retooling your workforce, rebuilding your infrastructure, and reimplementing your business processes in order to make those things work well. Yes, those are "one time" costs but they're huge and get spread over a 3-5 year (or more!) time frame. Initiatives like that also require a huge amount of support from upper administration advocates to make them work. That's even before the campus politics come into play.
Further, the devil is in the dollars. Proprietary software tends to cost more, yes, but scope and features are limited, as is the ability to "tinker" under the hood. What you see is what you get, so make it work to the best of your abilities. With Free and Open Source Software it's quite the opposite -- tinkering under the hood is a requirement. In the end you can and probably will end up paying equal amounts of money for alternate solutions. The main difference being that the allocation of funds goes from capital investments on software and hardware to human resources to employ people to understand and run the applications you want.
My personal belief is that the latter method of having High Quality human resources who are able to address and resolve issues in Free and Open Source Applications at your disposal leads to a higher level of service with a better ability to adapt to the organizations' business needs. But lots of people don't like that line of thinking because no one wants to be in the position where the only one they can blame is themselves. It's also nearly impossible sell upwards for that reason.
In either case, I don't think you'll see a widespread change in services from one method to another without some for of "tipping point" event (such as a mass failure of a critical system caused directly by the product and an inability for the vendor to fix it) or the application of a distinct vision from an enlightened member of upper administration who has the political clout to make it happen. There needs to be some catalyst to turn the fear of the unknown into a desire to innovate at a organizational level.
The system has proven to be exceptionally stable and scalable. Spending large sums of money on these systems is totally unnecessary.