Well guys, I'm calling it. On Tuesday, April 26, 2011, NYU's redesign of Albert was pronounced a failure. Not because it has the visual appeal of my sixth grade video game blog, or because the course search breaks in half of my browsers, or because bugs were still being worked out last night, hours before it was due to be used by the first students registering for the fall.
The new Albert is a failure because its creators did not recognize that any piece of software that takes years to develop and has no significant student testing and a built-in assumption of non-competition is unlikely to ever reach its full potential. We are left with NYU's version or nothing.
Without competition from the university's diversely talented developer community, the services provided by sites like Albert are destined to suck. Yet that competition is made largely impossible because the data underlying those services (when classes meet, class survey results, etc.) are locked away, inaccessible to those outside the hallowed datacenters of ITS.
There are a few notable exceptions that continue to deserve praise: Public Safety provides daily crime logs on its website, and CAS has a course evaluation search. Yet Stern evaluations are closed to non-Sternies, while other schools don't release evaluation data to the public at all. NYU's Office of Institutional Research, which collects rich data on university services, prevents students from accessing the majority of its information. With the new Albert, class schedules are locked away in a website that's broken on its good days and buggy on its best.
NYU should take a cue from the MTA, which began releasing troves of real-time transit data for free over a year ago. Its advertisements say it well: "Instead of developing transit apps ourselves, we gave our info to the people who do it best." By releasing this data, the MTA allowed developers to compete with one another to create more useful web and mobile applications, most of which are now available at no expense to either the MTA or to the public. The MTA saved money, and the resulting apps were better for the competition they faced.
If NYU is serious about making useful web services available to students, it needs to get serious about open data and take steps to encourage student competition.
To that end, today I am launching OpenCourseSearch.org, a website that aims to liberate one segment of NYU's data: course schedules. Through a program that "scrapes" Albert for the underlying class data (the source code to which is being made available for free), this site frees both the data itself and the means for others to obtain it down the line. Anyone is encouraged to take the data and create a better course search than I have. And I'm hoping students at other universities will take up the challenege to write a scraper for their own school, whether for course schedules or something else entirely.
It's time for developers to take the first step, and show students and the university the benefits open data brings. Then maybe they'll agree to let us have it.
