On Friday the University of Waterloo made public the cancellation of the project that was working on the replacement to a PeopleSoft based system the University uses for co-operative education job matching. There is no doubt student media will have some colourful words for the whole thing in the coming weeks although my favorite student voice (OMG UW) has been rather tame. It could be just because the real trolls are off coding somewhere during reading week.
To students, I will say this again in this post, thank the CECS staff you come into contact with for the top quality jobs and the pay scale averages they hold employers too. Jobmine might be a hassle but you wouldn’t have the opportunity to dislike it so much if they weren’t doing a few things amazingly well.
What killed it?
It was 4+ years of work, it had an excellent report and detailed research driving it, and it did have a capable team (Disclaimer: I was on it, I left it, but those that replaced the original team were certainly more than capable). Co-op students and staff that worked on it will have their opinions about specific situations or personalities but I am going to try and take a higher level view. I have thought about this for a long time but writing it down helps me think it through… so my gut feeling still is that…
Software development was placed in the leadership role of large scale organizational and cultural change.
No one decided to do it this way, not sure anyone realized that is what was happening until it was too far along, but given the report on Co-op and the large scale change that will take years (5+) to implement, the desire to do something now with a software system that is difficult to use (but works) must have seemed like the easiest thing to do quickly. The problem with that, software projects in committee laden environments are really easy to slow down and frustrate. When you toss in impending large scale change the anxiety of the change is directed at the one project that represents it.
My view is that WaterlooWorks is a victim of the process in an atmosphere of large scale change. It was a project that hit head on the cultural issues facing all of higher education and one issue the University of Waterloo is working hard to tackle (at all levels). It was too much change too early. Only in November 2010 was the organizational change in CECS announced, I am certain it was hoped the change would come earlier but I think if you consider what is happening on campus as a whole it took the time it took so it could be successul. The software project started in 2007 but staff knew then change was coming and WaterlooWorks was the first sign of that change.
In the end was the software problematic? Yes, the memo says it was. But it was a product of the process and no one can be blamed for that, we must learn from it as an institution and move on.
Student time is not real world time (4 months = 1 year)
Why the push for software? A theory of mine is that the software project was driven by student time, show changes to Co-op quickly so that the current student FEDs President can see change in the few months that they are actually leading FEDs. Understandably, as students lives change so much every four months at uwaterloo they start to expect big things from themselves and the world they are in a similar time span. That is a good and a bad thing. There are many things that can be done in short time frames but when you are moving an organization the size of uwaterloo it just isn’t realistic. It should be, it will be, but not right now.
I find myself falling into that time warp at VeloCity — a lot changes in four months — so this is something I have only recently come to truly understand or appreciate. As I enter my third year there I gain a lot of perspective though and that time warp starts to develop patterns. I am just now figuring how to take advantage of that and I think it could work with software projects as well.
Building reliable software is hard
The automatic reaction from many that can build web apps will be: “well I could have built that in a term.”
Go for it. At its core the application is just a dating site — it takes resume’s and a light profile, arranges meetings, and lets both employers and employees rank one another. Then an algorithm runs (maybe less complex than eharmony), everyone is hooked up, and students go to work for four months.
What you don’t know is all of the stuff that goes on in the background. For example, each faculty and department has requirements or exceptions that aren’t the same. That requires a rules engine. Once you dig down you find you need a heck of a rules engine. Again not impossible. Start peeling away more and more of the process though and what is a simple application becomes a very large code base.
Build a new jobmine
I say this not to discourage students, please build it. Show people how simple you see the process working, experiment, learn. Take the ‘Apple’ approach and offer only the features needed for the process to work and see what you have. It could be better. My observation is that the complicated process is being simplified by employers anyway (startups — usually run by uwaterloo alumni — recruiting at their own events, collecting resumes, offering on the spot). This works when honest employers do it but it could hurt students as well. Keep in mind, there are a lot of shady folks out there looking for free talent with really crappy jobs. CECS does an amazing job ensuring the quality and pay levels of student jobs.
Rather than get all crazy with disappointment try and understand the circumstances and do what you can do to try and help. Also, you may find the co-op process annoying but take a moment to thank the folks at CECS as those pay scale averages you enjoy as students are 100% the result of the things the staff do extremely well. Maybe software projects aren’t their thing at the moment but what matters more to students — the quality of the jobs or the few days of swearing fits at a dated interface?
…all that said. I am sad it didn’t work. I hope everyone involved reflects on what was right and what was wrong then gives it another shot. I fear that taking so long to go through a process to get to a pilot too much was invested to feel anything but a deep loss.