On beta testing, PC magazine cover discs, and what an institution with nothing to lose taught me about power.
In 1997, four final-year computer science students at Leeds Metropolitan University were called into a room one by one and presented with a letter. Sign it, and confirm that all intellectual property in our final-year project belonged to the University. Refuse, and the institution would consider whether to award our degrees.
We had built an interactive tutorial for Borland Delphi as part of an assignment requiring real user testing. It was free. It was non-commercial. The technology was self-taught, not on the curriculum. Around thirty thousand people had downloaded it. PC Plus had included it on their cover disc. Apparently, this was the problem.
We signed.
The acronym, and why it still makes me wince
The product was called Delphi-Dolphin. The name was an acronym I worked backwards from a domain registration: Delphi ONline Programming Help INteractive. The web in 1997 was young, the available domain pool was already thinning, and the contrivance required to make the letters fit a recognisable word was significant. I am told the cringe will fade. Twenty-nine years later, it has not.
The application itself was something we cared about. Our brief required user testing. Putting the software on a website and letting strangers download it for free was the most direct route to the largest pool of testers we could reach. We were not building a business. We were trying to learn whether the thing worked.
It did. The download numbers told us so. Iterative. Responsive. Genuinely attentive to feedback. This was the intention, and people seemed to feel it.
A jewel case, a bottom strip, and an institution paying attention
The confirmed magazine placement is PC Plus SuperCD no.38, August 1997. The headline software on the disc was Borland C++ 4.5, followed by Hypercube HV++, and Simply Personnel. We sat on the bottom strip alongside a five-hundred-component Delphi library. One line of text. No name, no fanfare.
The bottom strip mattered more than it sounds. Cover discs were how the UK acquired software before broadband. Inclusion meant editorial assessment had judged the application worth distributing to a mass audience. For a student project, this was an unusual outcome.
The University noticed. What followed was not a celebration. It was a legal letter, presented individually rather than to the team collectively, threatening the most significant consequence the institution had available.
What the threat was really about
Consider what that threat asked of a 21-year-old four years into a degree programme. Four years of work. Four years of debt. The finish line visible. A signature against the prospect of losing all of it.
The University's position was weak on the merits. We had credited the institution in the software. None of us had earned a penny from distribution. The technology was not taught on the course. We had built the application in our own time, using skills we had acquired independently, in service of an assignment that explicitly required real user engagement. The result had exceeded the brief, which is what an educational institution should want from its students.
What the institution did instead was take advantage of the fact that we were young, unrepresented, and acutely vulnerable to the one threat that mattered. It was institutional cowardice with a legal dressing applied to it.
What I will say is that being summoned individually, presented with a pre-drafted letter, and threatened with the worst available consequence, taught me something about institutional power that has stayed with me. Institutions weigh what they have to lose against what you have to lose. When the asymmetry is large enough, the merits stop mattering.
What the project became, and what stayed with me
Delphi-Dolphin did not die with the letter. The website evolved into a PHP and MySQL community site with tutorial libraries, a discussion forum, an article submission workflow, and an RSS feed for Borland announcements. The HTML tutorial content was eventually handed over to Project Jedi's Education team. Active development continued until 15 April 2003. By then, the UK market was moving from Delphi to C Sharp, and I followed it.
Twenty-nine years on, I am building AI-powered products in the social housing sector. The technical tools are unrecognisable. The instinct is identical. Build something that works. Put it in front of the people who will use it. Listen to what they tell you. The discipline of working directly with users, rather than designing for an abstracted version of them, is the same discipline that made a student project resonate well enough to land on a national cover disc.
The institutional lesson has stayed with me in a different way. I have watched organisations behave like that University ever since: defensive, asymmetric in their use of power, more concerned with controlling outcomes than with the merit of what they are controlling. I have also watched what happens when an organisation takes the opposite view, treats the people doing the work as its strongest asset, and lets them put their work in front of real users without first protecting itself from the possibility of success.
You get more interesting outcomes the second way. You get more interesting people, too.
I still own delphi-dolphin.com. You are reading this on it. The University did not get to decide what the project meant in 1997, and they do not get to decide what it means now.