What We Lost When We Stopped Doing UML

Wednesday 1st October 2025, 6:30 pm.

Speaker: Mike Ritchie

Venue: Computershare Limited, Edinburgh House, 4 North St. Andrew Street, Edinburgh EH2 1HJ

Refreshments and networking from 6:00 pm for those attending in person.

This talk will be preceeded by the Edinburgh Branch Annual General Meeting at 18:15.

This event is free of charge and open to all, though registration is required - you may not get past venue security if you do not register in advance.

Synopsis

The Unified Modelling Language was born during a period of industry innovation around all things "object". It was the golden era of object-oriented languages, including Ada, C++, Eiffel, and Smalltalk. UML neither invented nor particularly extended object-oriented ideas, but its history and development are inextricably linked to object-oriented design paradigms.

A few specialised domains still rely heavily on UML. Still, thirty years from its first release, it's rare to find a modern software engineering team routinely using it for systems or software modelling - despite OO design approaches still being prevalent in software design. What went wrong?

This talk will briefly outline UML's origins - its family tree is more complicated than you might imagine - and the goals of modelling languages in general.

We'll explore some of the factors that slowed the adoption of UML and the advances in technologies and engineering practices that have either replaced or significantly reduced the need for structured systems modelling.

Finally, we'll ask whether it still has a place in modern software engineering and whether some of UML's positive aspects—there are some—have slipped between the cracks.

Many who worked in software engineering during the period of UML's emergence, if asked the question, "What did we lose when we stopped using UML?" would quickly reply, "Absolutely nothing".

Well, let's see.

About the speaker