The regional conference index, in the past tense

Between roughly 2007 and 2018 a dense circuit of small Ruby conferences existed alongside the two or three large ones. They were mostly single-track, mostly one or two days, mostly organised by a local user group rather than a company, and most of them have since stopped. This page indexes the ones with a public record, in the past tense, because that is what they now are.

What the years column means

It is the span the public record covers, not a claim about how long each event ran. Several of these certainly held editions outside the years shown, and a few of the earlier ones left very little behind at all. Where the record is thin the entry is shorter rather than filled in, because a confident date that cannot be evidenced is an invention, and an index of invented dates is worse than no index.

What is not in the table: attendance figures, ticket prices and speaker lists. Those were published inconsistently, they changed year to year, and reproducing them would give the false impression of a dataset rather than a list of events that happened.

The shape of the circuit

Reading the table by first year, the pattern is clear enough. A first wave around 2007 and 2008, mostly in the United States and mostly grown out of a user group. A second, larger wave from 2011 to 2015, spread much wider geographically. Then a thinning after about 2016, with the European and Asian events continuing after most of the American ones had stopped.

What the circuit shared was a format rather than a subject: a hundred to three hundred people, one track, a room borrowed from a university or a company, and a programme assembled by two or three people in their evenings. That format is described from the inside on organising a conference.

Venues, and why they are not listed

Almost every event in the table used a room it did not own: a university lecture theatre, a converted industrial building, a cinema out of hours, a company’s largest meeting room. The venue changed year to year for most of them, which is why the table names a city rather than a building.

The city is stable. The room almost never was.

The pattern worth taking from it is that venue choice at this scale is driven by what can be borrowed cheaply for one or two days, and the borrowed room then sets the capacity, which sets the ticket price, which sets how much sponsorship is needed. That chain runs in one direction and it is set out on the organising page, with the money side on conference sponsorship.

Why so many of them stopped

Rarely for a single reason, and rarely because of a lack of interest. The recurring ones are organiser fatigue after three or four years of unpaid evenings, a venue that stopped being available or affordable, and a sponsorship base that thinned as companies concentrated their budgets on fewer and larger events.

Several stopped deliberately and said so, which is a healthier ending than the more common one of simply not announcing a next edition. An event that says it has finished leaves a record; one that goes quiet leaves a website that slowly stops being true.

Where the sources come from

Each entry is drawn from what the events themselves published and from community indexes that tracked them at the time, principally the Planet Ruby conference list and RubyEvents. Where two sources disagreed on a year, the narrower span was taken.

The two editions of one of these events with the fullest surviving record are documented in detail at 2007 and 2008, and the programme design that almost all of them shared is on programme and tracks. For anybody looking at the circuit from the speaking side, the call for speakers is where it starts.

The kinds of room these events borrowed, with named examples, are on venues.

Questions