Key takeaways
- The conditions that made a once-a-year conference session the main venue for unfinished work no longer exist.
- The same function now happens continuously, in early access, game jams and public test branches.
- The clearest descendant is the experimental feature toggle in live games, which asks players to do what the workshop audience used to do.
A session where developers show unfinished games to a room of peers made sense in a world where there was no other way to do it. No storefront would carry a prototype, no audience existed for a build with three levels, and the only distribution for an idea was a projector and an hour at a conference.
That world ended. The interesting question is what replaced the function, because the function was never the session – it was the feedback.
Where the feedback moved
Three places. Early access put unfinished builds directly in front of paying players, which is the workshop’s premise with money and scale attached. Game jams industrialised the Carnegie Mellon rules, so a weekend now produces thousands of prototypes rather than fifty in a semester. And public test branches let live games ship a change to willing players before committing to it.
Each of those is larger and faster than the session. Each is also noisier. A room of three hundred designers gives you a considered reaction. A hundred thousand early access players give you a review score, which is a different instrument measuring a different thing.
What was lost
Selection. The workshop had curators, and curation is what made a line-up worth reading a decade later. The 2018 session put nine games on a bill and three of them mattered enormously. A jam produces two thousand entries and no one can tell you which three matter. A curated bill of fourteen avoided that problem simply by having a person choose.
The other loss is the permission to fail in public. Early access punishes an idea that does not work – refunds, reviews, a store page that will follow the game forever. A session slot costs nothing if the answer is no, which is exactly why the answer was allowed to be no.
What survived, in an unexpected place
Open a modern live game’s world creation screen and you will find a list of toggles marked experimental: features the studio has built but not finished, with a warning that they may change, break your save, or never ship at all. Turning one on is opting in to test an unfinished idea and report back.
That is the workshop’s bargain, moved from a conference room into a settings menu and offered to millions of people instead of three hundred. The audience is no longer designers, which changes what kind of feedback comes back – but the structure is the same: show the unfinished thing, watch what people do with it, keep it or drop it.
This site covers one of those games, and the mechanics of its toggle system are explained in the workshop archive page, which is where the two halves of this domain’s history meet. Whether that is a fitting inheritance or a diminished one is a fair argument. Our own view is that it is both: the reach is enormous and the judgement is gone, because a million players reporting a crash is not the same as a room of peers telling you the idea is boring.
Does the session still matter?
As a venue, less than it did. As a record, more. The line-ups preserved on this domain are one of the few places where you can see what a generation of designers thought was worth trying, including the parts that failed – and failure is the data that early access reviews and jam pages systematically delete.
A session whose best evidence is a list of failures is an odd thing to defend, and it is the only defence that holds.
That is the argument for keeping the archive rather than redirecting it: not sentiment, but that the record of what did not work is the part nothing else preserves.
What early access actually measures
Early access looks like the workshop at scale, and it is measuring something different. A conference audience answers the question “is this idea worth pursuing”. A paying early access audience answers “is this worth the money now”, which is a product question wearing a prototype’s clothes.
The difference shows in what each format kills. The session killed ideas that were boring. Early access kills ideas that are unfinished, which is every idea at the point where feedback would be most useful.
The scale problem in jams
A modern game jam produces thousands of entries over a weekend. That is a triumph of the format described in the seven-day rules piece and it creates a discovery problem the original never had.
Fifty prototypes from one semester could be reviewed by the people who made them. Three thousand entries cannot be reviewed by anyone, so the good ones surface through the same popularity mechanisms as everything else online – which selects for what is shareable, not what is interesting.
What a modern version would look like
If someone rebuilt the workshop today, the valuable parts to keep would be curation, a hard time limit per presenter, a room of peers rather than customers, and permission to show something that does not work. The parts to drop are the annual schedule and the physical venue, both of which were constraints rather than features.
Our own view is that nothing currently occupies that space. Storefront demos, jam pages and public test branches all collect reactions at enormous scale and none of them provide judgement. A curated hour with three hundred designers in it is a small thing that no amount of scale replaces.
Why this domain keeps the record
The archive here is not nostalgia. It is a list of what a generation of designers thought was worth trying, with the failures still attached – and the failures are the part that every modern venue deletes. A store page for an abandoned prototype disappears. A jam entry from 2019 is unreachable. A session recap from 2011 is still readable, because someone kept the page.
Images: Thekla Inc., via The Witness Steam store page.