Productivity · Retrospective
Building for a ceremony date: the third pro bono panel

Chinese New Year falls on 17 February this year, and from today until 23 February a Yellow Jambhala panel I built pro bono is part of a festival at Bodhi Meditation PJ. It is the third of these panels.
The first was a Ksitigarbha blessing wall for the temple's Ullambana ceremony on 5 and 6 September: an LED wall running my animation, with three kiosks where visitors wrote a wish. The second, for Medicine Buddha, ran at the Medicine Buddha Birthday Festival on 15 and 16 November. Each one pairs an animation for the hall with a way for visitors to write a wish that the panel then carries. Yellow Jambhala is a deity of wealth, which makes it a natural subject for the new year.
The build itself is small. What makes it different from most software I have worked on is the date.
A date that does not move
A product launch can slip a week, and usually someone can explain why. A ceremony cannot slip. The festival opens on the first day of the new year whether or not the panel is ready, and the people in the hall are there for the ceremony, not for a screen.
So the planning runs backwards from that day. What has to be finished, and seen working, well before it? What can be decided late, and what cannot? The answer that matters most is about scope. When something is not working well close to the date, the choice is to cut it, not to fix it late. A smaller panel that runs is worth more than a richer one that might.
No rollback window
In web work, a bad release can usually be rolled back in minutes, often before most users notice. A hall during a ceremony offers no such window. A blank or frozen screen is seen by everyone present, and there is no quiet moment to swap versions.
That changes what I want from the design. The panel's resting state should be the finished animation, running on its own. Wishes from visitors add to it, but if they stop arriving, nothing should break: the animation carries on and looks as intended. The question I ask of every part is what it looks like when that part fails, and whether that is acceptable in front of people at prayer.
A week is not a weekend
The first two panels ran for two days each. This one runs for seven. A build that has to last a week asks for different things from one that has to last a weekend. It should start cleanly every day, survive being switched off and on again, and not depend on me being in the room. Anything that works only because someone was watching it closely on the first day is not finished.
Reuse is what makes pro bono last
Pro bono work fits beside everything else only if each new piece is cheaper than the last. This third panel follows the same pattern as the first two: write a wish, and watch the panel carry it. The deity and the animation are new; the shape of the experience is not. Building a new system for every ceremony would make each one a project. Keeping the pattern and changing the content makes each one a piece of work that can be finished in the time there is.
There is a second reason to keep the pattern. Anyone who wrote a wish at an earlier ceremony will already know what to do. Familiarity is a feature when the audience comes once or twice a year.
The meaning is not mine to design
In a devotional setting, the deity, the iconography and the words around them carry meaning that belongs to the temple and to the people who come to pray. My part is to make it run well and to keep the technology quiet. A blessing panel should feel like part of the ceremony, not like a demonstration of software.
What a fixed date teaches
Most software is built against dates that can move, and it shows. Scope grows because there is always next week. A ceremony removes that option. It makes me honest about scope sooner, it makes failure states a design question rather than an afterthought, and it makes reuse a practical necessity rather than a nice idea.
On the first day of the year, the work is either ready or it is not. I have found that a useful discipline, and a good way to begin a year.