Blog
Notes on scheduling, and on building this.
Writing about the craft of the master schedule, what we learn from schools, and the decisions behind nOS.
Writing about the craft of the master schedule, what we learn from schools, and the decisions behind nOS.
We would rather publish nothing than publish filler. The blog opens when the first piece is worth your time — the format and the standards are already set, and you can read both.
Not a content calendar — an honest list of what we think is worth saying. If one of these would be useful to you sooner, say so and we will move it up.
Working through the real arithmetic of a course that runs once — how quickly two singletons in the same period become unsolvable, and how to spot the collision before you build around it.
nOS computes a mathematical upper bound on how many requests any scheduler could satisfy with your data. Why that number is more useful than a percentage, and what it means when you hit it.
A close look at Columbia CPRL's About Time report — what it found about who builds the schedule, what tools catch, and which students absorb the compromises.
The design argument for a pending layer — including why the AI proposes rather than acts, and what we got wrong the first time we built it.
The practical version: what to do in August, what to lock by January, and which decisions are much more expensive to reverse than they look.
Every item above is a planned topic, not a published article. Nothing here is live yet.
The research on what building a master schedule by hand actually costs — with every figure sourced.
ReadThe realistic sequence, when in the school year to attempt it, and what happens if you decide to leave.
ReadGuides, references and tools — including the free browser-only schedule audit.
Browse