Started, finished or dropped
The day after GameKal’s first build, it gained a second screen that has nothing to do with release dates: a journal for the games you started, finished or dropped, kept separate from the release grid.
By the end of this post you’ll have a simple test for when a feature deserves its own screen instead of another column on an existing one.
This post covers why the journal is separate, the four states a game can be in, and the one rule that makes “dropped” work.
Two questions, two screens
The calendar answers one question: what is coming out, and when? Every row on it is about the future, or about a date that just passed.
The moment you track a game, a second question shows up: what did I do with it? Did you start it? Finish it? Give up halfway? That question is about you, not about the release, and it keeps going long after the release date is old news.
Putting both on one screen would make each worse. The calendar would fill up with games from two years ago, and your history would be chopped into months that have nothing to do with when you played. So they’re two screens that share one shelf.
Four states
Borrowed from the way Goodreads keeps a reading shelf, as covered in Two products, one calendar, every tracked game is in exactly one state:
| State | What it means | On the calendar? | In the journal? |
|---|---|---|---|
| Upcoming | you want it; it hasn’t started | yes | yes |
| Playing | you’ve started it | yes | yes |
| Completed | you finished it | yes | yes |
| Dropped | you stopped, on purpose | no | yes |
The dropped rule
The most important decision is in the last row: a dropped game stays in the journal and leaves the calendar.
Both halves matter:
- It stays in the journal because dropping a game is part of your history. Deleting it would make the record lie.
- It leaves the calendar because the calendar is a list of things you still care about. A game you’ve walked away from has no business taking up a day.
One state, two screens, two different answers, and both are right. That’s the test I use now: when the same piece of data needs to behave differently in two places, the places probably want to be separate screens.
Private by default
Ratings in the journal are private. They’re there to make your own list better, not to become a public review, and nothing about what you played has to leave your device to make the journal useful.
What I learned
Looking back, three lessons stand out.
Split by question, not by data. The calendar and the journal use the same shelf. They’re separate because they answer different questions.
Give “no” its own state. “Dropped” sounds negative, but without it every abandoned game sits in “playing” forever, and the journal stops being true.
Let one state mean different things in different places. Dropped is visible in one screen and invisible in another, on purpose.
What’s next?
With two screens and a shelf, the next problem was seeing the app the way a phone would, on a wide monitor: a device frame for the web preview.
Started, finished or dropped: three words, and the one that leaves the calendar matters most.