Movies

A film and TV catalogue that runs on my own machine, with 12.7 million IMDb titles in SQLite and posters and streaming availability from TMDB. My watch log is kept as plain files in git.

Movies is a film and TV catalogue that runs entirely on my own machine. It keeps about 12.7 million titles from IMDb's public datasets in a SQLite file. Posters, cast, summaries and where each title can be watched come from TMDB. Those requests to TMDB are the only thing that leaves the machine, and there is no account or sync.

Browsing happens on a single screen backed by a single query. Search is a filter on that screen rather than a separate page, so "science fiction I have not seen, matching 'star', newest first" is one question. The filters cover text, genre, year, runtime, rating, cast and crew, streaming service and age range, and any combination of them can be saved as a named view.

Each of the twenty-two genres has its own colour on the cards. Related genres share a hue, so horror and thriller are both red, and in practice there are nine colour families. That makes it possible to read a wall of posters by category before reading any text.

Where the data comes from: IMDb's datasets and TMDB fill a local SQLite catalogue. The watch log is kept in plain files next to it, and one browse screen queries all of it.

The watch log, ratings, reviews and lists I write are kept as Markdown and JSONL files in git. Those files can be read without the app, and the database only holds a copy derived from them. The interface can be switched between English and Turkish. The source is not public.