I am making this because I want to have highly customisable notifications for my phone given certain conditions are true, and, like Gary Kasparov once noted in a human vs computer vs human+computer tournement, the quantative abilities of a machine can be utilised most effectively when synergised with my human ability for qualification. Here's some examples of ways in which I would use this:
Let's say I am a programmer working in the square mile and I take the Hayes train from Cannon Street station to get home every day. I have this app and I decide to set up a notification with the following definition:
GIVEN: I am more than 100m from the work building (e.g. I just left work) It is past 5:30pm (e.g. I'm not going for lunch, I'm going home)
THEN: Bring up a notification with the departure times of the next three Hayes trains.
Note: I choose the next three, because I want to know how much longer I'm going to wait if I'm lazy and walk slower or I stop off for a snack at tesco.
Let's say I am me, but better at programming, and I have a combined/advanced mealplanner/spend-analysis programs and databases, and they are running on a server listening on port 80 of my machine at home. I have set up a notification with the following definition:
GIVEN: I am in a food shop (in tesco, sainsburys... etc.)
THEN: Bring up a notification that tells me the last 3 items of food I bought (from spending data) and also their nutritional value (from the mealplanner nutrition table).
Note: From this information, I know what food I've got in and what macronutrients I lack given my current inventory - and hence what kind of foods I ought to buy. Moreover - this could be automated, and the notification could simply be a list of suggestions.