User:Chriscf/On use cases

From OpenStreetMap Wiki
Jump to: navigation, search

If you want to write a feature proposal, and want to convince the minority of users that actually have a brain, you will probably want to frame your proposal in a way which suggests why your proposal is needed.

The exception

Features useful in navigation should always be mapped. Features are useful for navigation if there are many of them in strategic locations, not merely because there happens to be one on a street corner near you. If you can envisage "turn right at the X" as a common happening, beyond simply a handful of examples, it's probably a suitable feature to map.


A use case requires a few basic elements:

  • A simple goal
  • A credible user
  • A working solution


Goals are straight-forward expressions of intent, and are active rather than passive. For example, for an ATM, you might say "I need to get money". There are plenty of people who might empathise with this.

Straight off the bat, the following are not valid goals for individual features:

  • satisfying curious minds — OpenStreetMap is not in the business collecting data for the sake of it, we hope we can offer a viable alternative to commercial mapping products
  • reflecting reality — this applies to all features by definition, therefore it is not a useful argument to make in favour of a specific feature
  • we want to map X — this doesn't tell us anything useful about your proposal is a good idea; if you want to map something, just do it


Think about who will be consuming the information. There are applications we do not support. The following generally need some kind of guarantee of accuracy that can be backed up financially, something which we very explicitly exclude:

  • Research applications — as climatologists know, bad or incomplete data can invalidate results
  • Safety-critical applications — any situation where life is likely to be at risk as a result of bad data, e.g. use by emergency services,
  • Construction plans — these generally require accurate very-large-scale maps, which are beyond our technical capability

However, the following are non-obvious examples worth considering:

  • Haulage firms — almost every local paper ends up carrying stories about drivers who have followed their sat-nav and ended up in a ditch
  • Local authorities — often, these spend vast sums of money on large-scale mapping that we can't replace, but also on medium-scale mapping that we can replace