Proposed features/Second opinion correctness survey
|Second opinion correctness survey|
|Status:||Draft (under way)|
|Tagging:||survey=date(s) in international format jjjj-mm-dd|
|Definition:||Stating with the current date that a certain object has been surveyed and is considered correct|
Currently there's no visible information on the quality of data of an object, all data seem equally well except for the occasional FixMe tags. When a mapper currently checks an object on-site, he can't state the correctness of the data of an object afterwards if no corrections are needed? On the other side nobody knows if an object has been surveyed by others. There isn't any indication which object should be surveyed, therefore mappers in an already well mapped region can't improve the data any further.
A tag "survey" is created with multiple values "date" separated by commas (date in international format jjjj-mm-dd). A user can state with the current date that he surveyed and considers the data of an object as correct. Multiple users can survey an object and add their dates. Any user can state the correctness except users who have created or modified that object within the last 10 month. Second opinions should be set either by others or at least about one year later by the creating/modifying user. This gives a hint on how actual the data is and if an additional check is necessary. This would solve the long term obsolescence of data.
If yet another user thinks an object with a survey tag isn't correct and should be modified (because it has changed), he may simply modify it. Alas he may be warned, that the survey tag will be deleted and all users will be notified. Any notified user may now simply ignore this notice or survey and approve/revert the change again.
No user needs special rights, any user can either add new data or approve/modify existing data. No burocracy is needed, all users are equal. Yet it will give anybody a hint which data shall not be modified without good reason. If changes were done unintended, the user gets a warning and might reconsider his action.
Data without a survey tag isn't more wrong than other data. It's only a hint if or how often an object is checked. Sure a multiple checked object is more likely to be correct, yet that's only a hint not a fact.
Besides making it visible to anybody that the data is considered correct by multiple users and therefore most probably correct, vandalism on existing data will be more difficult. Yet while not preventing vandalism, with the notification vandalism on existing data is immediately detected.
Example & Description
None so far
I've dropped the username from the tag since it can be looked up in the history. Also a survey tag with just a date makes it less obtrusive while still showing the survey information.
Please write comments into the development forum thread during draft state.
- check_date=* with usage stats for existing tags