Proposal talk:Note suffix

From OpenStreetMap Wiki
Jump to navigation Jump to search

Current Use

Checking the Taginfo statistics and excluding all language suffixes, the current use seems to be very much in favor of note as a prefix.

I agree there is the possibility for rare ambiguities, but for a tag solely for human reading and not automatic evaluation this is acceptable and doesn't justify deprecating tags that are used hundreds of thousands of times. --Mueschel (talk) 21:39, 30 June 2022 (UTC)

Most of the most used keys with the note: prefix don't actually describe another tag, e.g. while there are >100k note:city tags none of them are used in conjunction with the "city" tag.
There are only three note: prefixed keys with more than >10k uses that actually add a note about an existing tag:
36 562 note:old_railway_operator
16 558 note:name
15 496 note:lanes
For the :note suffix key the most used keys that add a note about an existing tag are:
6 418 minsk_PT:note
5 963 fire_hydrant:note
3 924 maxspeed:note
3 569 surface:note
So yes while the note: prefix is used more often the difference appears to be less than one order of magnitude.
Yes while the values of these tags are "solely for human reading", I still think that there is much benefit to be gained by letting machines reliably map tag-specific notes to the tags that they belong to.
For example a GUI map editor should display maxspeed:note next to the input for maxspeed. In order for editors to do that soundly we need to establish an unambiguous convention.
--Push-f (talk) 21:49, 1 July 2022 (UTC)

Similar Tags

I don't think there should be a new tagging rule for a 'note' suffix separately. 'description' is a tag part used in a very similar way and should be defined in the same way as 'note'. --Mueschel (talk) 21:43, 30 June 2022 (UTC)

The 'description' tag is different in 1) that it is meant to be displayed to map users (whereas 'note' is only intended for mappers) and 2) that it is not generally used to add a description about a specific tag, which is exactly where the ambiguity arises that this proposal seeks to resolve. --Push-f (talk) 21:49, 1 July 2022 (UTC)

Sync with `check_date:*`

As far as I remember, @Westnordost decided to use `check_date:<OtherTag>` (as a prefix) for StreetComplete to work around the issue, that the `check_date` as a suffix has the disambiguity of being interpreted as a tag. Eg. 'parking:lane:check_date' is not an attribute of `parking:lane` (as `parking:lane:surface` would be). Another reason was, that using check_date as a prefix allows to scope sub-tags more easily. check_date:parking:lane would be for all the sub-tags of `parking:lane`. It is like "parking:lane:*:check_date". I think it would be great to use the same logic for check_date as for note (and description).--Tordans (talk) 15:17, 4 July 2022 (UTC)

The whole ambiguity of the note: prefix arises from having both language suffixes as well as key suffixes (as explained in the Rationale of the proposal). A date obviously does not have language suffixes, so it doesn't have the same problem. --Push-f (talk) 15:42, 4 July 2022 (UTC)
It sounds like a "one problem solved, two problems added" situation.--Tordans (talk) 08:23, 5 July 2022 (UTC)
Well the name:langCode=* convention is used millions of times. Are you suggesting that we should use a different syntax for language-specific notes? Because simply using note: as a prefix does not solve the ambiguity problem this proposal is seeking to address. Also I am not sure what problems you're referring to. Yes the :note suffix would get a special meaning and could not be used as a subkey with a different meaning. I don't think that this is a problem because a :note prefix with a different meaning would be confusing anyway. --Push-f (talk) 09:03, 5 July 2022 (UTC)

One object problem

There is only one osm object which has this problem:


It's not enough to change how I tag. Maraf (talk) 19:47, 10 July 2022 (UTC)

The status quo doesn't determine reality. It is actually 2 objects, 1 note:bin=* node 3916622268. This is caused by under-mapping of those features and languages. It is commendable effort that destination:*=* tries to use destination:lang:*=* around destination:to=*, despite the lack of Tongan guidance (or even exits in the real world), and my reservations about it. The wiki's automatic compound documentation fails too, although it is not authoritative, and could somehow be fixed. --- Kovposch (talk) 06:52, 11 July 2022 (UTC)