[Tagging] Deprecation - waterway=riverbank vs water=river

[Dave F]

On 10/02/2021 13:36, Tomas Straupis wrote:
>
>    It is important to understand at least the basic rules of IT:
>    1. Do not touch it if it works.

Amend something if it can be improved.

>    2. If you want to change something in widespread use, benefits of
> additional work for change should be seriously outweighed by the
> benefits of the new thing

A few hours in JOSM

> (and water=river has no advantage over
> riverbank whatsoever it tags exactly the same thing).

It fits in with all the other natural=water entities, making data 
filtering easier
It allows canals/streams etc to be accurately tagged with 
water=canal/stream.


>    One has to understand, that these key=value pairs are not only
> something "in the database". These are values in huge amount of
> products, QA rules, training materials, books as well as peoples
> minds.

They all have to work with what's in the database. The two schemes have 
run side by side for a few years now. All major renderers render both 
schemes


> Any change of widespread tags is very bad.

Not changing something for the better based on fear of change is worse.

>
>    What we have now:
>    1. waterway=riverbank is original (older) tagging.
>    2. waterway=riverbank outnumbers water=river in usage.
> waterway=riverbank 251 166 and water=river 152 993*.
>
>    Therefore water=river is the clear candidate for removal if we look
> at basic principles of IT.

Majority doesn't equate to quality.

>
> P.S. Most water=river usage is generated by rogue behaviour of iD
> coders who push their own opinion on different tags and force
> unknowing users to change tags by lying about "deprecated" tags.

That sounds like a lie to me.

DaveF