Counterintuitive warning about unusual Unicode characters with super-/subscript

[anonymous]

Hello, there.

Here, in France, numerous roads branching from a main one use the same ref, with a letter, which is often in subscript or, mostly, superscript; note that this ref can also appear in destination:* tags. The problem is that, in such cases, the JOSM (15492) validator warns about unusual Unicode character but, if I want to tag what's on the ground, I must use the super-/subscript. That could lead users to erroneously drop the subscript/superscript letters, replacing them by standard ones, whereas using super-/subscript is closer to reality. Could JOSM stop complaining about superscript/subscript characters, at least for all [*:]ref[:*] tags?

Awaiting your answer,

Regards.

[Klumbumbus]

Please link some examples.

[stoecker]

And pictures of the related traffic signs if possible - I didn't fully understand the description.

[Don-vip]

And also your status report. This ticket lacks a lot of needed information. If you're not at ease with English, please post in French, I'll translate.

[Don-vip]

Found one: ​https://www.openstreetmap.org/way/342883660

@anonymous: It would have saved me a lot of time if you had linked an example, as requested :(

[Don-vip]

Not sure this is a good idea to use the subscripts/superscripts available in Unicode:

• most of the characters are in the phonetic extensions block. Wikipedia says these characters are intended to indicate secondary articulation.
• as they are meant only for phonetic usage, the table is incomplete (there is no C for example): ​https://en.wikipedia.org/wiki/Unicode_subscripts_and_superscripts#Latin_and_Greek_tables

[anonymous]

@Don-vip: I simply did not check the ticket, so I didn't saw you were asking for an example. Sorry for the delay.

I'm aware that these sub-/superscript were not meant for this usage, but using them, when appropriate and possible, seems to improves OSM data by giving a ref closer to reality.

[Don-vip]

It is better visually, but if you search in my example "D36A" instead of D 36ᴬ you won't find the road in nominatim. It's just an example, you can't expect all data consumers to be able to recognize phonetic extensions in a non-phonetic string. My guess would be to write it as D36A in ref and D36ᴬ in another tag. visual_ref or something ? It would only be needed for the renderer.

I blame the departments for this stupid decision. Why didn't they simply use D36A? Using a letter as exponent is really strange.

[Don-vip]

In 15614/josm:

fix #18322 - allow uppercase superscript latin characters in *ref tags