Hello again!

This time these signs are properly tagged.

The traffic signs added in this commit are using the "Traffic signs_ES" JOSM tag preset by yopaseopor.
This adding of traffic signs comes from an agreement from the Catalan community for mapping the traffic signs using his preset and with the country IDs in osm.wiki/Tag:traffic_sign:id=, human-readable values in osm.wiki/Tag:traffic_sign=, and adding a relation similar of the type=node for tagging the traffic signs, where is attached and the place of the road where it starts to take effect. The plugin in question has been updated several times to reflect and be able to work with all of this. Think as if it were a "regional standard", it was talked on the Telegram community group, and no one opposed it.

This traffic sign tagging scheme is very complex and detailed, but this also causes it to be super verbose and state the obvious in many cases.
But also I consider that having a verbose scheme is good because it allows more flexibility. For example, if there is more than one destination in a single direction, there are numbered tags for specifying all the destinations.

The fact of splitting values into different tags, makes them easier to edit and also, if in the future, the system for tagging traffic signs changes (again), then we would be able to easy adapt the current signs into the new system via automated edits; instead of having to potentially redo the signs because all the values were mixed.

This explains the new tags, although this commit has added 3 different roundabout destination signs (one of them being the one you've linked), so I'm pretty sure that most of these tags are also in the other signs.

The only other (simpler) standard for roundabout destination sign tagging that I know is the Spanish roundabout sign preset from the RoadSigns JOSM plugin (osm.wiki/JOSM/Plugins/RoadSigns).
Using this plugin, the sign of the node you've linked would look like this:

destination:colour=blue|white|blue
destination:ref=A-27|C-37|A-27
destination:sign=pier;plane;ave||
destination=Tarragona|Alcover;Reus;Prades|Valls;Lleida
junction=roundabout
traffic_sign=ES:s200
turn:destination=left|through|right|slight-right

For this sign, this preset works flawlessly. The tags are short and few, and everything fits perfectly. I really like how it looks, it's simple yet effective.
Unfortunately, this is also a pretty simple roundabout sign, and more complex signs would have great difficulties for "fitting in".

On the same roundabout, on the exit leaving the A-27 and entering the roundabout from the south, there's also this sign:

- OSM node (mapped with the same preset as the one sign you've linked): node/12645203477#map=19/41.270094/1.232110&layers=N
- Bing Streetside: https://www.bing.com/maps/?cp=41.269972~1.23202&lvl=15.0&pi=3.4&style=x&dir=350.4

You can see that the left roundabout exit has C-37 as the main road reference and Alcover as it's primary destination, but there's also a secondary reference of C-14 with Montblanc.
This is because Montblanc is an important city, and to reach it, you have to first go to Alcover via the C-37, and once there you have to grab the C-14 road towards Montblanc. This roundabout does not directly connect with the C-14 road.

How would these references be tagged in the RoadSigns plugin? My guess would be "destination:ref=C-37;C-14|A-27|C-37", but such an approach will clash front-facing with other signs that have several primary references.

Signs like this one: https://www.bing.com/maps/?cp=41.141731%7E1.233511&lvl=17.0&mo=om.1&pi=0.4&style=x&dir=213.3
(not mapped in OSM yet, I think)

This left roundabout exit goes towards Tarragona and has the road references of A-27, A-7 and T-11. This is because after the left roundabout exit, there's a brutally massive junction in which you can go to either A-27 (the main one that this roundabout connects to), the A-7 or the T-11, and you can enter Tarragona through any of these three.

Following the RoadSigns plugin, it's destination:ref would look something like this: "destination:ref=A-27;A-7;T-11||A-27", and this tagging makes sense here, but now we still have the question of what to do the C-14 Montblanc case of the previously-mentioned sign.
Placing "C-14 Montblanc" with both secondary reference and secondary destination text all as a secondary destination (or as a new line of all the destinations together) might work, but this method will probably not highlight the C-14 reference text as much as the other references (thus defeating the reason of why we place references in other tags).

Do you want another example of a complicated roundabout sign? Here's a big one: https://www.bing.com/maps/?cp=41.279575%7E1.257056&lvl=17.0&mo=om.1&pi=-3.2&style=x&dir=251.5
(this one is not yet mapped in OSM, I'll probably tackle it once I go arround the Valls beltway)

On this sign, the RoadSigns plugin would have problems rendering the white background of the left exit while being yellow for the secondary destination of Camp de Tarragona with the AVE symbol.
And then there's the hodgepodge of the right exit, with a primary reference of the N-240 road towards Valls and Lleida with white background, a secondary reference of C-51 towards El Vendrell with a white background, and a tertiary blue background box of the top-level motorway AP-2, with European route reference E-90 and the sign of paid toll.

Representing this sign using the RoadSigns plugin would be impossible, or we would have to do a lot of compromises and "tag approximations" that then we would have to "decrypt and unroll" should we want to render this sign into an image or a model or interpret it for whatever reason or propose may arise.

The approach of the RoadSigns plugin works well in simple signs, but unfortunately fails on the more complex ones. The current, by being verbose as heck, we have a chance to adapt it better to whatever surrealist painting of a sign comes. I don't know of any other traffic destination sign tagging system other than the one used here and the RoadSigns plugin, but if there is, it'll probably need more tags than normal for specifying these and all the edge cases that may arise, and then the current situation of new tags will also reappear.

Sure, this complex tagging scheme is an overkill for simple signs like the one you've linked, but will be useful for better detailing the more complex signs. And there are many other tagging schemes in OSM that are overkill for very small features (for example: a very short bus/metro line that still needs plenty of tags for the ways, stations and relation about names, operators, networks, and such…).

Traffic signs are very complex, there are many of them, and very different, so it makes sense that it's tagging system is also complex and detailed.

And for your point of the Tarragona pier, plane and AVE symbols being all in different destination symbol tags, sort-of implying that there are all alone in new lines; I agree with you.

Due to splitting values into their own tags, this also means placing values like those symbols in independent tags. Although, I suppose that there will be very little confusion; as it is highly unlikely that these symbols will be alone in their lines. I will probably change this in an upcoming commit, because I also don't like it much xD

Sorry for such a massive reply. I hope that the examples that I provided help in understanding the logic of this scheme.

I understand your logic behind some of your tag changes and rearrangements, but I disagree on some.

The name fits a description better, true, but so are literally all the names of monuments, plaques and memorials. Very few memorials do have a "title" in it that is its official name. Many of those follow a scheme of "type of memorial + short description of it". Look around in OSM and you would see that the vast majority of memorials are mapped with a name, some may come from an official government memorial database.

Never heard anything about the -t- thing; but I personally still prefer ":translation:".

I think this will be an easier way for applications to drink this data easier, searching for "inscription:translation:XX" tags seems faster than deciphering "XX-t-XX" tags. This also is more clear to a human reading the tags that the contents are a translation to Spanish.

About the "source:inscription:es" tag, sorry, my mistake. The "source:inscription:es" tag confused me because the tag ends with a Spanish ":es", but the language in the value is Catalan. Your "source:inscription:es" should be "source:inscription:translation:es", because the original inscription is not in Spanish, only the translation is.

I'll revert the translation tag back to "inscription:translation:es", readd the name, move the source of the translation to "source:inscription:translation:es" and move the "inscription:translation:es:description:XX" to "description:XX".

I like the sound of "inscription:translation:es:description:ca" xDD and I would like to add it also, but I guess that description will be more used, so I'll go with that for now.

Hello! Your contributions of this changeset make no sense.

You've moved the translated inscription to a new tag "inscription:es-t-la"
This -t- setup is the first time that I heard about it, and despite it is used elsewhere (https://taginfo.openstreetmap.org/search?q=-t-), it seems undocumented (no wiki page nor anywhere else online talking about it that I could find).

I've done it with the "inscription:translation:es" because in taginfo there's also previous usage of the inscription:translation tag (https://taginfo.openstreetmap.org/search?q=inscription%3Atranslation; it was used in a "inscription:translation:pl" before my changeset) specifically for "inscription".
Besides the ":translation:" tag (https://taginfo.openstreetmap.org/search?q=%3Atranslation:translation), I personally think is cleared and easier to understand than language codes with a t in the middle.

Also, you've moved the name value to the description, deleted the multilingual descriptions and moved the Catalan description to the Spanish source (source:inscription:es).

May I ask your logic behind these changes?

Buenas! He revertido tus cambios en mi changeset/161710534 (changeset/161710534).

Passatge d'en Gassó es una vía privada, cierto.
Pero su forma y nombre aparece en el mapa tipográfico del ICGC y en fuentes de geo.valls.cat, así que considero que es apropiado añadirlo a OSM. Además de que su forma es perfectamente apreciable desde imágenes por satélite actualizadas (como la ortofoto del ICGC que has usado como fuente en este changeset).

La página de la wiki "Limitations on mapping private information" (osm.wiki/Limitations_on_mapping_private_information) menciona que limitaciones y restricciones se aplican en el mapeado de propiedades privadas.
El mapeo de Passatge d'en Gassó cumple con estas directivas debido a los siguientes puntos que la página de la wiki menciona:

- "Mapping private buildings, private roads (including driveways), and private parking is completely acceptable. Add access (access)=private=* (access=private) as appropriate to roads, parking lots, etc. However, annoying artifacts may appear in applications like Pokémon Go with short driveways."
- "It is also considered fine to map existing roads/paths, if using them is forbidden; use access (access)=no=* (access=no)/access (access)=private=* (access=private) rather than deleting them."

Tu acción de directamente eliminar vías privadas no cumple con estas directivas. Añadir un par de "access=private" y "barrier=fence" no cuesta nada y hace que el mapa este mas completo.

Heyas! Thanks for pointing this out.

Your comment has made me aware of how weirdly I mapped this sign.
I used a JOSM traffic sign preset for tagging this, and it seems I applied the 
Spanish-style orientation in city sign preset, when this sign is a Catalan-style orientation in city.
The difference between the two is that the Spanish-style is colourful and bigger, while the Catalan-style is mainly in boring white background.

This sign has nine panels of destinations with different arrows, text and icons each, so it would be impossible to condense all of this into 6-ish or so tags.
The traffic sign OSM wiki page says how to map traffic signs in general (as in orientation, if it's in a pole, etc.), but not what values to use in each sign, so it doesn't help in this case.

Also, looking at the tags it put out, it seems to combine roundabout-like exit with numbered destination panels in a weird way.

I'll remove the tags and refactor this sign in an upcoming changeset.

Buenas! En este changeset has movido el nodo del survey_point NGW79 de Valls (nodo con ID 1142949994; node/1142949994).

El problema es que este survey_point tenía highways conectadas y actuaba como intersección; y al mover el nodo, la intersección se ha movido y ha quedado desajustada.

He hecho este changeset para arreglarlo (ID 160510112; changeset/160510112), donde he desconectado las vías del survey_point y las he reconectado a un nuevo nodo de intersección en el mismo sitio de antes.

La causa de este problema ha estado muy probablemente por el hecho de que antes este survey_point estaba en medio de la calle y al trazar las vías de la zona se ha usado su nodo como nodo de intersección por el mero hecho de que el nodo ya estaba allí (aunque puede que parte de la culpa la tenga yo en base a unos edits de hace unos meses, ups… 😅).

No te estoy echando la culpa de esto, ni de lejos. Solo te dejo este comentario para que no te estrañe si luego te sale mi changeset en el historial de dicho survey_point.

Salutaciones y felices fiestas

Bones! Nope, el "hop=yes" es un typo; hauria de ser "shop=yes", disculpes 😅 

He arreglat el typo amb aquest commit: changeset/160517321

Poso nodes com a "shop=yes" si no sé quin tipus de local és. En aquest cas hi ha un aparador gegant d'articles, però desconec del cert si és part del basar del costat o una tenda independent, no sembla que estigui retolat. Quan hi tombi per la zona, provaré de completar-ho.

Gràcies per l'avís i disculpes!