reverted, vandalism

changeset/176854146

reverted, vandalism

changeset/176854139

thnx!

reverted wrt way/1263096014 & way/1261354298 : even though they are not "official", these ground paths do exist

changeset/176837569

once again reverted - no separate way exists; this is not an entrance; there is no sidewalk. Continued unconstructed edits may qualify as vandalism and lead to blocking

changeset/176791417

once again reverted - no separate way exists; this is not an entrance; there is no sidewalk. Continued unconstructed edits may qualify as vandalism and lead to blocking

changeset/176791417

once again reverted - no separate way exists; this is not an entrance; there is no sidewalk. Continued unconstructed edits may qualify as vandalism and lead to blocking

changeset/176791417

reverted - no separate way exists

changeset/176761230

reverted - not every crossing of building and footpath is an entrance; it's marked separately, node/10803344311

changeset/176761221

reverted - not every crossing of a footpath and a building is an entrance. StreetComplete should be fixed instead

changeset/176650298

rv wrt way/65069849 - unexplained removal of name (repeatedly)

changeset/176649663

Just a small note: the road leading to a parking lot should be tagged just as highway=service, without service=parking_aisle. Even the "main" road within parking lot that is necessarily used by cars to get in and out is just plain highway=service. The tag service=parking_aisle should only be used for smaller inner roads within the parking lot that separate rows of individual parking spaces.

hhhm, the link above was corrupted. Enter https://nt0.napr.gov.ge/NGCache?x= {x}&y= {y}&z= {z}&l= ORTHO_GEORGIA_4 but without the spaces after = symbols

I think it's better to split way/1061621209 in two parts: one north of the fence, needleleaved; and another south of the fence, broadleaved (or mixed).
Also, I recommend using imagery from maps.gov.ge, it's far more clear and precise. When editing in iD, open background settings to the right, click the three dots next to "Custom" and enter https://nt0.napr.gov.ge/NGCache?x={x}&y={y}&z={z}&l=ORTHO_GEORGIA_4

rv wrt node/13279188501 - it can be clearly seen in aerial imagery, if you use https://nt0.napr.gov.ge/NGCache?x={x}&y={y}&z={z}&l=ORTHO_GEORGIA_4

changeset/176277845

reverted - blind following of iD suggestions which are often faulty, e.g. not all Bank of Georgia branches are branded as Solo, TBC payment terminals actually accept coins etc.

changeset/176192029

reverted - blind following of iD suggestions which are often faulty, e.g. not all Bank of Georgia branches are branded as Solo, TBC ATMs actually accept coins etc.

changeset/176191893

"duplication and ... single responsibility principle"
Sounds like a totally artificial and far-fetched problem in contrast to the risk of leaving out vital info from some renderers.

Why remove the building from the relation in Kojori?! Or why remove addr:street from the remaining buildings?! They live together perfectly well, take a look at this relation, for example: relation/15056525

The system works fine with two strategies being used as the same time.

And I totally agree that a relation encompassing all parts of a street is useful for adding tags for the whole street, like wikidata and etymology. Although I differ on two points: imho, names in different languages are still better kept at all parts of a street (being quite a vital part of information that is better to be provided for even the most basic of renderers), and I prefer relation with type=street rather than type=associatedStreet for the reasons I once explained in the telegram chat.

Although they may be signposted as "supermarkets", according to OSM standards shops such as Kalata, Ori Nabiji, Nikora and Carrefour City are designated as convenience shops. In contrast to this, Carrefour Market, Nikora XL, Agrohub and most Goodwills are mapped as supermarkets.