Zhihang Yao

agree, according to the [postgreSQL documentation ](https://www.postgresql.org/docs/current/datatype-character.html), using `text` has no bad side effect on performance

> There is a naming inconsistency: "old" names from v. 4 use the _ to separate words in column names (which I like, for the sake of readability). These columns...

I would like to have a proposal: Rename `objectid` -> `identifier` for the tables `feature`, `address`, `implicit_geometry` etc. Rename the current `identifier` -> `global_identifier` Rename the current `identifier_codespace` -> `global_identifier_codespace`...

> I'm a bit hesitant with `global_identifier` though. Sounds like we would guarantee uniqueness. The idea is, the CityGML `identifier` values can directly go into this column, because they must...

sorry, let me correct my text. I mean CityGML `featureId` here which should go into the `global_identifier` column. The CityGML `identifier` can be non-unique and going into the `identifier` column....

This is a very nice overview table.

Offline discussion: No changes needed.

I darkly remeber that we've discussed this topic when starting the design of the V5 and decided to use `feature` instead `object` due to some reasons. I also feel `object`...

I prefer the first proposal of @clausnagel. I also agree that using `feature` is much more common in GIS world mentioned by @muratkendir. Also, distinguisching `object` and `feature` sounds still...

According to the offline discussion, we decided to keep the names `feature`, `objectclass` etc. having been used in the v5-alpha-1 release.