reconsider mapcss function osm_timestamp()

[Klumbumbus]

For ​https://www.openstreetmap.org/api/0.6/node/4471457459 the mapcss function osm_timestamp() returns an integer number (1477838904) which is not really useful I guess.
It would be better if it would return the human readable format (2016-10-30T14:48:24Z)

[taylor.smock]

If this change is made, we'd probably want to be able to parse timestamp values as well, e.g.

way[osm_timestamp() < "2016-11-01T00:00:00Z"] {
  // Do something
}

I don't think we currently do so.

[stoecker]

Why is the integer not helpful?

[mikeho]

JOSM, Window: "History for way/node <number>" ( Ctrl + h ) shows the text "Version 3 created on 1/20/20 6:16 PM". This is more understandable for humans than "Version 3 created on 1615578551". In a MapCSS it is the same. That's why we need a function that outputs the osm_timestamp as a formatted string, for example osm_timestamp_formated_string() ( like the strftime() function ).

The format "2016-10-30T14:48:24Z" can be an alternative to national date formats like "30.10.2023, 14:48:24 Z".

node[...]::layer_xyz { ...  text: concat("Last edit by ", osm_user_name(), " on ", osm_timestamp_formated_string() ); ... }

The result "Last edit by mikeho on 2016-10-30T14:48:24Z" is easy to understand. "Last edit by mikeho on 161030123" says nothing to humans.

[Klumbumbus]

Replying to stoecker:

Why is the integer not helpful?

For "human reading" it is useless. If it can be used for sorting or something like that, then 2 functions would be the best.