rename a given key

hi,
cause is this contribution: http://wiki.openstreetmap.org/wiki/Talk:Tag:boundary%3Dprotected_area#Misleading_tag_names
what do you think: would it be usefull and easy to rename a key (protect_id to protect_type)?
on the other hand, this argue might be dissolved, if required, be a further key like (the existing) site_id … thanks for answers, cheers, crom

First off, the proposed page does not seem very coherent. There exist already leisure=nature_reserve, boundary=national_park and a tag called http://wiki.openstreetmap.org/wiki/Proposed_features/conservation suggested by someone you might have heard about (http://wiki.openstreetmap.org/wiki/User:Timbl). There also exist existing tagging schemes for Ramsar sites (http://wiki.openstreetmap.org/wiki/Ramsar_site), and some conservation areas in the UK are also tagged with their legal status, such as SSSI. So there is a danger of tag proliferation, although this is understandable given that it has not yet received in depth attention.

Second. The OSM standard for using external IDs is to use the ref=* tag, in this case I would expect to see it name-spaced, e.g., ref:iucn_id. Use of tags with id and type in their names is not really a good idea as they may get very confusing on some objects.

Thirdly, bulk search and replace on tags is pretty much abhorred by active mappers: it adds loads of cruft to the local change logs/history, may obliterate something being done locally, or a current piece of mapping which is work in progress. See http://wiki.openstreetmap.org/wiki/Automated_Edits.

Lastly, the whole idea of a numeric based tagging system is also not to be encouraged. So I’d much rather such tags were removed and replaced by something meaningful: the wiki is a documentation tool, not a meta-data dictionary.

Hello SK53,
thanks for your view!

please you have details, proposals?
It includes all protected areas with 20 or 30 items,
stops stagnancy and growing overflow (“proliferation”) and its still open like before.

they were “incoherent” “too”, technical an by a wrong use - or not use resp. areas are not tagged.

there is any known “problem” with the UK´s nature-reserve system and the IUCN.
It should be possible to solve that and order the ile-once into the table.

what you think about use & meaning of ID (identifyer) - is it “misleading” for you?

You say I´ve also, beside this assumed ID-“syntax-error”, to add “ref:”
But only for IDs and IUCN or protect_id are classes or something, no IDs.
So here no ref:

so - the best tag for osm would be still (out of two votes):
protect_type=4
(with numeric value like now)

So I should change first these wiki-side, primarily rename the IDs and than changing the keys: I´ve never done somthing like this, but I can try to download a data.osm of the key protected_area (60MB), change the key protect_id (1238 hits in 1 files; changing: only protect_id with clear numbers from 1 to 99) and … upload data.osm with JOSM again - may be in some smaller files? Anybody knows, if that works? Or any osm-guru can do this … ?

differing, for some. Numbers, letters, etc. … variable, somehow or other? I think, it might looks more friendly to join an abstract code (simple number) with my local name, than my local name with a foreign name? Because I stay in my known space.

On a small level it should be passable…? and boundary=administrative, the admin_level?

cheers, crom

Hello,
this wiki-page was been updated:
http://wiki.openstreetmap.org/wiki/Tag:boundary%3Dprotected_area
regards, crom

Using a level type tag and forcing everything into a hierarchy is the fundamental problem: it doesn’t work very well with administrative boundaries I have no expectation that it will work with nature conservation zones. The proposal is also so complex it seems designed to deter mappers.

hi SK53, thank you.
its an area-decoration by double-tag, a preselection, non-hierarchic. Better?
Therewith can´t you query, design and render an aera? And its no problem to find the included bounded objects? You have tipps to solve problems? regards, crom

My tip is “go out and do some mapping”.

hi,

How you think to get a p_a into a database? In any case you have lines (two N/E - points) and you have to decorate them with data (we ignore the case, to decorate all objects up to the boundary).
What the hack is the problem? Its the lack of a system to catch all the hundreds of types and on the other hand the database analysis technic. In a database query you can separate with p_a in a first step lots of data.

landuse=conservation
Most don´t grasp the protection “stuff”. Conservation is no leisure, no landuse and its not plane. Its a boundary and its administrativ, “Non-physical”! You will need different changing planes/layers below. With other terms than boundary, you will get into misleading spheres and uses (Timbl agrees: “In general you can not deduce the natural state of land from the fact that it is conservation land”).

For p_a there are lots of concers, but realy no constructiv postings. Its a mess!
http://lists.openstreetmap.org/pipermail/tagging/2011-March/007008.html
People using this scheme and they come along with this “complex” proposal. The page and usage seems not too difficult or challenging.

Sure, there might be some problems. On my view, they are more in using relations and - the fact of non-rendering. Therefor the formus are. And those, who should know about OSM-render, don´t give any constructiv input. Only “go back to school”.

e.g.:
Ghana-site: Muni Lagoon 14/08/92 Central 8,670 ha 05º22 N 000º40 W:
boundary=protected_area
protection_title=Ramsar site
protect_class=98
IBA:ref=GH034
IBA:ref:source=Important Bird Areas, *
IBA:ref:website=http://www.birdlife.org/action/change/ramsar/ibas_ramsar_africa/Ghana.pdf
IBA:ref:operator=*
name=Muni Lagoon

or one term less:
landuse=conservation
ramsar=yes

(“yes” is no value …)

You can´t show all about 20 p_a-types on the map There should be prior reserves to render (first seven). Others, like the globals, can become activated by checkbox.

thoughts:

  • the “ref”-use above is ok. like this?
  • I need a capable procedure to tag zones (I, II, …) of p_a´s (with relation an child/parent?). Thats an other thread.
  • Its better to use existing ways - if existing - or draw for p_a´s better own once? (I don´t like this bundles of lines. But it might be a decision from case to case).

thanks for constructiv responses.

regards, crom

btw. & for sure: Potlatch is nice, thanks Richard!
in the edit-mode, sometimes I noticed situations, that an aera covers other info (nodes, ways).