[Tagging] Feature Proposal – RFC – natural=peninsula (Was: Feature Proposal – RFC – place=peninsula)

[Warin]

On 19/01/19 10:22, Graeme Fitzpatrick wrote:
>
> On Sat, 19 Jan 2019 at 09:09, Markus <selfishseahorse at gmail.com 
> <mailto:selfishseahorse at gmail.com>> wrote:
>
>     It certainly can be phrased better (this isn't my strong point), but i
>     wanted to make it clear that a peninsula can also be part of a bigger
>     peninsula.
>
>
> OK, how about "A natural=cape can be part of a natural=peninsula, a 
> natural=peninsula can be part of a larger natural=peninsula, but a 
> natural=peninsula cannot be part of a natural=cape"?
>
>     I've updated the proposal accordingly.
>
>
> Good, thanks, but that also raises an awkward (& unanswerable?) 
> question about "Please do not map very large peninsulas like 
> subcontinents as multipolygons as they strain the servers too much and 
> are hard to maintain"?
>
> How big is "too much strain" & who can say it's straining too much?
>
> Is Cape York Peninsula OK, but Italy too big?
>
> How can anybody tell?
>
> BTW I'm in no way complaining or objecting to the idea (I'll be voting 
> for it when it get's there!), it's just the question of the technical 
> limitations that may be involved?
>

A found a guide somewhere that said 300 was a good maximum number of 
members for a multipolygon.
The northern Blue Mountains tree relation was over 600 and climbing. I 
thought that was a bit high and might need attention, particularly as I 
was adding more members.
I split it into 4 IIRC .. and go them all under 200 members each.

I have not looked as yet at the other large tree relations that I have 
in the past edited. In the 'to do list'.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20190119/e1591a3c/attachment.html>