Re: Geometry / geography woes
"nikolai.berkoff via postgis-users" <[email protected]> Wed, 10 Sep 2025 13:39:15 +0000
| Newsgroups | gmane.comp.gis.postgis |
|---|---|
| Message-ID | <yCWJTAmliMs0rAIIXWmBYf8G5MBRTFYsoH1GSOYd70Bq153r4tZPPP6Yc1FuUvEy6vGs9joSMVQI65oa18Zd9bvrTCvn_rKsPWoY9ib8jTk=@pm.me> |
Thank you, A local coordinate system would not work for us as we are calling ST_Subdivide on the ranges of all species in the IUCN Redlist, so it is very much global problem. I believe ST_Segmentize would solve this for us. Would you consider a documentation update to ST_Subdivide and the the FAQ warning about this? For the FAQ something like: NOTE: If you plan to re-project geometries after using ST_Subdivide(), be aware that the subdivision process may introduce additional vertices. These new vertices do not affect the geometry in the original coordinate system, but they can alter its shape when re-projected — especially if the original vertices span large distances. To mitigate this, consider applying ST_Segmentize() to your geometries before subdivision, ensuring that vertex spacing is appropriate for the target projection or geography-based calculations. and putting the example Regina gave in https://postgis.net/docs/ST_Subdivide.html ? I would happily make a PR.
signature.asc
(application/pgp-signature, 603 B)
-----BEGIN PGP SIGNATURE----- Version: ProtonMail wsC5BAEBCgBtBYJowX91CZCP0vdACs4dP0UUAAAAAAAcACBzYWx0QG5vdGF0 aW9ucy5vcGVucGdwanMub3JndV9NObFmgEK6qsX0n+GIvDKlH83OVlcvreo6 cdhK8A8WIQTRiXkt+tp5DMsBczKP0vdACs4dPwAA7/gIAJjA+eAb6Tlv0GyD CWKdsj8CoC5R2guW1GpsHfebY89N7lLZQorQs3W3Arv0482u90ptK5C9KAtO hcnibOX+rUgKUg9iDOnyf5wqILgSn72pjge8OFC2i4WAVo013MwKDgea8jWl +g+mMetqTG9X+WQA4IwbUT8pH0HxXDbsd/Mm0T43Rdzm//MZEUCwXn035I/8 kXUicbxwoYNfCRfLmUIB4Gtie13USToV9jmFjr4i85f/a0eNfnl30M+nR9iR LcRhxB7uvLeq+b9j05vOmkKQZIays0BDgzsOirRj4u5MsCayQhOPfSsFNXMd yCvcscBiprK9MIu94MC7fSmHblo= =EPdl -----END PGP SIGNATURE-----