Qui c’è un topic http://getsatisfaction.com/opengeo/topics/features_across_the_international_dateline a cui ho partecipato
e il mio punto di vista è l’ultimo dei commenti; lo riporto qui per immediatezza, anche se in inglese
Cristoforo Abbattista 19 hours ago
I’m experiencing the problem just now.
I’m convinced that PostGIS doesn’t need to take care of it, because the digitized feature has the correct coordinates.
Only the modules which draw the maps need to take care. That means OL when retrieving features from Feature servers (WFS case) or Geoserver when serving images representing maps (WMS case).
The problem derives from the fact that OL or Geoserver do not manage such case in coherence with the other points they draw on the earth.
What does it means: even for points laying far from the date line, OL and GS make the choice to draw the shortest projected path between the points; in fact they could always choice to draw a line circumnavigating the earth to connect them even they are at the equator, but they don’t! Why, insted, do they choose to do that for points near the date line? I know why, but they need to correct.
Regards and waiting for solution.
Cristoforo Abbattista
PS: indeed it seems that a good applicable workaround is to manage such case as multiline or multipolygon features touching themselves on the date line
Vorrei conoscere il parere della lista e in particolare come si comporta qgis.
Ciao
Cristoforo
PS: Per inciso ArcGIS fa la stessa cosa che fanno tutti; ossia, secondo me, sbaglia.