>* For my import I will update the pointers of edges my own, could you*>* >Sandro point me to*>* specifications of next_*_edge.*
I guess shoulkd better spend time to optimize the actual ISO function rather than think to do own low level function. The controls to do a right ISO topology are many and skipping them to have fast load mean to have a not really topological dataset. Also there is never in the topology table to maintain really the topology. The correctness of topology is all on the ISO functions. Another point is that often 8really often I guess) the user think to have a set topological. But instead it is not so. As example. The only product I know capable to produce a really topological set is arcinfo -workstation version (with coverages) or oracle with topology model. No other product is capable of this. please note che rules of arcgis don't work on coverages so they don't grant toe really topology face-edge-node. So If an user write own function to load the postgis topology tables. And it's dataset is not really topology and it write more fast routine skipping some complex controls to be faster. It will produce simply some tables anemd topolgycal but not really topologicals. And so grows the caos. So who like write a truly topological load system should get and follow the ISO specs. The actually version of postgis topology is write followinfg the ISO specs row by row. Regards, -- ----------------- Andrea Peri . . . . . . . . . qwerty àèìòù -----------------
_______________________________________________ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users
