22 Members
Geometry Engine Open Source - https://geos.osgeo.org1 Servers

Load older messages

Timestamp Message
13 Nov 2019
20:43:43@freenode_darkblue_b:matrix.orgdarkblue_bI want to understand how MakeValid relates to the practices of Four-Letter-Company, OpenStreetmap and others
20:44:01@freenode_darkblue_b:matrix.orgdarkblue_b.. regarding 2D building footprints in massive online mapping
20:44:43@freenode_darkblue_b:matrix.orgdarkblue_bSandro wants me to do work and not talk so much, so I wont ask there right now :-)
20:44:51@freenode_darkblue_b:matrix.orgdarkblue_bhe is right
20:45:16@_slack_osgeo_UJM07FH5H:matrix.orgMartin Davis pretty sure MakeValid just nodes and polygonizes all the lines, and then keeps only the polygons which form a valid MultiPolygon
20:45:32@freenode_darkblue_b:matrix.orgdarkblue_bbasically that is what I think too
20:45:47@freenode_darkblue_b:matrix.orgdarkblue_bbut there is a lot of detail in that ..
20:45:59@freenode_darkblue_b:matrix.orgdarkblue_bit hasnt failed yet on these test cases
20:46:40@_slack_osgeo_UJM07FH5H:matrix.orgMartin Davis I would hope not! If it does then time for some bug-fixing...
20:50:44@freenode_darkblue_b:matrix.orgdarkblue_bok - I think the colon in the info box means .. it occurs between vertice N and M
20:51:30@freenode_darkblue_b:matrix.orgdarkblue_bso .. Ring-CCW Vert[0:1 0] POINT( x y )
20:51:36@_slack_osgeo_UJM07FH5H:matrix.orgMartin Davis .. "on segment N"
20:51:53@freenode_darkblue_b:matrix.orgdarkblue_b ok yes "on segment N"
20:52:21@freenode_darkblue_b:matrix.orgdarkblue_bmaybe that was easy.. didnt occur to me until now
20:53:21@_slack_osgeo_UJM07FH5H:matrix.orgMartin Davis
[]  POLYGON(14)
Ring-CW  Seg[1 0-1]  LINESTRING ( 160 210, 220 210 ) Len: 60.0
20:53:28@_slack_osgeo_UJM07FH5H:matrix.orgMartin Davis If you hover over an actual segment it shows
20:53:51@freenode_darkblue_b:matrix.orgdarkblue_bdid you get a MackeValid() one too ?
20:54:03@freenode_darkblue_b:matrix.orgdarkblue_bGEOS 3.8 of course
20:54:20@_slack_osgeo_UJM07FH5H:matrix.orgMartin Davis not yet... I actually don't have 3.8 running yet
20:54:27@freenode_darkblue_b:matrix.orgdarkblue_bMULTIPOLYGON(((200 260,270 210,196 66,100 190,200 260),(200 210,160 210,160 170,220 170,220 210,200 210)),((200 210,212 185,188 185,200 210)))
20:54:35@freenode_darkblue_b:matrix.orgdarkblue_bthere you go :D
20:54:57@freenode_darkblue_b:matrix.orgdarkblue_byes! thats why hard-headed berkeley guys are useful.. hardly anyone is testing this
20:54:57@_slack_osgeo_UJM07FH5H:matrix.orgMartin Davis thx
20:56:29@_slack_osgeo_UJM07FH5H:matrix.orgMartin Davis actually the colon separates element and ring indexes. So Ring-CCW Vert[0:1 0] means vertex 0 of ring 1 (a hole) in element 0 (a Polygon)
20:57:09@freenode_darkblue_b:matrix.orgdarkblue_boh ah hm
20:58:09@_slack_osgeo_UJM07FH5H:matrix.orgMartin Davis a bit cryptic I guess... could prob. have a bit more text in that descriptor
20:58:49@freenode_darkblue_b:matrix.orgdarkblue_bI did not figure it out, after looking thrice
22:46:46@freenode_robe249:matrix.org@freenode_robe249:matrix.org left the room.

There are no newer messages yet.

Back to Room List