Changes between Initial Version and Version 1 of Ticket #3712, comment 18


Ignore:
Timestamp:
Feb 22, 2017, 12:23:36 PM (5 years ago)
Author:
Björn Harrtell
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #3712, comment 18

    initial v1  
    55@flippmoke:
    66
    7 I did have a look at wagyu and while my C++ and geometry math skills are too limited to understand it completely it seems to be very well executed. I'm familiar with snap rounding but not the rest of the concepts. I can see how it works with small scale errors that occur because of scaling but I wonder how it can handle severely invalid input, for instance a large bow tie like POLYGON (0 0, 10 0, 10 5, 0 -5, 0 0).
     7I did have a look at wagyu and while my C++ and geometry math skills are too limited to understand it completely it seems to be very well executed. I'm familiar with snap rounding but not the rest of the concepts. I can see how it works with small scale errors that occur because of scaling but I wonder how it can handle severely invalid input, for instance a large bow tie like POLYGON ((0 0, 10 0, 10 5, 0 -5, 0 0)).
    88
    99I'd also love to know more about the algorithm and perhaps help bring it to PostGIS. One way to do that would be to include it in GEOS. As you might know the upstream of GEOS is JTS and I think there would be the appropriate place to get it implemented first. JTS/GEOS already has snap rounding. I'll bring up the subject on the JTS mailing list and hopefully get some of their excellent geometry math brainpower involved.