[fdo-internals] RE: Motion: Vote on FDO RFC 38 - SHP multi-polygons

[Traian Stanev]

+1 Traian


From: fdo-internals-bounces at lists.osgeo.org [mailto:fdo-internals-bounces at lists.osgeo.org] On Behalf Of Orest Halustchak
Sent: Monday, September 07, 2009 9:33 PM
To: FDO Internals Mail List
Subject: [fdo-internals] Motion: Vote on FDO RFC 38 - SHP multi-polygons

Hi,

There were no further comments on this RFC: http://trac.osgeo.org/fdo/wiki/FDORfc38 - Multi-polygon support for SHP.

I would like to motion a vote to accept this RFC.

Thanks,
Orest.

From: fdo-internals-bounces at lists.osgeo.org [mailto:fdo-internals-bounces at lists.osgeo.org] On Behalf Of Orest Halustchak
Sent: Thursday, August 27, 2009 4:50 PM
To: FDO Internals Mail List
Subject: [fdo-internals] Back to FDO RFC 38 - SHP multi-polygons

Hi,

Regarding: http://trac.osgeo.org/fdo/wiki/FDORfc38 - Multi-polygon support for SHP.


In our last discussions on this RFC, there was concern about performance impact.



Dan has done some prototype performance testing to determine the impact of this proposed change. He used some datasets that included complex multi-polygons including Norway data and Australia data that had detailed multi-polygons for all the islands (100's of loops for each feature). Following are some statistics with before and after timings. With each of these datasets, the time to render on the screen was much longer than these load times and there is no difference that an end user would see in these cases.



Australia: 9 multipolygon features with 3 of them having between 800 and 900 polygon loops each, others averaging about 100 loops each.

      .015 seconds to .125 seconds

Europe: 54 multipolygon features with Norway having 862 loops, with others of 653, 509, 359 loops, and the rest averaging about 100 loops each.

      .016 seconds to .17 seconds

Queensland parcels: 2,000,000 polygons

      Unchanged at 1.6 seconds



So, yes it takes longer to process, but the overall timings are still sub-second for the complex cases. There is also no impact on cases of regular polygons.


Dan will be back from his vacation soon and will be able to do the implementation. So, I'd like to get to a vote on this if nobody has objections.

Thanks,
Orest.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/fdo-internals/attachments/20090908/5a03ff5d/attachment.html