<html>
<head>
<title></title>
<meta content="MSHTML 6.00.6000.16850" name="GENERATOR"></meta>
</head>
<body>Hallo
<div align="left"> </div>
<div align="left">I have a thought about indexing. Idon't know much about them but I have read some headings from the diskussions about deviding features to get more out of the geometries.</div>
<div align="left"> </div>
<div align="left">I guess the issues about comparing big to small will be more common when handling "the whole world".</div>
<div align="left">Wouldn't it be possible to just devide the index. </div>
<div align="left"> </div>
<div align="left">The bounding-boxes would be just like the divided features bboxes but just be indexes.</div>
<div align="left"> </div>
<div align="left">I think one example to discuss this from is a query asking for all towns close to the atlantic oscean. With a boundingbox to the atlantic the index will give us most of the towns in the world. </div>
<div align="left"> </div>
<div align="left">The idea is to build the index like, lets say, for example one bounding box to the representation of the atlantic in each long/lat degree.</div>
<div align="left">Then the index could fist scan all towns sharing "lat/long-degree boxes" with the atlantic from some static ID of the boxes, then select from them using the smaller boundingboxes to the atlantic. </div>
<div align="left"> </div>
<div align="left">Or maybe it would be better to divide the indexes in relation to the size of the object. I don't know</div>
<div align="left"> </div>
<div align="left">I say like Regina:</div>
<div align="left">Crazy thought so feel free to knock it down.<br />
</div>
<div align="left">or maybe it's alreagy out there.<br />
</div>
<div align="left"> </div>
<div align="left">/Nicklas</div>
<div align="left"> </div>
<div align="left"><br />
2009-06-18 Paul Ramsey wrote:<br />
<br />
Folks,<br />
><br />
>I may be getting a contract to bring in geodetic support in the next<br />
>little while. The challenge is to bring in this potentially big piece<br />
>of new functionality, and have it into a released version in around 9<br />
>months. I have been thinking about different ways to approach the<br />
>implementation and placed some thoughts here.<br />
><br />
>http://files.getdropbox.com/u/1184727/geodetic-plan.txt<br />
><br />
>I'm interested to hear your thoughts about what the most useful<br />
>approach would be. The client wants to see this work in the core<br />
>distribution, and OpenGeo has an interest in the longer term vision of<br />
>PostGIS, so the most expedient approach is not necessarily required,<br />
>we can choose something that balances our long-term vision with the<br />
>short-term need to have working code in a numbered release.<br />
><br />
>Yours,<br />
><br />
>Paul<br />
>_______________________________________________<br />
>postgis-devel mailing list<br />
>postgis-devel@postgis.refractions.net<br />
>http://postgis.refractions.net/mailman/listinfo/postgis-devel<br />
><br />
></div>
</body>
</html>