Comments on: Strange behavior by design of the spatial function Filter in SQL Server 2008 http://blog.jordogskog.no/2010/02/27/strange-behavior-by-design-of-the-spatial-function-filter-in-sql-server-2008/ Open Source (mostly) GIS in general and PostGIS in particular Sat, 19 Feb 2011 22:34:03 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: Nicklas Avén http://blog.jordogskog.no/2010/02/27/strange-behavior-by-design-of-the-spatial-function-filter-in-sql-server-2008/comment-page-1/#comment-6394 Nicklas Avén Sat, 19 Feb 2011 22:34:03 +0000 http://blog.jordogskog.no/?p=78#comment-6394 the point here is not the false positive answers from the filter function. The point is that the filter function in some situations give different answer in the select clause and the where clause depending on that the where part gets an answer index-based answer and the select part does not. /Nicklas the point here is not the false positive answers from the filter function. The point is that the filter function in some situations give different answer in the select clause and the where clause depending on that the where part gets an answer index-based answer and the select part does not.
/Nicklas

]]> By: David http://blog.jordogskog.no/2010/02/27/strange-behavior-by-design-of-the-spatial-function-filter-in-sql-server-2008/comment-page-1/#comment-6308 David Wed, 16 Feb 2011 17:23:57 +0000 http://blog.jordogskog.no/?p=78#comment-6308 You are correct, it may return false positives. However, the intended use is to get a smaller chunk of data out of a huge database really really fast. And THEN you use more precise filtering to narrow that smaller piece of data down to your final correct answer. If you were to only use the precise filters first, it would take far longer than this two step process. In a sense, it's like looking something up in a [real, paper book] encyclopedia - you would grab the book with generally what you want in it, like articles starting with En-Ez, and then look in the index in the back of that book to find exactly the article you want. You are correct, it may return false positives. However, the intended use is to get a smaller chunk of data out of a huge database really really fast. And THEN you use more precise filtering to narrow that smaller piece of data down to your final correct answer. If you were to only use the precise filters first, it would take far longer than this two step process.

In a sense, it’s like looking something up in a [real, paper book] encyclopedia – you would grab the book with generally what you want in it, like articles starting with En-Ez, and then look in the index in the back of that book to find exactly the article you want.

]]> By: Isaac Kunen http://blog.jordogskog.no/2010/02/27/strange-behavior-by-design-of-the-spatial-function-filter-in-sql-server-2008/comment-page-1/#comment-101 Isaac Kunen Fri, 05 Mar 2010 07:02:07 +0000 http://blog.jordogskog.no/?p=78#comment-101 Hi Nicklas, Well, the trackback didn't work (I assume user error on my end) but I've posted a response here: http://blogs.msdn.com/isaac/archive/2010/03/04/filter-one-odd-duck.aspx The short version: yes it's odd. But it's a good thing. Cheers, -Isaac Hi Nicklas,

Well, the trackback didn’t work (I assume user error on my end) but I’ve posted a response here: http://blogs.msdn.com/isaac/archive/2010/03/04/filter-one-odd-duck.aspx

The short version: yes it’s odd. But it’s a good thing.

Cheers,
-Isaac

]]>