Go to content Go to navigation and search

Home

Straws in the Wind Blog Articles


Search

Browse

RSS / Atom

Email me

textpattern

Creative Commons License
All Blog Articles, Data Models and Free Source Code by Simon Greener, The SpatialDB Advisor is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.

The Sad State of SQL Spatial Standards - Take 2

Friday November 30 2007 at 07:23

There is a useful exchange going on over on the Microsoft Forums in relation to the paucity of information on geographic objects (as against planar objects) in the OGC SQL standard: this is in the context of SQL Server 2008 “Katmai” Spatial implementation. This exchange occured over Microsoft’s reading of the OGC standards in respect of the coordinate ordering of data in their geography data type (as against their geometry planar data type). The upshot is that the Microsoft engineers chose to have the coordinates in a geography ordered Latitude/Longitude which is at odds with how Oracle, PostGIS and many other vendors order the ordinates (they use Longitude/Latitude).

In the exchange I note that Google Maps uses Latitude/Longitude and observe that it is not so much the coordinate ordering that is the issue, or even “standards compliance”, rather it is about the functionality of the implementation.

Creative Commons License

post this at del.icio.uspost this at Diggpost this at Technoratipost this at Redditpost this at Farkpost this at Yahoo! my webpost this at Windows Livepost this at Google Bookmarkspost this to Twitter

Comment [2]

I can see this causing some issues in the future:

* Will this mean that where OGIS standards are inadequate, will we see Microsoft fall in line with standards used by other products, or will we see further standards according to Microsoft?

* What is this going to do for those who are looking for interoperability between other products and SQL Server? Who will budge to build the functionality to allow data to be shared between say Oracle and SQL Server?

— Thomas · 30 November 2007, 09:20 · #

> I can see this causing some issues in the future:

All decisions made early in the life of any product, especially where they relate to data structures and organisation, have effects. Let’s hope Microsoft’s decisions on spatial in 2008 are positive.

> * Will this mean that where OGIS standards are inadequate, will we see Microsoft fall in line with standards used by other products, or will we see further standards according to Microsoft?

I assume by “standards used by other products” you mean how others have interpreted the OGC standards or what they have done in their own implementation. As you correctly observed in an email to me:

“However on further reflection, I can see an advantage in the Microsoft implementation – outside of a spatial environment, we trade information in Lat/Long rather than Long/Lat, so for someone who wasn’t a previous spatial database user coming to use the spatial components in SQL server, its relatively simple (ie, you’re not having to think in reverse for your database like we do for Oracle).”

This is quite true: what we are experiencing now in the GIS industry is the destruction of the comfortable walls we have built around our narrow vertical market. Google, Microsoft and Oracle as all taking geospatial to the masses: and those masses talk Latitude/Longitude in a loose way that makes our professional legs shake!

> * What is this going to do for those who are looking for interoperability between other products and SQL Server? Who will budge to build the functionality to allow data to be shared between say Oracle and SQL Server?

Given the GIS industry generally produces end-user software the coordinate ordering, hopefully, will be handled by these products. So, for example, if we translate data from SQL Server to Oracle Locator using FME then it will “know” to change the ordering (2008 beta does this for me). If I have to hand-build a database-to-database translation (ie not using any third-party software just the databases’ technologies) then I will have to remember to switch the ordering. Yes, this is a pain, but as long as I know I am happy.

Finally, I will let your email have the last word:

”... Regardless of what the outcome is, one thing that the OGC should be taking home with them for further consideration is that ambiguous standards definitions are going to do damage to the ideal of having an open standard. I find it somewhat odd that the storage ordering for the co-ord pairs is open for interpretation.”

regards
Simon

Simon · 4 December 2007, 06:38 · #