Hibernate Spatial 1.1.1 released

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Hibernate Spatial 1.1.1 released

Karel Maesen
Administrator
Hi list,

Hibernate Spatial 1.1.1 has been released. This is a minor bugfix release. See the release notes at: http://www.hibernatespatial.org/release_notes.html.

This will probably the last release of the HS 1.x version for Hibernate 3.x, unless of course more major bugs are reported. From now I'm going to concentrate on integrating Hibernate Spatial in Hibernate Core 4.x.

You can check the progress on this on github (https://github.com/maesenka/hibernate-core/tree/HHH-6509) and in the hibernate JIRA (https://hibernate.onjira.com/browse/HHH-6509). Note that in the Hibernate JIRA, there is a component 'Spatial' in the Hibernate Core project.

Regards,

Karel


PS: for those who feel like forking my github project. I will regularly rebase my work against the Hibernate Core master branch. This could bite you if you're not careful.




_______________________________________________
hibernatespatial-users mailing list
[hidden email]
http://www.hibernatespatial.org/cgi-bin/mailman/listinfo/hibernatespatial-users
Reply | Threaded
Open this post in threaded view
|

Re: Hibernate Spatial 1.1.1 released

pmgmendes
That's good news! Nice job.

But why the HIBSPA-21 (http://www.hibernatespatial.org/jira/browse/HIBSPA-21) wasn't contemplated for this release? It's a little strange how it still continues marked as an Improvement when it's clearly a bug in the way the SDO_GEOMETRY type is being dealt with. The provided patches aren't enough?

Best,
Mendes
Reply | Threaded
Open this post in threaded view
|

Re: Hibernate Spatial 1.1.1 released

Karel Maesen
Administrator
Hi Mendes,

I considered this an improvement - rather than a bug - because it produces no wrong or incorrect results, only sub-optimal ones. It is rather a case of not supporting a feature of Oracle's spatial types, rather than a mistake. But I guess, this is debatable.

It was not considered because it takes a bit more work, esp. in making sure that I have good tests to ensure consistent behavior. I plan to do this work in the context of the next major release of HS. If it is urgent I can see what I can do.

Regards,

Karel

On 03 Jan 2012, at 23:19, pmendes wrote:

> That's good news! Nice job.
>
> But why the HIBSPA-21
> (http://www.hibernatespatial.org/jira/browse/HIBSPA-21) wasn't contemplated
> for this release? It's a little strange how it still continues marked as an
> Improvement when it's clearly a bug in the way the SDO_GEOMETRY type is
> being dealt with. The provided patches aren't enough?
>
> Best,
> Mendes
>
> --
> View this message in context: http://hibernate-spatial.1140993.n2.nabble.com/Hibernate-Spatial-1-1-1-released-tp7148271p7148442.html
> Sent from the Hibernate Spatial - Users mailing list archive at Nabble.com.
> _______________________________________________
> hibernatespatial-users mailing list
> [hidden email]
> http://www.hibernatespatial.org/cgi-bin/mailman/listinfo/hibernatespatial-users

_______________________________________________
hibernatespatial-users mailing list
[hidden email]
http://www.hibernatespatial.org/cgi-bin/mailman/listinfo/hibernatespatial-users
Reply | Threaded
Open this post in threaded view
|

Hibernate Spatial 1.1.1 released

pmgmendes
> I considered this an improvement - rather than a bug - because it produces no wrong or incorrect results, only sub-optimal ones. It is rather a case of not supporting a feature of Oracle's spatial types, rather than a mistake. But I guess, this is debatable.

If you're talking about "oriented points" - I didn't know this type of point geometry until 5 minutes ago - it makes sense to store the point coordinates in the SDO_ORDINATE_ARRAY. But the actual SDO_ELEM_INFO that is being generate does not describe this kind of point. http://docs.oracle.com/cd/B19306_01/appdev.102/b14255/sdo_objrelschema.htm#i1001898

The Oracle's function ST_isValid() only returns 1 (true) when:
 - for a point only geometry -> SDO_POINT = (12,14) , SDO_ORDINATES = NULL, SDO_ELEM_INFO = NULL;
 - for an oriented point geometry -> SDO_POINT = NULL, SDO_ELEM_INFO = (1,1,1, 3,1,0), SDO_ORDINATES = (12,14, 0.3,0.2)

But most part of the free open source software for geospatial (gdal, geoserver, mapserver, and others) that support Oracle as a datasource does not recognize oriented points.

> It was not considered because it takes a bit more work, esp. in making sure that I have good tests to ensure consistent behavior. I plan to do this work in the context of the next major release of HS. If it is urgent I can see what I can do.

It makes sense, as tests were not provided. It's not urgent. Thanks for asking and for your time in this great project.
_______________________________________________
hibernatespatial-users mailing list
[hidden email]
http://www.hibernatespatial.org/cgi-bin/mailman/listinfo/hibernatespatial-users