This discussion is archived
8 Replies Latest reply: Jan 10, 2013 1:51 PM by B Hall RSS

Re: GeoRaptor 3.2.1 Released for SQL Developer 3.x

FlyingGuy Explorer
Currently Being Moderated
Hey,
Referencing this that was posted in early September 2012:
GeoRaptor 3.2.1 Released for SQL Developer 3.x

I am supporting GRE's and the like at UC Berkeley and I really want to hook them up with this plug because it will make their life MUCH easier. But I have some pretty basic problems:

1. A VERY simply table that has an 8307 linestring. The query that GR defaults when the table is added :
SELECT rowid,"ID","NETWORK_ID","BEG_NODE_ID","END_NODE_ID","LENGTH","SPEED_LIMIT","DETAIL_LEVEL",
"IN_SYNCH","CREATED","MODIFIED","CREATED_BY","MODIFIED_BY","MODSTAMP","NET_GEN_ID",t.GEOM as GEOM 
  FROM VIA.LINKS_HOLD t 
 WHERE SDO_FILTER(t.GEOM,?,?) = 'TRUE'
This seems to do something but nothing is ever displayed. The table definition is as follows:
CREATE TABLE LINKS_HOLD  (
ID NUMBER, 
NETWORK_ID NUMBER, 
BEG_NODE_ID NUMBER, 
END_NODE_ID NUMBER, 
GEOM MDSYS.SDO_GEOMETRY , 
LENGTH NUMBER(9,3), 
SPEED_LIMIT NUMBER(3,0), 
DETAIL_LEVEL NUMBER(3,0), 
IN_SYNCH NUMBER(1,0), 
CREATED DATE, 
MODIFIED DATE, 
CREATED_BY VARCHAR2(30 BYTE), 
MODIFIED_BY VARCHAR2(30 BYTE), 
MODSTAMP TIMESTAMP (6), 
NET_GEN_ID NUMBER
)
And some sample data. I am only including the geometry
GEOM(SDO_GTYPE, SDO_SRID, SDO_POINT(X, Y, Z), SDO_ELEM_INFO, SDO_ORDINATES)
--------------------------------------------------------------------------------
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18477, 34.02929, -118.18156, 34.02932))
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18268, 34.02393, -118.18269, 34.02631))
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18269, 34.02631, -118.18268, 34.02393))
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18257, 34.02731, -118.18387, 34.0273))
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18387, 34.0273, -118.18257, 34.02731))
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18256, 34.02631, -118.18269, 34.02631))
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18269, 34.02631, -118.18256, 34.02631))
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18131, 34.02632, -118.18256, 34.02631))
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18256, 34.02631, -118.18131, 34.02632))
SDO_GEOMETRY(2002, 8307, NULL, SDO_ELEM_INFO_ARRAY(1, 2, 1), SDO_ORDINATE_ARRAY(-118.18257, 34.02731, -118.18256, 34.02631))
Since the query I tried to substitute for the system generated one was:
select geom from links_hold ;
When I tried to run that the system responded with and error and pasted the following into the clipboard:
SELECT rowid, GEOM as GEOM FROM ( 
select
rowid,geom
from
via.links_hold
) 
? MDSYS.SDO_GEOMETRY(2003,8307,NULL,MDSYS.SDO_ELEM_INFO_ARRAY(1,1003,3),MDSYS.SDO_ORDINATE_ARRAY(-180.0,-165.78279266572636, 180.0,165.78279266572636))
? 
Please keep in mind that very same query will display all the links where I run in the generic map view (GMV) of SQLD and
since GMV has little in the way of additional features it is rather frustrating to see your tool not really working, or what
is more likely is that I have a lack of understanding because I cannot find documentation that will allow me to
move forward w/o bugging you.

So Any light you can throw on this would be of great assistance. Thanks much!
  • 1. Re: GeoRaptor 3.2.1 Released for SQL Developer 3.x
    Simon Greener Journeyer
    Currently Being Moderated
    Flying Guy,

    The non-display is a major bug that was discovered after GeoRaptor 3.2 was released. I knew it was there a few days after
    the release - it got past our testers as well. I hoped to have a release out very soon after but a bunch of things conspired to
    stop this from happening. Anyway, no one complained so I figured no one was using it enough to find the bug.

    The bug is fixed which you can test if you download the Early Access release from our SourceForge website.

    There are some other new functions in this EA that you might discover. The main one being the ability to have a feature count
    (those drawn within the map window) displayed on the view / layer tree. See GeoRaptor preferences. Another is the restoration
    of right mouse click menus on the sdo_geometry columns of table/mv/view connection nodes. Also, if a table has more than one
    sdo_geometry column, the first is no longer selected if you right mouse click/map etc on table node. It will display a pick list
    of sdo_geometry columns for you to chose from. (There has also been a major revamp of connection management as I noted
    problems since SQL Developer 3.2.)

    Please download and try it. Consider joining the GeoRaptor lists to help with testing and documenting. If you find bugs in
    the EA please email the right georaptor list, or use the GeoRaptor SourceForge site to register the bug. Or email me
    directly simon at spatialdbadvisor dot come.

    I hope to have a release ready by the end of the week but I have not heard back from the small group of testers
    that we have something that is ready to go.

    Wrt documentation we have added some where users have created some for me with my input. Please volunteer to
    create documentation.

    The simple fact of life is that:

    1. GeoRaptor is downloaded by lots and lots of people (a good thing).

    But:

    <complain>
    2. Very few responses ever occur to postings made about GeoRaptor on this forum or on our sourceforge forum: disheartening;
    3. Very few bug reports or enhancement requests ever get created at sourceforge: disheartening;
    4. I have become the only developer which is a bad thing. I known of some Java developers who have used GeoRaptor but
    most/all expect someone else to fix bugs or develop the tool. I have spend thousands of hours on a tool that I find useful because SQL Developer
    is my main Oracle access tool. But I have never, ever heard from more than 10 people over the past 3 years that they too find it useful: disheartening;
    5. Attempts to contact downloaders routinely fall on deaf ears or fail: disheartening;
    6. Requests for documenters, developers, testers are ignored or fail because we can't contact anyone who downloads GeoRaptor: disheartening.
    </complain>

    Like a lot of open source projects I do not get paid for anything I do. But what annoys me is that no one - other than a tiny brace of people - will help with anything
    even when we ask.

    I have also decided that major new functionality added to GeoRaptor eg GeoRasters; WMS; WFS; etc will only be done when sufficient funding, via a fundraising website
    makes a set amount. This is something that the PostGIS have started doing and I think it only fair considering how many months each year I spend on development "in
    the dark" for no income.

    regards
    Simon
  • 2. Re: GeoRaptor 3.2.1 Released for SQL Developer 3.x
    FlyingGuy Explorer
    Currently Being Moderated
    OMG!

    You have NO idea. We are moving from PostGre to Oracle. I have people from PhD's to GSR's and undergrads starting to explore Oracle w/Spatial and this tool is a freeking godsend. I will beta test ANYTHING and you can rest assured you will get feedback. On my wish list right now is the ability to package up pre-bundled sets of layers that I can put our git stream and point new people at them and say, go get these files, put them in your .sqldeveloper/georaptor ( or whatever ) then open georaptor and you will have 5 pre-defined networks with all the links, nodes and labels.

    I will download the latest tomorrow.

    Cheers!
  • 3. Re: GeoRaptor 3.2.1 Released for SQL Developer 3.x
    Simon Greener Journeyer
    Currently Being Moderated
    Flying Guy,
    OMG
    Yes, it is that bad.
    You have NO idea. We are moving from PostGre to Oracle.
    Wow, the opposite of every situation I have heard of recently.
    I have people from PhD's to GSR's and undergrads starting to explore Oracle w/Spatial and this tool is a freeking godsend.
    Amazing: others don't let us know that the tool is so useful.

    What is a GSR BTW?
    I will beta test ANYTHING and you can rest assured you will get feedback.
    If you have a SourceForge account, register yourself on the georaptor-devel@lists.sourceforge.net email list.
    On my wish list right now is the ability to package up pre-bundled sets of layers that I can put our git stream and point new people at them and say, go get these files, put them in your .sqldeveloper/georaptor ( or whatever ) then open georaptor and you will have 5 pre-defined networks with all the links, nodes and labels.
    OK, in the new EA (not a previous release) is the ability to export a single layer, selected set of layers or an entire view to an export properties list that can be imported into another SQL Developer/GeoRaptor session.

    So, you will need to create the layers, define the styling etc and then export the layer styling metadata to an XML file. Send it to your colleagues who can import it. As long as the connection name is the same the layer definition will work.
    I will download the latest tomorrow.
    Let me know, either by direct email or via the above georaptor sourceforge list.

    regards
    Simon
  • 4. Re: GeoRaptor 3.2.1 Released for SQL Developer 3.x
    FlyingGuy Explorer
    Currently Being Moderated
    So a GSR is a Graduate Student Researcher.

    PostGre and PostGis are to be respected but even with 9.2 is still does not do partitioning well and does not support the sophisticated partitioning methods of Oracle. Also moving data around is a major pain in PostGre since data files get tied to the size of disk sectors, memory and so many other things they do to get performance. Also we tackle very large real time data feeds which in PostGre put us in an almost perpetual state of vacuum. The need to fix the transaction IDX problem before I will take them seriously for high transaction environments. Oracle needs to make some serious changes to handling spatial queries since a query on a geom is fast fast but if you put an AND condition with it you can be into a table scan before you know it. Also spatial has a bit of a "bolted on feel" that postGis does not. But even with those complaints it is still pretty much the best solution, IMO. And it is a hard sell for people who have only known PG in this arena.

    Looking forward to tomorrow!

    Cheers
  • 5. Re: GeoRaptor 3.2.1 Released for SQL Developer 3.x
    B Hall Explorer
    Currently Being Moderated
    Simon,

    Now that I'm working for a company that uses ESRI software for editing, I'm finding the validate geometry feature very handy to show (graphically) what the issues are. Great work.

    My plan is to provide the data owners with regular "bad geom" reports and show them how to use GeoRaptor to help identify the issues the reports point out. This of course is rather important, as they are using MVs calling features such as sdo_intersection that like to exhaust all session memory when they encounter bad polygons.

    BTW, Is there a way to put a general row limit on the number of objects georaptor tries to render?

    Thanks,

    Bryan
  • 6. Re: GeoRaptor 3.2.1 Released for SQL Developer 3.x
    Simon Greener Journeyer
    Currently Being Moderated
    Bryan,

    Yes I can add this to GeoRaptor.

    Don't forget, though, that a GeoRaptor non-point layer can have pixel filtering turned on which, for small scales for polygon areas etc, can result in large improvements is search and draw times. Accessed via SQL table for Layer Properties dialog.

    The current Early Access release is a version that could be released and I am keen to get it out because of the existing visualisation bugs in the last public release.

    Please add this request to our Sourceforge feature request page.

    I will add this capability to the GeoRaptor today and update the EA. I will then write back to this forum to ask you to test the release. I would prefer all testing comments to come to me directly at my email address (which you know).

    Announce: I have implemented the desired change. A new EA with version number 3.2.2.0793 or above has the change.
    Note 1: Tools>Preferences>GeoRaptor>Miscellaneous>"Search/Display/Query Limit (feats):" is where the limiter is set.
    - *0 means no limit is applied*.
    - Any number >=1 will implement a limit that is applied across all Query and Spatial layers but not Graphic layers.
    - This limiter is NOT layer specific.
    Note 2: Pixel filtering in the SDO_FILTER operator is default for all Query and Spatial layers.
    - This may interfere with how you interpret the displayed results where a feature/query limit is applied.
    Note 3: This EA has in it the ability to display the number of features drawn in a view's map via a preference.
    - The number is prefixed the the layer's visible name in the view node.
    - Tools>Preferences>GeoRaptor>Spatial View>"Display Layer's Feature Count?" is where the display is controlled

    Please let me know how your testing goes.

    regards
    Simon

    Edited by: Simon Greener on Jan 8, 2013 10:02 AM

    Edited by: Simon Greener on Jan 8, 2013 10:35 AM
  • 7. Re: GeoRaptor 3.2.1 Released for SQL Developer 3.x
    Simon Greener Journeyer
    Currently Being Moderated
    Bryan,

    I have been thinking of doing more work on the validate code if I can find time..... and not within the current impending release cycle.

    What I am considering doing is this:

    1. Try and use JTS or custom Java to find all instances of an error not just the first for some error types. Currently validate.. with.. context stops after finding the first error. But for things like 13356 or 13349 it is possible that such errors may occur elsewhere in the geometry.
    2. The draw/zoom feature/error stuff was inherited from the old GeoRaptor code - though I have done a lot of work on it to improve it. I would like to do the following:
    On validation where errors occur:
    * Create two graphic layers named <table_geom>features and <tablegeom>_errors. 
    * Then either:
    A. add all features and errors to these new graphic themes and use the right mouse click menus to find and zoom to each feature/error as required, Or
    B. Add feature/error to graphic layers individually (over writing what was there before).

    Do you have any views/comments on these two possible changes?

    regards
    Simon
  • 8. Re: GeoRaptor 3.2.1 Released for SQL Developer 3.x
    B Hall Explorer
    Currently Being Moderated
    Simon,

    I've downloaded the latest and have been using it today. I'll go add the request on SF you requested.

    As far as multiple errors go - I like the idea, but for 13356 type errors, I think it is unnecessary. From what I have seen so far, generally reducing the density of the data corrects the issue fine. For errors like crossing polygons however, that could be very useful if that just so happened in two or more areas.

    As far as the two layers (normal and error) go - I love the idea. As it is, I am loading the layer in addition to the error to get a reference, which isn't as bad now that you have a row limit. But having it do that for me would go a long way as to make the tool more friendly to my new EXXX customer base.

    Something that would also be most helpful would be to show feature's individual errors while zooming in or out, maybe until a different item is selected in the table. It is rather tedious to have to keep re-selecting that after zooming in for a closer look.

    Thanks again,

    Bryan