Accessing images using time dimension.

Yewondwossen Assefa assefa at DMSOLUTIONS.CA
Wed May 16 14:21:28 EDT 2007


Dave Weaver wrote:
> On 15/05/07, Yewondwossen Assefa <assefa at dmsolutions.ca> wrote:
>> Dave Weaver wrote:
>>> 1) What am I doing wrong (or not doing at all) ?
>>>
>>    Do you have your imagery displaying when doing a simple draw map ?
>> (either a wms getmap or mapserver cgi). This is to make sure that the
>> tile index and the pg layers are set properly.
> 
> Forgive my ignorance, but I'm not sure what you mean by a simple draw map.
> 
  All I meant was you can use mapserver cgi and do something like :
http://localhost:8080/cgi-bin/mapserv?map=/home/map/fgs/www/htdocs/maps/postgis2.map&mode=map&layers=sat 

  WMS get map is also valid but just trying to keep it simple and see if 
the map draws with your rasters.

> My current attempted access is via this WMS GetMap call:
> 
> http://localhost:8080/cgi-bin/mapserv?map=/home/map/fgs/www/htdocs/maps/postgis2.map&SERVICE=WMS&VERSION=1.1.1&REQUEST=GetMap&amp;amp;LAYERS=world,sat,radar&WIDTH=800&HEIGHT=400&FORMAT=image/png&TRANSPARENT=false&&amp;amp;BBOX=-90,-25,90,85&TIME=2007-05-11T11:00
> 
> How do I simplify it to test the tile index and pg layers as you suggest?
>  

Most of the time I have used postgis layers, I start by importing data 
from a shape file. From the postings you have sent, maybe there is a 
problem with your postgis layer. How about creating your tiles with 
tile4ms (creates a shape file), see if that works in your map file and 
then import the shape file in postgis ?

>>> 2) Am I right in thinking the geometry column can be null and the rows
>>> will still be considered, depending on image_time ?
>>>
>>   I think the geometry needs to be valid, each row of geometry should
>> normally represent the bbox of the image.
> 
> Fair enough. I've tried adding a proper geometry (see my other email in this
> thread) to no noticeable effect.
> 
>>> 3) Is the duplication of the metadata in the 2 layers above necessary?
>>> I've done it this way because that's what's in the example in the "WMS
>>> Time Support" document, but do I really need to repeat all this
>>> information?
>>>
>>   The matadata should only be necessary on the tileindex layer.
> 
> Makes sense. It would be nice if  the example I mentioned above was adjusted
> appropriately then :-)
> 
  I have added a not to the documentation regarding this.

>>> 4) How can I debug this - is there some way I can find out *why* it's
>>> not displaying the raster layer? Is there maybe some way to show which
>>> time periods it's considering?
>>    the wms time in your case ends up setting the filter element on your
>> postgis layer to something like :
>> "(date_trunc('minute',image_time) = '2007-05-11T11:00:00')" .
>>   Assuming that your imagery is displayed without a time request,you can
>> manually set the filter element and see if that is working.
> 
> 
> It seems unusal to me that there is no logged information from mapserver to
> help track down issues like this. I tried adding the "DEBUG ON"  directive
> to my map file and specifed a LOG location, but I still see no log.  I
> rebuilt following a "./configure [....] --with-debug" and still see no log
> file. Most frustrating!  Perhaps I did something wrong there - I will
> investigate further.
> 
>>> 5) Is there some basic HowTo document that explains what I'm trying to
>>> do? (I couldn't find one).
>>>
>>> Thanks a lot
>>> Dave.
>>>
>>
>> --
>> ----------------------------------------------------------------
>> Assefa Yewondwossen
>> Software Analyst
>>
>> Email: assefa at dmsolutions.ca
>> http://www.dmsolutions.ca/
>>
>> Phone: (613) 565-5056 (ext 14)
>> Fax:   (613) 565-0925
>> ----------------------------------------------------------------
>>
>>
>>
> 


-- 
----------------------------------------------------------------
Assefa Yewondwossen
Software Analyst

Email: assefa at dmsolutions.ca
http://www.dmsolutions.ca/

Phone: (613) 565-5056 (ext 14)
Fax:   (613) 565-0925
----------------------------------------------------------------



More information about the mapserver-users mailing list