Using WMS time dimension with rasters?

[Jeroen Ticheler]

Hi!
Happy to see this discussion and implementation progressing! This  
(seeing more clients & servers with time support in WMS)  is indeed  
very true. We are working to get temporal data online both as WMS and  
WCS. Having a good support for that in MapServer would be a real  
asset, and actually allow us to use MapServer also for these data sets.

In fact, this also gets to the problem I faced so far while  
implementing this; One common Map file will not do to have both a WMS  
and WCS with time support running, they are slightly different in the  
way they have to be configured (the WCS needs special layer that  
gives an error when accessing the service as WMS, at least that is  
what I ran into when following the different WMS and WCS time examples).
Are there plans to align the two services?

Greetings,
Jeroen

On 26 Oct 2005, at 21:20, Jeff de La Beaujardiere wrote:


>> Jeff Schmaltz wrote:
>> But looking at Appendix B & C of the WMS 1.1.1 spec, it's not  
>> clear to me if this combining of multiple extents and extent types  
>> is legal?
>>
>>
>
> Per Table C.1 of WMS 1.1.1, this is legal:
> 2004-09-21/2004-09-25/P1D, 2004-09-29/2004-10-31/P1D
>
> However, the example you gave with a single value in the middle of a
> set of ranges is neither forbidden nor explicitly permitted:
> 2004-09-21/2004-09-25/P1D, 2004-09-27, 2004-09-29/2004-10-31/P1D
> (sorry about that ambiguity in the spec).
>
> You would be better off making the middle one similar to the others:
> 2004-09-21/2004-09-25/P1D, 2004-09-27/2004-09-27/P1D,  
> 2004-09-29/2004-10-31/P1D
>
> -Jeff DLB
> Editor
>
> P.S. Although time support in WMS clients and servers has  
> traditionally been poor,
> I think we will see more of it as providers with interesting time- 
> dependent
> datasets start putting them on line.
>
>