MS RFC 30: Support for WMS 1.3.0

[Daniel Morissette]

Kralidis,Tom [Burlington] wrote:
> 
> FYI here's the original ticket behind this:
> http://trac.osgeo.org/mapserver/ticket/473
> 

Thanks. I have added a ref to the ticket in the RFC and vice versa.


> Market behaviour shows us that 1.1.1 is still the most heavily used.
> WMS 1.3.0 came out in 2003, and adoption has been very slim.  It might
> be something "good to have" in the OGC suite within MapServer, though.
> 

I fully agree with you. I'm not a big fan of WMS 1.3.0 either, we're 
doing it mostly for completeness.


> Can the XML output can be done via libxml2?  Note that OGC:WMS 1.3.0
> does NOT support OWS Common (which mapowscommon.c implements), but
> future versions will.  This would position MapServer OGC:WMS support for
> turnkey integration once OGC:WMS 1.4.0 is adopted (N.B.: most OGC specs
> are or will be to be deriving from OWS Common).
> 
> From our libxml2 discussions, one issue was bandwidth; I think the this,
> in OGC:WMS, would not be affected here (i.e. OGC:WFS GetFeature and
> OGC:SOS GetObservation would be the only voluminous XML cases I can
> think of).
> 
> To clarify, I am not suggesting mapwms.c be libxml2'd in whole, just
> 1.3.0.
> 

I'll keep libxml2 in mind during the implementation, but I do not plan 
to refactor and risk breaking any code to convert it to libxml2 as part 
of this upgrade.

Since WMS 1.3.0 doesn't implement OWS common, it won't benefit from any 
of the code that's already using libxml2. It will actually mostly reuse 
existing printf-based code. I think the right time to switch to libxml2 
for WMS would be when it will support OWS common and then there will be 
real benefits by reusing functions from mapowscommon.c.

With respect to 'bandwidth', I don't think the issue that we had raised 
against libxml2 for WFS was directly related to bandwidth: the issue was 
the size of the response and the amount of memory required to generate 
them. However you are correct that size of the responses is not that big 
a deal for WMS and as I wrote above I'll keep libxml2 in mind as I 
implement things but that won't be a priority since we would be mostly 
reusing existing printf-based code that's already well tested and working.

Daniel
-- 
Daniel Morissette
http://www.mapgears.com/