[postgis-users] PointM, PolyLineM, PolygonM

[Paul Ramsey]

Measure is commonly used to proxy other values into a line, vertex by 
vertex. For example, the "road miles" of a highway system often differ 
from the mathematical milage calculated from the geometries. So in order 
to determine the "location" of events calculated relative to road miles 
(company A has contract to maintain highway from mile A to mile B) 
"measures" are attached to the vertices indicating the road milage at 
each vertex, and events are located on the geometry by "measuring" 
relative to those values.

Since the measures are a vertex-by-vertex measurement, not a 
feature-by-feature one, they cannot be dropped into an attribute table, 
generally speaking.

Paul

strk wrote:
> pramsey wrote:
> 
>>What do people think about supporting measures in general? There are two 
>>possibilities:
>>
>>(a) keep the geometries as they are and have and "measure dependant" 
>>functions assume that the coordinates are (X,Y,M)
>>(b) expand to 4-d standard geometries, (X,Y,Z,M)
>>
>>I lean towards option (a), since most people aren't even using xyz 
>>geometries now, let alone xyzm. A little flag on the geometry could 
>>indicate whether the geometry is xym or xyz.
>>
>>Any then of course we'd need to figure out what useful "measure-smart" 
>>functions would be useful for a LRS implementation in postgis.
> 
> 
> I do not know what are measured geometries used for, what about, in the
> meanwhile, putting the measure out of the geometry ? I really can't imagine
> when would you need a measure since you have associated attributes with
> values and SQL !
> 
> --strk;
> 
> 
> 
>>P.
>>
>>strk wrote:
>>
>>>Hello Jeff,
>>>are you planning to include any kind of support for
>>>shapefile "measured geometries" types in shp2pgsql ?
>>>

-- 
       __
      /
      | Paul Ramsey
      | Refractions Research
      | Email: pramsey at refractions.net
      | Phone: (250) 885-0632
      \_