[Geotools-devel] Parameters & ease of use

[Jody G. <jga...@re...>]

Martin Desruisseaux wrote:
> Jody Garnett a =E9crit :
>=20
>> parameter   ------synchornized----   difficult to use, but consist=
ent
>=20
> I noticed too that more convenience methods may be needed in parame=
ters.=20
> The hard thing is to find the right methods. For example I'm pretty=
 much=20
> against using ParameterDescriptor as keys, because it would not wor=
k=20
> with different implementations (even inside Geotools, e.g. EPSG dat=
abase=20
> and various MathTransforms have different ParameterDescriptor for t=
he=20
> same "False northing" parameter: alias, identifiers, remarks and ra=
nge=20
> of valid values may be different, but it still the same parameter a=
s far=20
> as projections are concerned). ParameterDescriptor is metadata abou=
t a=20
> parameter, and the key is the name.
>=20
> I'm interrested in learning which kind of difficulties you encounte=
r, so=20
> we can try to improve on that.

Hard to describe - the the ParameterDescriptors *do* in fact describe=
=20
the set of valid ParameterValues. When I tried to create a user so th=
at=20
end users could construct a valid set of Parameter Values it got=20
complicated when considering ParameterGroupDescriptors.

Since that time I think we have refined the interface so that=20
ParameterGroupDescriptors are always 0..* - so perhaps it is easier n=
ow.

Basically we have two extremes - DataStore PARAM is a simple paramete=
r=20
system based on Bean properties and the needs of GeoServer.=20
ParametersValue is a complicated one based on a specification (aka a=
=20
data model).

What I would like is something in between - where we have the=20
completness, while retaining an obvious way to apply it when making
user interfaces.

Perhaps I should try making a user interface out of it again based on=
=20
the last 6 months of experience.

Jody