Thanks Martin,

I agree with both proposals, so I started from customizing the NADCONTransform 
and exploring the epsgs. By indexing I meant the tree structure algorithms and 
I agree with what you have suggested.

Regarding your question about algorithm package - if I remember well it is used 
just by builders and I think that It was as private as possible. I'll look into 
it. I'm currently away - I'll do more when I'll come back home (tomorrow).

Bests,
Jan.
______________________________________________________________
> Od: [EMAIL PROTECTED]
> Komu: Jan Jezek <[EMAIL PROTECTED]>
> CC: Jesse Eichar <[EMAIL PROTECTED]>,  geotools-devel 
> <[email protected]>
> Datum: 05.06.2007 16:16
> Předmět: Re: GSoC report&questions
>
>Hello Jan
>
>I didn't looked deeply at the NADCONTransform code. It is something I
wanted to
>do for a while, but didn't had the time yet. But there is some notes:
>
>If you go on the path of something similar to NADCONTransform, it would
be nice
>if NADCONTransform could be refactored into something generic (not
specific to
>NADCON), maybe with a NADCONTransform subclass if needed - but ideally
the
>parent should should be generic enough that we don't need.
>
>About the way to specify the mapped positions (WKT, Parameter or URL),
maybe we
>need to provides both. Maybe we could begin with URL for the following
reason?
>
>* May be easier to implement (no special ParameterDescriptor).
>* Allow to retrofit NADCONTransform in your framework.
>* Needed anyway for usage with the EPSG database.
>
>The later point seems of special interest to me. The EPSG database
defines
>transforms based on some definition files (mostly gridded data I guess),
with
>filenames declared in the "Coordinate_Operation_Parameter Values" table.
This
>apply to EPSG operation methods 9613, 9614, 9615, 9620, 9634, 9658, 9662,
9663,
>9664, 9665. It would be nice if you could take a look at what EPSG said
about
>those operation methods and design your framework (including the way to
specify
>mapped points) in a compatible manner, if possible.
>
>About the indexing mechanism, I'm not sure what you means. Do you means a
R-tree
>or something similar? If so, is it envisageable to take the
implementation in
>the shapefile plugin, document it if needed, and move the implementation
to the
>referencing module? We can not introduce a dependency to the shapefile
plugin
>since it would be a circular dependency (shapefile already depends on
>referencing), but shapefile could use the implementation in referencing
if we
>move it there.
>
>       Martin
>


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Geotools-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-devel

Reply via email to