[GRASS-dev] r.series: extentions as plugins

[Markus Neteler]

On Thu, May 21, 2009 at 11:26 PM, Helena Mitasova
<hmitaso at unity.ncsu.edu> wrote:
> On May 19, 2009, at 6:08 PM, Markus Neteler wrote:
>> would it make sense to add support for extentions as plugins in
>> r.series? We plan to add some function fitting which might not be
>> of interest to all - on the other hand a plugin style for extensions
>> could simplify the addition of custom functionality.
>
> Markus,
>
> it would be hugely appreciated if you could add any planned capabilities
> to the core r.series rather than plugins -
>  function fitting would be certainly of great interest.

Helena,

the point wasn't to keep the plugin(s) for me but to facilitate the
development of extensions. That every group interested in an extensions
hack the core r.series code might destabilize everything.
Having it separated out as plugin separates core from user functionality.

> We were able to get lots of very interesting, useful and unique results
> when applying the module to time series of lidar data and we have been
> thinking
> about adding more options too. We got around the variable time step issue
> using inserted null rasters (this should probably be included on the man
> page)

(yes, please add or send me a text snippet)

> but providing the time steps as an option may be a better approach in case
> there are big differences - you may have this issue with function fitting as
> well unless you have constant time step. Maybe we should keep in touch
> about r.series to coordinate the development.

Perfect.

Markus