I'm using Tiles2 so I configured the TilesResult for my views.
To set up the lists for dropdowns in those tiles, I am coding my own Result,
rather than relying on a weakly related Action to do it.
My result has to extend TilesResult, but it seems to be an unnecessary
dependency, because really my Result has little to do with Tiles.
Wouldn't it be beneficial if the Results could be chained together like the
interceptors?
My requirements specify a mechanism that allows for:
- caching of lists (countries again in ApplicationScope for example)
- parameterization (e.g. a list of codes allowed in a particular category -
requires the categoryId)
- localization of dropdown beans (i.e. country names)
- naming the Tile
So I would define my Results globally specifying the Tile (or other view type),
and optionally the caching, the parameter(s) and the localization, and then use
it wherever needed.
This would tie up the view coding with the view itself, and overcome the problem
with validation failure returning flow to the view when the required dropdown
lists are not present.
Does that make sense for future development?
Regards
Adam
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
