Yes, once we get to disconnected objects, like RowSets, it all gets
murky. After all a RowSet would not need to come from a database either.
It could be a flat file, or a spreadsheet, or created dynamically like a
collection. But most often it would be a DBMS.
My real question was where you were going to add these to the framework.
As a toplevel package, like "org.apache.stuts.pager"?
I think in the pre-1.0 era, we might have put this into the util
package, but now that we're moving all that off the commons, I wasn't
sure where this would fit in.
The RowSet taglib also has since nice formatting features for basic
column retrieval. I just keep wondering if there's a way to put all this
together into something that could standalone, either in contrib or the
commons.
After all, the first thing I'm going to want to do is page through a
RowSet ... and those dbTag guys would probably like to page through a
result set as well. So the pager package would seem to have at least
sequential cohesion with the dbTags.
Oleg V Alexeev wrote:
> Yes, Ted, I agree with your about disconnected tags and
> database-centric tags. But PagerTag can be used not only for database
> results listings. It work for any environment. To use it developer
> must set appropriate parameters and use PagerIterator to get links
> sequence.
