On 2/2/07, Igor Vaynberg <[EMAIL PROTECTED]> wrote:
exactly what will you improve in core by having joda time? why cant it be a
dependency on the extensions? or why not make it a core module all onto
itself?

The most important reason for that is that we could ship the core
project with a dateconverter that *just works* including applying time
zone differences, and that can be used as a solid basis for a more
extensive date/ time support. I don't feel like writing that against
the JDK's date/ time API, and I doubt anyone else wants to do that,
but I feel we *should* have such a converter. I don't think many
people are aware of this as you typically won't notice timezone issues
quickly when working with dates without caring much for the time, but
it is actually problematic to not take the time zone into account (as
we experienced for Teachscape where we have a server running in NYC
and testers working on the west coast).

Furthermore, it would just be a better basis for other date/time
components, which may or may not be included in core, but would all
depend on the date converter.

Also, considering how many people are using DatePickers (my uneducated
guess is that almost every project uses at least one), I feel we
*should* provide a good alternative for the date picker that might
even be put in the core project. Another great component is the date/
time picker, which I can contribute with Joda time (again either in
the core or extensions project).

Before people get up screaming against considering moving components
from extensions to core, please remember that the original purpose
(and mine still) of wicket-extensions was to have a kind of incubation
project where components we don't feel should be part of core right
now but that we want to support can be put in, but that could be
*considered* to be moved over time if they are good enough and heavily
enough used. Ideally, at any given time components in extensions would
contain components that are not typically used in every project such
as captcha and palette but also components that are still maturing/
under consideration to be moved. DatePicker - even thought the current
one failed that test - falls into the last category in my opinion.

Eelco

Reply via email to