That sounds great Jesse. Looks like you guys have already given this a lot
of thought. I'm more than happy to get involved and contribute where I can.
Thanks for all the input.

--Erik

On 10/24/05, Jesse Kuhnert <[EMAIL PROTECTED]> wrote:
>
> You should probably just go download the source :)
>
> The part component that used to exist did almost exactly what you are
> describing with the Region component.
>
> I've changed it since to use some of the new functionality possible with
> the
> recent refactoryings of having most of the tapestry logic broken up into
> different hivemind services that you can interact with. I have an
> enhancement worker contribution that actually subclasses all component (on
> classes that make sense) renderComponent methods to enable capturing their
> output without having to have Part/Region components getting involved.
>
> I think every example in the demo has had all of it's javascript
> eliminated
> now except perhaps for the ProgressBar component. JS includes have been
> figured out along with everything else..
>
> I you'll find that switching over to tacos will save you a lot of time.
> All
> of the hard work has already been done and these sort of core concepts
> figured out. All that is left to do now is add cooler/better components
> and
> more functionality, which has become a lot easier as more and more
> functionality has been added to the core.
>
> The only javascript that needs to be included by users is the dojotoolkit
> package, which has been done on purpose. I could have included it into the
> core as well, eliminating any need for js package inclusion, but it really
> goes against what the dojo guys seemed to be wanting to do, so I didn't...
>
> It would still be nice to have a "default" dojo profile included with
> tacos,
> but until I have time to submit a patch to tapestry for somehow possibly
> overriding private component assets it will probably be the way it is now.
> (It does come with dojo anyways after all, you just have to copy and paste
> into your javascript directory, which I'm hoping won't be putting too many
> people out..)
>
> Come help :) We can always use more contributors....
>
> jesse
>
> On 10/24/05, Erik Dreyer <[EMAIL PROTECTED]> wrote:
> >
> > I'm curious to see how they solved the problem of having to capture the
> > output of a specified set of components. In my implementation the Region
> > component had some special logic in it's renderComponent() method to
> > capture
> > it's body. But being able to specify any old component without the need
> > for
> > Region is powerful. It more easily allows existing apps to be
> retrofitted
> > with this functionality. I'm also curious to see how their
> implementation
> > deals with the required client-side JS. One of the features I tried to
> add
> > to mine was not having to write any client-side JS.
> >
> >
> >
> > On 10/24/05, Jesse Kuhnert <[EMAIL PROTECTED]> wrote:
> > >
> > > Yes, I think the region component was called "Part" in tacos before
> the
> > > library got re-vamped.
> > >
> > > What would be even more interesting would be if someone were able to
> > > figure
> > > out a way to do a real swing-like graphical update. Ie no special
> > > components/or having to specify which components should be updated. If
> > it
> > > could figure what has changed, or may render differently than it did
> > > before
> > > on a request and automagically repaint those components that would be
> a
> > > really cool feature. :) It's something I've wanted to do but it
> remains
> > > one
> > > of those pie-in-the-sky when you have time sort of features that
> aren't
> > > important enough right now to put everything else on hold...
> > >
> > > Maybe you should come check out tacos and contribute whatever you are
> > > doing
> > > there?
> > >
> > > On 10/24/05, Markus Joschko <[EMAIL PROTECTED]> wrote:
> > > >
> > > > hi erik,
> > > > I have a deja-vu. Have you looked at the tacos library? It also does
> > > > this (and a lot more), but it is not restricted to regions but can
> > > > repaint every component seperately.
> > > >
> > > > markus
> > > >
> > > > On 10/24/05, Erik Dreyer <[EMAIL PROTECTED]> wrote:
> > > > > Hi Phil,
> > > > >
> > > > > I'm working on a small set of components that do exactly that.
> > > > Basically,
> > > > > like Swing, you have something that creates an event (link,
> button,
> > > etc)
> > > > and
> > > > > a registered listener for that event. You also have a list of
> > Regions
> > > > (new
> > > > > component) that will repaint based on the results of that event.
> > > Regions
> > > > are
> > > > > special components that capture their body (innerHTML) for
> > > transmission
> > > > back
> > > > > to the client where a simple script sets the innerHTML of the
> > > registered
> > > > > Region elements.
> > > > >
> > > > > It is very close to ready. I'll post them on this list soon.
> > > > >
> > > > > --Erik
> > > > >
> > > > > On 10/24/05, Phil Ulrich <[EMAIL PROTECTED]> wrote:
> > > > > >
> > > > > > Is there any way to force a single component to re-render
> itself?
> > > > > > Right now I have a DirectLink that, when clicked, changes the
> > source
> > > > > > binding of a Table component. (Well, actually, it changes the
> > > contents
> > > > > > of the object that the source is bound to.) However, I've tried
> a
> > > > > > number of different methods, but it doesn't seem like after that
> > > step
> > > > > > anything that I call will force the Table to render itself with
> > the
> > > > > > new source contents. Does anyone know a way of doing this?
> > > > > >
> > > > > > Thanks,
> > > > > > Phil
> > > > > >
> > > > > >
> > > ---------------------------------------------------------------------
> > > > > > To unsubscribe, e-mail:
> > [EMAIL PROTECTED]
> > > > > > For additional commands, e-mail:
> > > [EMAIL PROTECTED]
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> ---------------------------------------------------------------------
> > > > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > > > For additional commands, e-mail:
> [EMAIL PROTECTED]
> > > >
> > > >
> > >
> > >
> >
> >
>
>

Reply via email to