tapestry-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Howard Lewis Ship <hls...@gmail.com>
Subject Re: Let's fix Environment state
Date Fri, 04 Jan 2013 17:48:23 GMT
I don't see it as broken. Why do you feel it is broken. Use cases, not
broad pronouncements, please.

What drove this is the Form component; in an Ajax form submission, the Form
puts the FormSupport object into the environment.  Then as a consequence of
the Ajax render, it needs to partially render again ... my memory of this
is faulty. I can't remember what drove the need to cloak  I think it was
there were multiple unexpected things in the Environment, interacting
poorly and the cloak approach was the clean way to approach the problem.

On Fri, Jan 4, 2013 at 3:49 AM, Denis Stepanov <denis.stepanov@gmail.com>wrote:

> I like the concept of Environment, but Howard did broke it by introducing
> cloak/decloak. You would think you could just add new environmental service
> in the filter or somewhere else and assume it would be available everywhere
> during rendering of the request, but it is not true, cloaking will hide all
> your services without warning, and you end up spending your time trying to
> figure out why did it disappear.
> The cloaking is only used in the AjaxPartialResponseRenderer, I assume it
> is there to simulate "clean" request when rendering multiple ajax responses
> in the one request.
> I would say it is possible to replace the cloak/decloak functionality with
> similiar save/restore functionality: instead of hiding the stack before
> rendering a new response it would be better to restore the state of the
> environment stack.
> What do you think?
> Denis

Howard M. Lewis Ship

Creator of Apache Tapestry

The source for Tapestry training, mentoring and support. Contact me to
learn how I can get you up and productive in Tapestry fast!

(971) 678-5210

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message