I have a minor issue with using "xhr" vs "ajax" as ajax > xhr
It is possible to perform ajax without using an XmlHttpRequest. For
instance, in some browsers it is not currently possible to ajax upload a
file. Instead, you must post to a hidden iframe. I haven't looked at the
code but is possible that tapestry uses this method when posting a form to
a zone.
My 2p ;)
On Thursday, 12 January 2012, Paul Stanton <p...@mapshed.com.au> wrote:
> Igor,
>
> The zone wildcard idea is a bit flawed in that this 'should' work without
any zones on the page whatsoever. A zone update is not required for
AjaxResponseRenderer to be valid, therefore relying on a zone being present
is illogical.
>
> While a rare case, what would be wrong with executing an eventlink (or
the like) who's event handler just adds a script command to the
AjaxResponseRenderer?
>
> tml:
>
> <a t:type="eventlink" event="click" xhr="true">click</a>
>
> java:
>
> private void onClick()
> {
> ajaxResponseRenderer.addCallback(new JavaScriptCallback() {
> @Override
> public void run(JavaScriptSupport javascriptSupport) {
> javascriptSupport.addScript("alert('hello');");
> }
> });
> }
>
> As I mention in the Jira issue (which you should all vote for) you could
easily ensure backwards compatibility with some simple logic, ie:
> isXhr = xhrParam || zoneParam != null
>
> The problem is that the whole reliance on ZoneManager (tapestry.js) for
ajax response processing is out dated. Its left over from the initial
architecture that an event was tied to a single zone for request/response
handling.
>
> In my opinion Tapestry.ajaxRequest (or alternative) should be able to
handle anything AjaxResponseRenderer can throw back at it.
>
> If this were fixed, it would also resolve my other pet hate - the
inability to initialise an xhr request (who's handler makes use of
AjaxResponseRenderer) from script without linking to a zone/ZoneManager.
>
> I love tapestry but I hate this feature of it.
>
> Regards, Paul.
>
> On 6/01/2012 9:22 AM, Paul Stanton wrote:
>
> +1 vote for this, I previously raised this last January.
>
> https://issues.apache.org/jira/browse/TAP5-1404
>
> On 6/01/2012 7:37 AM, Igor Drobiazko wrote:
>
> This not a good idea, as we consequently would need to provide
> AjaxActionLink, AjaxPageLink, AjaxForm, etc.
>
> On Thu, Jan 5, 2012 at 9:02 PM, Muhammad Gelbana<m.gelb...@gmail.com
>wrote:
>
> Or maybe a whole new component with a whole different name ? Instead of
> EventLink we can have AjaxEventLink or something..just an idea.
> I bet you will find this page very interesting Lance
> http://tawus.wordpress.com/2011/10/01/tapestry-5-3-new-features-part-2/ It
> demonstrates many different flavors of ajax calls in tapestry.
>
> On Thu, Jan 5, 2012 at 8:09 PM, Dragan Sahpaski
> <dragan.sahpa...@gmail.com>wrote:
>
> On Thu, Jan 5, 2012 at 5:35 PM, Igor Drobiazko<igor.drobia...@gmail.com
>
> wrote:
> Maybe we should reuse zone paramater by passing a special zone id like
> zone="*". There is already a special id ^, which means the first
>
> container
>
> zone.
>
> I like this idea.
> I''m using "^" a lot so at least for me it's a good concept.
>
> Cheers
>
>
> On Thu, Jan 5, 2012 at 2:40 PM, Lance Java<lance.j...@googlemail.com
>
> wrote:
> Correct me if I'm wrong but currently, if you want an eventlink to
>
> use
>
> ajax, you must provide a "zone" parameter. This may have been fine in
>
> early
>
> versions of tapestry 5 but there are now circumstances where we want
>
> an
>
> eventlink to be done via XHR but shouldn't need to specify a zone in
>
> the
>
> tml. These include:
>
> 1. MultiZoneUpdate where the zone(s) are determined in the event
>
> handler
>
> 2. AjaxResponseRenderer where zone updates and javascript execs can
>
> be
>
> invoked in the event handler
>
> As a suggestion, eventlink could have a new boolean parameter for
>
> ajax
>
> eg
