click-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Adrian A." <a.adrian.t...@gmail.com>
Subject Re: Confusing page class mapping
Date Tue, 22 Jun 2010 08:12:52 GMT
> After Malcolm's comment I don't think we should move in this direction.
Well, I don't think this is "moving in that direction" since this is how 
click works since 0.x :). I still have some very old versions around, 
and also quite a few applications and even presentations that rely on 
this pattern :).

> The classname should be
> treated as an absolute name. Logically that makes sense and both IDE implementers interpreted
it
> this way too. Having absolute classname makes hotlinking quite straightforward, no mucking
around
> with the package name.
>
> Another advantage of this approach is its possible to define a page outside the defined
"package"
> without having to create another<pages>  element.
I wouldn't do that either :).
Another point of the <pages> element was also to separate "modules" of 
functionality, but without introducing a "module" tag in click.xml .

The idea of the "<page>" element was to use it *only* for the 
"exceptions" from the rule, the rule being specified in the "<pages>" 
element (e.g. automapping, autobinding, common package part, etc.), and 
not to be used on it's own, as click.xml should be not a "dependency 
injection" setting (as it looked for some users after the introduction 
of the Click Services) :).

Adrian.


Mime
View raw message