The basic question is, "do we want to make it easier for 
new developers to switch from Struts/JSP to Tapestry?".

I think the answer is yes.  More users is more 
acceptance is more likely to use Tapestry on your next 
consulting engagement.

The RAD concept I'm floating is basically 
saying "Tapestry is like JSP, but much more natural.  
You can start with ordinary looking HTML built like a 
normal JSP, and expand up from there."

Outside of NeXT and IB, I don't know anyone who has 
created a good component-based WYSIWYG component 
editor.  Tapestry is trying to let ordinary HomeSite and 
the like be the WYSIWYG editor.

So we have a tradeoff; beginning users put more cruft in 
the HTML, but it still is valid HTML (with extra 
arguments) ... it's not HTML + JSP Directives + XML-like 
tag library references + includes.

I don't like having two paradigms in Tapestry, and this 
RAD paradigm is not nearly as capable as the exiting 
paradigm.  I especially don't like the fact that the 
attributes are treated like OGNL expressions (for 
anonymous components) and are treated like static 
bindings (for informal parameters on a traditional 
component).

But if we can get people to sit down and get something 
working in Tapestry in an hour, then everyone wins.

Again, will it be abused?  Yes.  Will some people get 
confused about what Tapestry really is?  Yes (but they 
would probably not know about it or understand it 
anyway).  Is it a lot of work?  No.

The idea is growing on me.  I'm not in 100% agreement 
with Marc; he is confused and thinks that this 
simplification is what most people should use whereas I 
see it as a bait-and-switch to ease initial adoption, 
and get users hooked when they leave the RAD part behind.

--
[EMAIL PROTECTED]

http://tapestry.sf.net
> 
> On Thursday, Oct 10, 2002, at 18:07 Europe/Zurich, Adam Greene wrote:
> 
> > In regards to Tapestry RAD being able to load *.html and wrap it with a
> > generic specification.  What about adding a :
> 
> All this sounds "interesting", but would most likely turn into an 
> exercise in futility.
> 
> If you want to provide true RAD functionality in Tapestry it has 
> nothing to do with the underlying framework. Instead, the only obvious 
> choice would be to provide something akin to NeXT's InterfaceBuilder 
> for the web: wysiwyg "drawing" of the UI. And don't even waste your 
> time mentioning WOBuilder: it's a utter failure, not even remotely 
> close to IB. Also, something like IB implies two things: a set of 
> useful, generic components to get started. And a plug-in API ala 
> IBPalette to allow third party to provide their pre-build components. 
> With such a tool there is no need to handle any html template or 
> description files: the tool takes care of them. Your only "worry" is to 
> provide an (optional) implementation.
> 
> Such a tool, coupled with a "rapid turn around mode", would go a long 
> way to provide RAD capacities: draw & run. Nothing much to it.
> 
> Cheers,
> 
> PA.
> 
> 
> 
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> Tapestry-developer mailing list
> [EMAIL PROTECTED]
> https://lists.sourceforge.net/lists/listinfo/tapestry-developer


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Tapestry-developer mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/tapestry-developer

Reply via email to