Reorder Front Scripts?

[Richard Gaskin]

Ken Ray wrote:

> On 7/21/06 11:44 PM, "Scott Rossi" <scott at tactilemedia.com> wrote:
> 
>> Is there a way to change the order of front scripts without
>> removing/reinserting scripts?
> 
> Not as far as I know, Scott.
>  
>> It seems that Rev's inspector palette reinserts itself into the front every
>> time it is opened from the main menu and, because it apparently doesn't pass
>> all object manipulation messages, blocks front scripts I have that are used
>> for control layout.
> 
> It should be a "good citizen" and pass them... perhaps the next rev of Rev
> would do that?

In general, yes, and I know one IDE that already does, at least most of
the time.

But one issue with IDE message passing comes up anytime someone
wants to make an app that uses the pointer tool: how the IDE should
handle mouse messages while the pointer tool is active?

SuperCard separates pointer- and browse-tool behavior with a distinct 
suite of messages just for the pointer tool (pointerDown, pointerUp, 
etc.).  This is  useful and convenient, allowing a separation of 
behavior that works well since it's extremely rare that you'd ever want 
the same behavior with the browse and pointer tools.

In contrast, Revolution has only one set of messages for both tools, and 
to make matters more challenging the behavior of those messages is not 
consistent between tools (for example, mouseDown is sent when you click 
on a card away from controls with the browse tool, but not with the 
pointer tool, mouseEnter and mouseLeave are never sent with the pointer 
tool, etc.).

With only one set of messages, if the IDE passed these then clicking on 
an object with the pointer tool to resize would also trigger any 
mouseDown script in its message path.

You could write all of your scripts to check the tool before executing, 
but that's a lot of work.

Simpler would be to have a separate set of message devoted to supporting 
the pointer tool, as established long ago by SuperCard and requested for 
Rev in BZ#2606:
<http://support.runrev.com/bugdatabase/show_bug.cgi?id=2606>

This is no panacea for message passing however, since those using the 
pointer tool suite of messages would still need to be conscious of the 
IDE's behavior and be prepared to purge IDE front- and backscripts which 
trap them without passing.

But at least it isolates the scope of the problem to the subset of 
developers making custom drawing environments -- a task so daunting in 
Rev (esp,. compared to the simplicity of doing so in SC) that it's 
currently only for advanced intermediate to expert developers anyway.

--
  Richard Gaskin
  Managing Editor, revJournal
  _______________________________________________________
  Rev tips, tutorials and more: http://www.revJournal.com