Events and Methods [Was: [Laszlo-dev] RE: Idle callbacks are the devil's tools!]

[P T Withington]

I think what Oliver is thinking of is Adam's proposal that there  
should be an <event> tag, similar to an <attribute> tag that would be  
used to declare the events that a class sends.  Then we could update  
the schema and people could just use `onfoo=` instead of `<method  
event=`.

On 9 Jun 2005, at 17:04, Max Carlson wrote:

> This seems like a bad idea to me.  I've had a lot of cases where I  
> wanted to bind a method to an event, invoke it by name directly.
>
> Another thing that would be nice is if onevent could take a list,  
> allowing a single method could bind to multiple events.  One case  
> where I've wanted this is when I want a drawview to redraw when its  
> height or width are changed.  As of now, I have to have the event  
> bindings invoke a third method.
>
> -Max
>
> Oliver Steele wrote:
>
>> I think Tucker had a proposal to do away with the distinction  
>> between  event= and name=, and to instead specify that any method  
>> whose name  begins with 'on' is automatically registered as an  
>> event handler for  that event.
>> On Jun 8, 2005, at 2:30 AM, Max Carlson wrote:
>>
>>> I've been stumped over this very mistake a number of times.  It   
>>> would be nice if the compiler warned for events that don't begin   
>>> with 'on'.  I've filed a feature request at http://  
>>> www.openlaszlo.org/jira/browse/LPP-401
>>>
>>> -Max
>>>
>>> Don Hopkins wrote:
>>>
>>>
>>>> Sarah spotted the mistake I made in the following test. I was using
>>>> event="foobar" instead of name="foobar". It works correctly when  
>>>> I  use
>>>> name="foobar", since delegates call names, not events.     -Don
>>>> -----Original Message-----
>>>> From: Don Hopkins [mailto:dhopkins at DonHopkins.com] Sent:  
>>>> Tuesday,  June 07, 2005 5:15 PM
>>>> To: 'OpenLaszlo development development'
>>>> Cc: 'Don Hopkins'
>>>> Subject: Idle callbacks are the devil's tools!
>>>> Am I doing this wrong, or is there a bug in LzIdle.callOnIdle?   
>>>> When you run the following code (enclosed), the foobar method  
>>>> doesn't
>>>> get called. The documentation for callOnIdle is worded  
>>>> strangely.  It says:
>>>> LzIdle.callOnIdle(d) Calls the given delegate at the next idle   
>>>> event. This can be used for a
>>>> non- recursive callback.
>>>> Shouldn't that be "non-repeating"? (I don't think "recursive" is  
>>>> the
>>>> right word to describe a repeating callback.) The documentation   
>>>> should
>>>> be explicit about the fact that the delegate is called only  
>>>> once, and
>>>> then unregistered. (And the behavior of the code should match the
>>>> documentation! ;-) To work around this problem, I'm just   
>>>> registering an LzIdle.onidle
>>>> callback, and unregistering it when it happens. I want to defer  
>>>> some
>>>> layout processing until after all replication has happened. But
>>>> unfortunately the replication seems to be happening via onidle
>>>> processing too, so we're both waiting in the same line, and it   
>>>> calls my
>>>> lazy layout callback before all the replication is finished. Is   
>>>> there a way to register a lower priority onidle callback that will
>>>> only run once all the pending replication and instantiation is   
>>>> finished?
>>>> I'd like for my idle callback to go to the end of the line each  
>>>> time
>>>> something changes. If I unregister and reregister my onidle   
>>>> callback (instead of checking
>>>> to see if I'm already registered and not doing anything in that  
>>>> case,
>>>> keeping my place in line) will my delegate go to the end of the  
>>>> line
>>>> behind any pending replication callbacks? But that would require  
>>>> more
>>>> processing, so I'd prefer a lazier way to be lazy. I am   
>>>> considering using a timer instead, but that would introduce an
>>>> unnecessary delay, and it still might fire too early. It would  
>>>> be  nice to have a reallyidle event that gets fired when the
>>>> idle queue is empty.     -Don
>>>> <canvas
>>>>   width="100%" height="100%">
>>>>   <attribute name="del" value="null"/>
>>>>   <method event="oninit"><![CDATA[
>>>>       this.del = new LzDelegate(this, 'foobar');
>>>>       Debug.write("calling LzIdle.callOnIdle", this.del);
>>>>       LzIdle.callOnIdle(this.del);
>>>>     ]]>
>>>>   </method>
>>>>   <method event="foobar" args="args"><![CDATA[
>>>>       Debug.write("foobar should be called, but isn't...", args,
>>>> this.del);
>>>>     ]]>
>>>>   </method>
>>>> </canvas>
>>>> _______________________________________________
>>>> Laszlo-dev mailing list
>>>> Laszlo-dev at openlaszlo.org
>>>> http://www.openlaszlo.org/mailman/listinfo/laszlo-dev
>>>>
>>>>
>>>
>>> _______________________________________________
>>> Laszlo-dev mailing list
>>> Laszlo-dev at openlaszlo.org
>>> http://www.openlaszlo.org/mailman/listinfo/laszlo-dev
>>>
>>>
>
> _______________________________________________
> Laszlo-dev mailing list
> Laszlo-dev at openlaszlo.org
> http://www.openlaszlo.org/mailman/listinfo/laszlo-dev
>
>