The pattern the Kernal has been using is to keep a reference to
'this' in a slot called owner. The reference is passed in at construct
time. From the Kernel spec:
new TextSprite( Object:owner )
The constructor has one argument:
- a reference to the owner object
Is implemented as:
LzTextSprite = function(owner) {
if (owner == null) return;
this.owner = owner;
It would be good to standardize on this pattern, unless anyone has
objections.
-Max
Henry Minsky wrote:
I've been getting away with binding some other variable to
"this" to capture it...
On 7/14/06, P T Withington <
[EMAIL PROTECTED]> wrote:
I
think that has to be the case. Just as the API's have to use plain
_javascript_ types (cannot use LFC types), we need to use a plain
_javascript_ mechanism for events across this API.
Note that for ECMA4 it is proposed that a method closure will capture
`this`. Perhaps there is a way to simulate this in the compiler, if
that would help?
On 2006-07-14, at 12:29 EDT, Henry Minsky wrote:
> So did we decide that it is a policy that kernel functions don't
> send Laszlo
> Events back to the LFC, they just use callback functions?
> I don't really have a strong feeling one way or the other on this.
>
>
> On 7/14/06, Max Carlson <[EMAIL PROTECTED]>
wrote:
>>
>>
>> Henry Minsky wrote:
>> >
>> >
>> >
>> >
>> > Or be more like XMLHTTPRequest, where the request
type is
>> just a
>> > parameter, instead of having a method for each
request type?
>> >
>> >
>> > I actually like that better, given that we may want to
add other
>> > parameters.
>> >
>> This does sound better. I'll update the Kernel API Spec.
>>
>> -Max
>>
>
>
>
> --
> Henry Minsky
> Software Architect
> [EMAIL PROTECTED]
--
Henry Minsky
Software Architect
[EMAIL PROTECTED]
|
_______________________________________________
Laszlo-dev mailing list
[email protected]
http://www.openlaszlo.org/mailman/listinfo/laszlo-dev
