[Laszlo-dev] R5388 DHTML runtime error

[André Bargull]

Short way: Your delegate and your method have got the same name!
Long way. So, by creating the delegate and assigning it to the field 
"this.onx_del", you're replacing the slot for the method. Whenever the 
"onx"-Event is send, the delegate tries to invoke the method "onx_del", 
but in your object, the slot "onx_del" points to the delegate instead of 
the method. Too bad!

One last point, why don't file all these errors/bugs/etc. in JIRA, as it 
is the appropriate way of reporting bugs? I don't want to be harsh, but 
a little bit of clarification from your side would definitely clear up 
my confusion. Thank you!

Qrowd Support wrote:

> Nightly build 5388 crashes the DHTML runtime with this code:
>
> <canvas id="maincanvas" width="100%" height="100%" validate="false"
> debug="true" proxied="false" bgcolor="0x00aaaa">
>     <debug x="650" y="200" width="800" height="400" fontsize="12"/>
>
>     <window x="100" y="100" width="200" height="200">
>         <handler name="oninit">
>             this.onx_del = new LzDelegate(this, "onx_del", this, "onx");
>         </handler>
>         <method name="onx_del">
>             Debug.write("onx_del: "+this.x);
>         </method>
>     </window>
> </canvas>
>
> DHTML IE7 and FF both crash with debug and non-debug runtimes.
>
> FF debug DHTML gives this runtime error:
>
> ERROR: http://localhost:8080/lps/includes/lfc/LFCdhtml-debug.js:4041:
> $5_d.c[$5_d.f] is not a function
> <javascript:window.parent.$modules.lz.Debug.displayObj(5)>
>
> The test code is borderline nonsensical...SWF doesn't display the debug
> output...but the DHTML should fail gracefully.
> -- Robert Yeager Founder, Qrowd http://www.qrowd.com/blog 
> http://www.qrowd.com <http://www.qrowd.com>
-- 
Mit freundlichen Grüßen,
André Bargull
Intensis GmbH

Tel.: +49 (0)231 5522-900
Fax.: +49 (0)231 5522-901
E-Mail: a.bargull at intensis.de
________________________________
Intensis GmbH
Ruhrallee 9
44139 Dortmund