Aha! These are the kinds of things we need to stamp out before that last bit of shim can be deleted.
I think we should yank out that line and just make LzDataText inherit from LzMiniNode (instead of instance). That seems a lot more honest. Really LzDataNode as a mix-in requires either LzNode or LzMiniNode as a superclass. That's the minimal change we could make. We still need to sort this out better, like split LzNode in two or something. Maybe next time we are all in one room we could do that. On 2006-10-11, at 22:46 EDT, Henry Minsky wrote: > DEBUG: Shim <javascript:top.$modules.lz.Debug.displayObj(1) > >LzDataText<javascript:top.$modules.lz.Debug.displayObj(0) > >.setAttribute > [2] > > I got this while running a regression test, test/lztest/lztest- > lzdatatext.lzx, > > There is this code inside the class definition of LzDataText: > > prototype.setAttribute = LzNode.prototype.setAttribute; > > Didn't we have a mini-node with setAttribute method defined ? I > guess that > was on LzDataElement only? > > I should run backtrace to see where this is geting called.. not > sure who > needs to be calling setAttribute > on an LzDataText. Maybe in it's initializer it's getting called > someplace.. > > > -- > Henry Minsky > Software Architect > [EMAIL PROTECTED] _______________________________________________ Laszlo-dev mailing list [email protected] http://www.openlaszlo.org/mailman/listinfo/laszlo-dev
