[Moose-dev] Re: Mondrian in VW: event question

Hi Alex, Yes, it is the intended behavior, because the rule is that all settings (shape, figure and interaction) get reset after a #nodes: or #edges:from:to: command. The reason is this. If you add several nodes of the same type, in 99% of the cases you will have them in one collection, and if you want them to behave the same one #nodes: command and one set of settings is all you need. If they are not in the same collection, then you can probably simply construct a collection. If you want them to behave or appear differently, then you want to have control over the settings for both cases. If you add several nodes of different types, then you want to have several #nodes commands. However, because there are several types the settings that work for one set of nodes do not work for the settings of another type. Let's change your example a bit: | view | view := ViewRenderer new. view interaction popupText: [:e | (e + 1) printString]. view node: 1. view node: 'a'. view open If you propagate the settings to the 'a' node, you will get an error because String does not understand #+. That is why we decided to have the settings reset after each command. There are indeed a few cases when you will have to duplicate some of your settings code, but the compromise is quite good. Cheers, Doru On 1 Apr 2009, at 12:05, Alexandre Bergel wrote: -- www.tudorgirba.com "It's not how it is, it is how we see it."