Re: [gatein-dev] New field on node type gtn:visible - how to migrate the data?

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 On 01/31/2014 11:41 AM, Julien Viet wrote: Cool, thanks! I was talking with ppalaga about it yesterday, and I'll give it a try soon. I'm probably missing something, but it isn't clear to me what would be the benefits of this approach, other than "it is contextually correct" (which is in itself a valid reason). I hope to find a possible answer when I get this implemented, but I couldn't find anything on the documentation that would explain the difference or perhaps show the benefits of this approach. Perhaps there's an advantage on how the data is stored? In data consistency? In other words, this is how the code looks like in the PR, without the new mixin inheriting gtn:visible: // NavigationServiceImpl Visible visible = sourceNav.adapt(Visible.class); ... // do operations related to Visible.class if (sourceNav.isAdapted(RestrictAccess.class)) { ... // do operations related to RestrictAccess.class } else { chromatticSession.setEmbedded(sourceNav, RestrictAccess.class, restrictAccess); } With the RestrictAccess inheriting Visible, I suppose that the code would look like this: if (!sourceNav.isAdapted(RestrictAccess.class)) { ... // it's only Visible, so, probably an old node: replace Visible by RestrictAccess on the sourceNav } // do operations related to Visible.class // and operations related to RestrictAccess.class And this is how it currently looks like on NodeData: if (navigation.isAdapted(Visible.class)) { ... // operations about Visible } if (navigation.isAdapted(RestrictAccess.class)) { ... // operations about RestrictAccess.class } With the change, the second conditional could be placed inside the first, saving perhaps an unnecessary check if the *first* renders false (which, from what I understand, would rarely happen). Juca.