tag:blogger.com,1999:blog-876358347971598886.post7459778411414054181..comments2009-11-30T14:21:44.955+02:00Comments on nothingmuch's perl blog: Modeling identity with KiokuX::Usernothingmuchhttp://www.blogger.com/profile/03975438115490089158nothingmuch@woobling.orgBlogger3125tag:blogger.com,1999:blog-876358347971598886.post-29371891951336493362009-05-20T21:29:50.768+03:002009-05-20T21:29:50.768+03:00Oh, and if you do need to wrap them, use delegatio...Oh, and if you do need to wrap them, use delegation, not subclassing. There are actually 3 different concrete set types depending on their materialization state (Transient, for sets that were created by the user, Deferred for lazy loaded sets created by the linker, and Loaded for vivified sets created by the linker. Deferred and Loaded behaviors are swapped by reblessing).nothingmuchhttp://www.blogger.com/profile/03975438115490089158noreply@blogger.comtag:blogger.com,1999:blog-876358347971598886.post-27950424196373250002009-05-20T21:04:14.764+03:002009-05-20T21:04:14.764+03:00No. You could, but it's a much simpler object, so ...No. You could, but it's a much simpler object, so usually no functionality is necessary (it's just a dumb collection). It's derived from the Set::Object APInothingmuchhttp://www.blogger.com/profile/03975438115490089158noreply@blogger.comtag:blogger.com,1999:blog-876358347971598886.post-34158440462102639022009-05-20T18:13:46.329+03:002009-05-20T18:13:46.329+03:00I'm a bit confused about Kioku Sets, should you su...I'm a bit confused about Kioku Sets, should you subclass them like you do with DBIC ResultSets in order to model behaviors operating on Set, or should there be a different way to approache this?johnhttp://www.blogger.com/profile/01931127394884250158noreply@blogger.com