Pierre van Rooden wrote: > > >No. The only thing is that an mmbase restriction has a flag > >'always/oncreate/onchange/never', which main goal was e.g. to avoid > >problems with a restriction like 'field value > now', which is a nice one > >during creation of the node, but when you save it again on a later time, > >the field value may have become smaller then 'now', so you don't want to > >enforce the restriction then. > > > > > I still feel that a basetype should not have any restriction, even > 'soft' ones, other than that the data entered is compatible for its type. > Sub-type should be stricter than the base type, not looser.
I would have loved to see the database type 'string' with specializations 'string' and 'field'. But that is not possible. 'specializations' are now merely a way to define inheritance in datatypes. If we want that the dataype 'string' behaves more or less the same as it did, then I think this is best (it must somehow look like the guitype 'string' is the idea). I think we agree that it is unfortunate that now string itself has restriction which is a bit stricter then the restriction on the sub-type 'field'. But happily it is only an unenforced restriction, so there is no real problem. Michiel -- Michiel Meeuwissen mihxil' Peperbus 111 MediaPark H'sum [] () +31 (0)35 6772979 nl_NL eo_XX en_US _______________________________________________ Developers mailing list [email protected] http://lists.mmbase.org/mailman/listinfo/developers
