The problem with this "filter" approach is that it needs to be built into the core above partitions, not at the level of interceptors. So that we can see it as a regular attribute and use in any operation. (But this makes it really hard to implement.) A related problem we have due to as-interceptors-implementation of virtual attributes is here:
http://issues.apache.org/jira/browse/DIRSERVER-824 We need to have virtual attributes just above partitions (or below the searcn engine). On Sat, Mar 28, 2009 at 11:48, Emmanuel Lecharny <[email protected]>wrote: > Ersin ER wrote: > >> Hi, >> >> The problem you mentioned and many more (collective attrs, dynamic groups) >> of that type are related to computed attributes. The solution is to >> develop >> a Virtual Attribute Computation layer on top of partitions. >> > I totally agree. Subentry should be seen as filters, instead of implemented > the way it is. I like to see them as 'layers' applied on top of the Dit. > They have the same structure than the Dit, but they restrict the number of > element you can "see" (in this case, "restrict" means applies a set of > ACIs). SO when processing an operation, you add the subentry filters on the > result, or even better, you use the subentry as a filter to limit the number > of returned elements. Then you don't need to modify all the entries the way > we do. I don't know if I explain corectly what I have in mind, but anyway... > > -- > -- > cordialement, regards, > Emmanuel Lécharny > www.iktek.com > directory.apache.org > > > -- Ersin ER http://www.ersiner.net
