Justinas, the currently proposed rules for ADL1.5 (see the draft), see section 5.3.12 are that at-codes are optional. This allows many leaf and near-leaf level nodes not to have to have useless at-codes attached to them. The same goes for any node that is a single child of an attribute, like HISTORY etc. The motivation for this is to allow for simpler paths, and keep the size of archetypes down to what is needed.
The C_OBJECT.node_id property is mandatory in the AOM. What we do in the current implementations is that the node_id is set to the value 'unknown'. This is currently not properly specified in the spec, but is implemented in the ADL workbench <http://www.openehr.org/svn/ref_impl_eiffel/TRUNK/apps/adl_workbench/doc/web/index.html>. hope this helps - thomas On 31/05/2011 12:12, Justinas Prelgauskas wrote: > Hello, > > I've been looking through a very useful set of ADL 1.5 test archetypes > related to Specialization feature of OpenEHR via "Knowledge2" > repository ( > http://www.openehr.org/wsvn/knowledge2/TRUNK/archetypes/ADL_1.5_test/validity/specialisation > ). > > I think some of archetype nodes lack "at"-codes. > For example OBSERVATION archetype > "openEHR-EHR-OBSERVATION.spec_test_obs.v1" uses such types as HISTORY > and ITEM_TREE. > However, both HISTORY and ITEM_TREE types are LOCATABLE. > Thus I think all such objects must have "at"-codes declared. > This must be true for all LOCATABLE objects, because they have > mandatory property "archetype_node_id". > Or maybe I am missing something? > > Regards, > Justinas Prelgauskas > > * > * -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.openehr.org/mailman/private/openehr-technical_lists.openehr.org/attachments/20110601/c5d8052c/attachment.html>
