| 43 | | It will be useful to start with a discussion about the requirements and what the features of patterns should be. What is expected from patterns? One view is that patterns should have the possibility to represent the mappings from templates and RDF to the full representation. Existing rule languages are not considered enough, and Victor has proposed a way to proceed using RDF. However, there is some uncertainty related to other possible requirements, e.g. what are the requirements from iRing Tools environment? We need input on the proposal from Victor from iRing Tools architects. Keith and Rahul have considered a similar approach as Victor. They will review material and provide feedback in the near future. |
| 44 | | |
| 45 | | A challenge today is the identification problem ("the intermediate identifier problem") - e.g. how to know that a set of properties actually belongs to the same stream instead of "inventing" new streams for each property. This problem is a challenge with TIPs because of the use of local identification. Currently there is no way to say that a stream is identical to some other stream. Therefore there is a need to expose the identifiers. Victor will create a thread in forum on alternative ways to solve the stream identification problem. What are the assumptions that related to a TIP today? One should model these assumtions to make them explicit and avoid ambigutity. |
| 46 | | |
| 47 | | One possibility is to require more during the modeling of the pattern (e.g. classify incoming og outcoming stream of a pump). Such an approach requires more work on the design of the pattern. |
| 48 | | |
| | 43 | It will be useful to start with a discussion about the requirements and what the features of patterns should be. What is expected from patterns? One view is that patterns should have the possibility to represent the mappings from templates and RDF to the full representation. Existing rule languages are not considered enough, and Victor has proposed a way to proceed using RDF. However, there is some uncertainty related to other possible requirements, e.g. what are the requirements from iRing Tools environment? We need input on the proposal from Victor from iRing Tools architects. Keith and Rahul have considered a similar approach as Victor. They will review material and provide feedback in the near future.[[BR]] |
| | 44 | |
| | 45 | A challenge today is the identification problem ("the intermediate identifier problem") - e.g. how to know that a set of properties actually belongs to the same stream instead of "inventing" new streams for each property. This problem is a challenge with TIPs because of the use of local identification. Currently there is no way to say that a stream is identical to some other stream. Therefore there is a need to expose the identifiers. Victor will create a thread in forum on alternative ways to solve the stream identification problem. What are the assumptions that related to a TIP today? One should model these assumtions to make them explicit and avoid ambigutity. [[BR]] |
| | 46 | |
| | 47 | One possibility is to require more during the modeling of the pattern (e.g. classify incoming og outcoming stream of a pump). Such an approach requires more work on the design of the pattern.[[BR]] |
| | 48 | |
