Prior to v7.0, Object ID's (OID's) were used for lining up data points between a client and the IOServer. Thus when changes were made to OID's for tags on an IOServer, these changes would then have to be copied out to all clients otherwise there was a potential for a mismatch. So, for example, if TagA was OID 1 and TagB was OID 2 and then these were swapped, you may end up with a situation where the client is asking for OID 1, expecting it to contain data for TagA, but the IOServer responds with data for TagB, resulting in the client displaying incorrect data. Similarly, if the client requests data that is no longer valid at all on the IOServer, this may result in a crash. (For more information on dealing with these issues in v6.x, see Q4641).

From v7.0, this is no longer an issue. This is because the client requests data based on Tag name, rather than OID. Thus if you change tags on the IOServer, you need only restart the server in order for changes to take effect. (The exception is if you have altered pages to point to different tags, in which case the client projects must be updated either manually or using the copy parameter detailed in Q2267.)