1 Reply Latest reply on Nov 3, 2006 11:41 AM by

    External references

      I agree. There is a tendency toward rabid fear of in-context relations which is not completely rational. Watch out, some of these people like to break all external relations. Locking them is a more sensible thing to do because locked relations can be unlocked, broken external relations cannot be unbroken.

      If you really want to see some of them flip, turn on the mulitple contexts switch. not that its a good idea.

      On the other side, though, overuse of in-context can cause you a lot of problems. It's hard to say how much is too much, but you need to make sure that you are using external relations in a controlled way. If you use them a lot, relations can become "circular" (parts referencing themselves) or daisy chained (long string of references), both of which are bad situations from understanding design intent and also for assembly performance.

      As the other guy said, parts that are reused in other assemblies probably shouldn't have external relations.

      As for the part about no mates, this thinking makes me shake my head and wonder why. This non-parametric approach to assembly modeling is silly and paranoid. I've seen other people get off on this as a good approach or "best practice" because none of your mates will ever blow up. That's true, but you don't get any benefit from them either. It's ok if your designs never change or don't move.

      Anyway, mates and in-context are good things, but don't use too much in-context. You shouldn't be afraid of it, but you should respect it. If you are disciplined and understand what is going on, there is little to worry about.
        • External references
          due to problems with parts created with external references used in other assemblies and how our PLM system works, we do not use external references either. For true top down design where all parts are only used in one assembly (no spin offs either), it probably works great.