Re: Multiple fields from Multiple tables Resident Load
I was hoping to avoid resident, but the problem I had was getting the complete list of UIDs into table 5 to apply the maps to. With only the mapping loads, when I tried to use fieldvalue() to efficiently get the complete list of UIDs, it just gave me an error, I assume because the mapping tables aren't considered REAL tables in the data model. That's why I had to resort to doing resident. It would be a much cleaner script if there were a way to avoid that.
Thoughts on doing it without the resident load?
If we COULD avoid the resident load, then I would indeed expect it to run faster with the mapping tables. But if the resident is necessary, I'm guessing it'd run slower. Testing would tell us for sure, but I'm not going to take the trouble, and testing results on random fake data might not be the same as testing results on the real data set.
I feel like I should advise people to keep the script simple even at the cost of efficiency. But I know our own publisher server is quite overloaded, so in practice, I end up doing anything I can do to be more efficient, even at the cost of complexity. I guess it depends on your environment. If you can get away with simple, I like simple.