Query Help Please

Question

Hi... having problems with a query.  Any assistance would be
much appreciated.
Two queries with identical columns: Villages_Query_1 and Villages_Query_2.
Both have these columns: Village_ID, Village_Name, Player_ID.
I need to find all records in Villages_Query_2 where the Village_ID's match but the Player_ID's have changed.
Example Village_Query_1
Village_ID
Village_Name
Player_ID
1
Houston
1
2
Dallas
2
3
Chicago
3
Example Village_Query_2
Village_ID
Village_Name
Player_ID
1
Houston
1
2
Phoenix
4
3
Chicago
3
4
New York
5
In this case, Village_ID = 2, has changed names (Dallas to Phoenix) and the Player_ID has changed (2 to 4).  In addition, a new record was added.
The eventual output I need is to be able to report the following:
Player 2 village "Dallas" was taken by Player 4 and renamed "Phoenix".
New York is a new village owned by Player 5.
How the heck do I do this??  I have been trying query after query... reading about query of queries and JOINS and and and... I am now completely confused.
Help appreciated.
Mark

Answer 1

Well... firstly... you do not use MS Access for that volume of data.  Plain and simple.  MS Access is for DBs like "My CD collection".  It's a desktop application, and is not intended to be used other than as a desktop application.
Part of the reason for it not being appropriate for the job is that it can't do things like bulk loading data, which is kinda what you're wanting to do here.  That aside, it's a single-user file-based DB which is simply not designed to work as the back-end for a web application (or any sort of serious application).
Anyway, I would approach this by putting all the data from the CSV files into the DB as is.  Then on the DB run a query which gets all your changes.  You're really going to struggle with the suggestions here to use valueList() to generate a list that is then used for a NOT IN(#list here#), because you're likely to have a mighty long list there.  Even proper DBs like Oracle only allow 2000 entries in a list like that (SQL Server is about the same, from memory), so I doubt QoQ will allow even that.  The reason the DBs put limits on these things is that doing a WHERE IN (#list#) is a really poorly-performing process.
If you've got all your data in the DB, then your query becomes pretty easy, and I'm sure even Access could cope with it.  it'd be something like this:
SELECT VB.village_id, VB.village_name AS village_old_name, VB.player_id AS player_old_id,
VU.village_id AS village_new_id, VU.village_name AS village_new_name, VU.player_id as player_new_id
FROM villages_base VB
RIGHT OUTER JOIN villages_updates VU
ON VB.village_id = VU.village_id
WHERE VB.village_name != VU.village_name
(that's untested and I only gave it about 1min thought before typing it in, so don't quote me on that!)
Where VILLAGE_BASE is your original data, and VILLAGE_UPDATES is the data that indicates the changes.  I'm kinda guessing that this is the sort of thing you want.  Note: the "new" villages will be the ones which have NULLs for the village_id, village_old_name and player_old_id.
Getting all the data into the DB is going to be a matter of looping over the CSV file and doing an INSERT for each row.  And that will take as long as it takes, so you might need to get some control over your request timeouts.  However doing these inserts will take less time than all the QoQ logic suggested before, so you might be OK.  And the query should be quick.
What happens to the data once the report is written?  Does the "updated" data become the "live" data?  If so, after you run your report you're gonna want to do something like a TRUNCATE on villages_base, and INSERT all the records from villages_update into it (then TRUNCATE villages_update, ready for the next time you need to run this process).  Although don't take my word for it here, as I'm guessing your requirement here ;-)
Adam