Hi @nag78936
The difference between AutoNumber and AutoNumberHashN is that the first only has one parameter to work from and the others have a comma separated list of values.
If you want an auto-number based on a single field then just use AutoNumber.
You can have multiple fields with AutoNumber, but you need to concatenate them yourself, which is slightly inefficient, so you might opt for the hash functions.
The difference between the 128 and 256 is the length of string that is created before the auto-number is performed. This reduces the risk of collisions between two different values receiving the same auto-number. In practical terms it is very unlikely to occur even with AutoNumber128, but if you have a large number of rows you might want to opt for 256.
Also... I hardly ever use AutoNumber, as it can not work alongside persisted data in QVDs and incremental loads. In the main just using the longer string for joins is fine - the Qlik engine creates a key for the join anyway under the bonnet. Having long strings that are only used for joins does increase memory use though.
Hope that helps.
Steve
