Test Data Features
In order to test a new feature or system there has to be some data
for it to “practise” on. Some of it should be “real”,
so that if you were to test the input of a Goods Received Note,
for example, it is likely that you would want your correct company,
warehousing and chart of accounts details available, whereas your
inventory and Purchase Ledger data would be a dummy set (because
you would not want to corrupt your live data).
Nevertheless it can be dangerous to use well-known reference data
all the time; a feature of the new application may be to split or
merge accounts, offices, warehouses, or categories of the Nominal
Ledger. Data needs to be generated to model a situation where, for
example, half the inventory is placed in a new warehouse.
This would normally require a large number of simple but tedious
changes to many records in the test or regression test pack, so
that the data change is synchronised. Using a technique which involves
keeping all of the test data in a central repository, we can rapidly
emulate such changes at execution time, between one run and the
next.
- Back to Test Data Generation
Page
- Back to Services Page |
|
