This page seems to be outdated before I even wrote it 😔. But please read on.....
Long ago when I was young they invented the concept of Null values in relational databases. The concept made sense but caused quite some problems, especially in Reporting type applications.
This is the concept : zero means zero and Null means "I do not know"
Imagine you have three boxes with money in them. In Box one you have 1000 €, In Box two you have 5000 € and in Box 3 you have 0 (zero) €. How much money do you have ? 1000 € + 5000 € + 0 € = 6000 €. Easy.
Now imagine in Box one you have 1000 €, in Box 2 you have 5000 €, but in Box 3 you have Null (I do not know) €.
Back in these days, when asked how much money do you have, most RDBMs would have responded Null (I do not know) .
Logically, this is correct: if you sum an unknown quantity to something, the result is unknown.
But in practice, imagine you get a million records in your data source and among them just one record with a Null value, when you try to SUM the values your answer is Null (I do not know). Not good.
Fortunately (in my opinion), nowadays most Relational Database systems have adapted better to reality and just ignore Null values.
So, in our previous example, the answer would be : you have 1000 € + 5000€ = 6000 €.
I would like to ask you a favour: If you currently are using or know of a DBMS that does NOT by default ignore Null values, drop me a message (bocebad@gmail.com).
Anyway, KNIME has very good ways to deal with Null values, we will have a look in the TODO next page....