Difference between Null values and Low values

meenakshi_forum · Active User Joined: 27 May 2008 Posts: 121 Location: India

Hi,

What's the difference between Null values and Low Values ?

I want to know it in case of both File and DB2 table.

What are Null values and Low values in a file?

What are Null vlaues and Low values in a DB2 table?

Thanks

Robert Sample · Posted: Thu Jan 07, 2010 11:41 pm

NULL means the variable has no value -- LOW-VALUES is a value of hexadecimal zeroes. You cannot have NULL values in a file since NULL values means there's a SECOND field that specifies whether or not the variable has data in it. DB2 supports NULL values as it is a database concept, not a file concept.

meenakshi_forum · Active User Joined: 27 May 2008 Posts: 121 Location: India

Some people interchangeably speak Null or Low values, stating that somethimes low values are to be considered as Null, i am bit confused?

What's the cause of occurence of Low value and High value, and why do they occur?

How absence of values are treated in a file? As we say absence of value in a table implies Null; can we have absence of value in a file?

--Thanks

Robert Sample · Posted: Fri Jan 08, 2010 12:19 am

The confusion comes because C and similar languages use the concept of a null-terminated string (which is a string of bytes that ends with a LOW-VALUE or x'00' which means no bits are set). The null in null-terminated is not the same as the concept of NULL in a database (which means the variable has no value, and that requires an indicator flag that can be set to true if the variable is NULL and false if the variable has a value). HIGH-VALUE is x'FF' which means all bits are set. In any given collating sequence (EBCDIC, ASCII, or whatever), LOW-VALUE is the first byte in the collating sequence while HIGH-VALUE is the last value in the collating sequence.

The absence of values in a file means there's no data there -- period. The file is empty and has no records. If there are data records in a file, there are no missing values -- every byte means something in a file and you cannot have missing values (as I said in reply to your other post).

meenakshi_forum · Active User Joined: 27 May 2008 Posts: 121 Location: India

Thanks Robert.

Just imagine that a file has some fields say - Account Number, Name, Address, Account Opening Date and some more.

Suppose, just a assumption and imagination that a person's Address is not available; then how would the file look like.

How do Low values and High values occur in a file or DataBase?

How a descision is taken for Low values and High values in a file or Data Base? By decision i mean how it is determined that field should have either Low val or High val?

Thanks

Robert Sample · Posted: Mon Jan 11, 2010 1:52 am

Depending upon the application, the address could have spaces in it or low values or high values or anything else -- whatever the application determines should be placed in the file. However, there will be something in those bytes -- they will be present in the file with some value. Low values and high values have no specific meaning in a file -- it depends entirely upon the application. Some databases may denote specific meanings to low values or high values in some cases, but again it depends upon the database software and the application that is using those fields.

The decision to use low values or high values tends to occur during the design process; however, this is not a fixed and firm rule. For example, if the design process determines a field in a file should be a 2-byte binary field (COMP in COBOL terms), then a velue of zero for that field is x'0000' while a value of -1 for that field is x'FFFF' -- in this case both low values and high values have specific meanings for the field even though neither was probably specifically considered during the design. So sometimes the design includes them on purpose, sometimes by accident, sometimes neither one is used.

Bill O'Boyle · Posted: Mon Jan 11, 2010 1:59 am

Speaking only from a file perspective, each field (IE: Account Number, Name, Address, Account Opening Date, etc), have a pre-defined fixed-length area in the record layout.

Depending on the IN-HOUSE requirements, if a given field is initialized with LOW-VALUES, SPACES, HIGH-VALUES or some other SITE DEPENDENT default value, then this is how a given field can be recognized as not being present.

I can't speak for a DB definition/design, but I'm sure someone else will jump in....

Bill

PS. Robert was seven minutes ahead....

dick scherrer · Posted: Mon Jan 11, 2010 3:18 am

Hello,