Updated documentation

modules/database/src/std/rec/compressRecord.dbd.pod

@@ -40,7 +40,7 @@ the beginning or the end of the VAL array.
 
 =head2 Parameter Fields
 
-The record-specific fields are described below.
+The record-specific fields are described below, grouped by functionality.
 
 =recordtype compress
 
@@ -60,10 +60,6 @@ menu(bufferingALG) {
 }
 recordtype(compress) {
 
-=head2 Parameter Fields
-
-The record-specific fields are described below, grouped by functionality.
-
 =head3 Scanning Parameters
 
 The compression record has the standard fields for specifying under what
@@ -85,7 +81,7 @@ algorithms which can be specified as follows:
 
 The following fields determine what channel to read and how to compress the data:
 
-=fields ALG, INP, NSAM, N, ILIL, IHIL, OFF, RES
+=fields ALG, INP, NSAM, N, ILIL, IHIL, OFF, RES, PBUF
 
 As stated above, the ALG field specifies which algorithm to be performed on the data.
 
@@ -167,6 +163,18 @@ Compress N to 1 samples, taking the median value.
 
 =back
 
+The behaviour of the record for partially filled buffers depends on the field PBUF.
+If PBUF is set to NO, then the record will wait until the buffer is completely full
+before processing. If PBUF is set to YES, then it will start processing immediately.
+
+For example, if ALG is set to C<<< N to 1 Average >>> with NSAM equal to 4, N equal
+to 1, and PBUF set to NO, then the first three times that the compress record is
+processed it will remain in an undefined state. On the fourth process, the average
+of all four records will be calculated and placed into the VAL field.
+
+If PBUF is set to YES, then after each process the average of the first several
+elements will be calculated.
+
 The compression record keeps NSAM data samples.
 
 The field N determines the number of elements to compress into each result.