variable length records under fujitsu cobol (COBOL.NET)

I have created a small test COBOL program to produce a variable length
file in Fujitsi Cobol.NET.

I wanted to see how are the records stored. I was expecting to see each
record with leading 2 or 4 byte field holding the record length. I was
astonished to find the record had BOTH leading AND trailing 4 byte
field holding the record length.

Here is my program:

000010 IDENTIFICATION DIVISION.
000020
000030 PROGRAM-ID. Program1 AS "aritest.Program1".
000040
000050*=================================================================
000060
000070 ENVIRONMENT DIVISION.
000080
CONFIGURATION SECTION.
000090 INPUT-OUTPUT SECTION.
000100
000110 FILE-CONTROL.
000120
000130 SELECT TEST-FILEV
000150 ORGANIZATION SEQUENTIAL
ASSIGN TO MYFILEV.

000160
000170
000180
000190 DATA DIVISION.
000200
000210 FILE SECTION.
000220
000360
000230 FD TEST-FILEV
000240
000250 LABEL RECORDS STANDARD
000260
000270 RECORDING MODE V
000290 BLOCK CONTAINS 0.
000300
000310 01 TST-RECV.
05 VAR-TXT.
10 PIC X OCCURS 0 TO 100 DEPENDING ON VAR-LEN.
000320
000360
000370 WORKING-STORAGE SECTION.
000380 01 VAR-LEN PIC 9(2) COMP.
000390 01 KUKU PIC X(10).
01 DISP-NUM PIC 9(4).
01 II PIC S9(4) COMP.
000400


000430 PROCEDURE DIVISION.
000440
000450**************************************************************
000460
000470 A-PROGRAM-CONTROL SECTION.
000480
000490*----------------------------------------------------------------
000500
000510 DISPLAY "TEST PROGRAM COMMENCING".
000520
000580
000610 OPEN OUTPUT TEST-FILEV.
MOVE 5 TO VAR-LEN
MOVE 'ABCDEF' TO VAR-TXT
WRITE TST-RECV.
SUBTRACT 1 FROM VAR-LEN.
WRITE TST-RECV.
000690 CLOSE TEST-FILEV.

000710 END PROGRAM Program1.
000720
000730

Here is a dump of the output file:

0000000 005 000 000 000 101 102 103 104 105 005 000 000 000 004 000 000
005 \0 \0 \0 A B C D E 005 \0 \0 \0 004 \0 \0
0000020 000 101 102 103 104 004 000 000 000
\0 A B C D 004 \0 \0 \0
0000031


(I use an equivalent of the unix dump program, od -b -c)

Is there any way I can get rid of those trailing record lengths? They
are a real problem.

Thanks,

Ari

.