Mainframe Question
in [Cobol]
|
From: Lee on 2 Feb 2006 02:17 Fujitsu claims their (incorrect) result is the same as that produced by IBM COBOL on MVS or whatever they run nowadays. Can someone please confirm this? TIA 01 W-Depr-Rem Pic S9(15)V9(2) Comp Value +0. 01 W-Est-Life-Y Pic S9(11)V9(6) Comp Value +5. 01 W-Depr-Mhs Pic S9(7) Comp Value +2. 01 W-Value Pic S9(11)V9(6) Comp Value +14800. 01 W-Depr-Tot Pic S9(11)V9(6) Comp Value +0. 01 W-Depr Pic S9(11)V9(6) Comp Value +0. PROCEDURE DIVISION. Compute W-Depr = (2 / W-Est-Life-Y) * (W-Value - W-Depr-Tot) * (W-Depr-Mhs /12). *The result of the above calculation using Fujitsu is +00000000986.662720 *The same calculation in Microfocus is +00000000986.666666
From: on 2 Feb 2006 07:47 In article <vmiEf.40880$dW3.22826(a)newssvr21.news.prodigy.com>, Lee <flu8349nospam(a)nospam.yahoo.com> wrote: >Fujitsu claims their (incorrect) result is the same as that produced by IBM >COBOL on MVS or whatever they run nowadays. Can someone please confirm >this? TIA > > > > 01 W-Depr-Rem Pic S9(15)V9(2) Comp Value +0. > 01 W-Est-Life-Y Pic S9(11)V9(6) Comp Value +5. > 01 W-Depr-Mhs Pic S9(7) Comp Value +2. > 01 W-Value Pic S9(11)V9(6) Comp Value +14800. > 01 W-Depr-Tot Pic S9(11)V9(6) Comp Value +0. > 01 W-Depr Pic S9(11)V9(6) Comp Value +0. > > PROCEDURE DIVISION. > > Compute W-Depr = (2 / W-Est-Life-Y) * (W-Value - W-Depr-Tot) * >(W-Depr-Mhs /12). > > > > *The result of the above calculation using Fujitsu is >+00000000986.662720 > *The same calculation in Microfocus is +00000000986.666666 Using IBM Enterprise COBOL for z/OS and OS/390 3.2.0 and the same code as above but including: 01 W-DEPR-DISP PIC Z(11).9(6)+ VALUE ZEROES. .... MOVE W-DEPR TO W-DEPR-DISP. DISPLAY ' DEPR DISP = ', W-DEPR-DISP. .... .... the compile generates a return-code of 4 (IGYPS0001-W A blank was missing before character "1" in column 42. A blank was assumed.) and throws out the following in the JESlog: DEPR DISP = 986.662720+ DD
From: Michael Mattias on 2 Feb 2006 09:07 "Lee" <flu8349nospam(a)nospam.yahoo.com> wrote in message news:vmiEf.40880$dW3.22826(a)newssvr21.news.prodigy.com... > Fujitsu claims their (incorrect) result is the same as that produced by IBM > COBOL on MVS or whatever they run nowadays. Can someone please confirm > this? TIA > *The result of the above calculation using Fujitsu is > +00000000986.662720 > *The same calculation in Microfocus is +00000000986.666666 That 'same as' claim should probably be qualified, since you can get different results on IBM depending on compile-time options;e.g., the TRUNC and FLOAT options come to mind for arithmetic. Besides, in the example you have here, this code is not written correctly if you want that many digits of precision in the answer. Your intermediate values are almost certainly getting rounded 'somewhere' and depending on where that occurs it will affect the final result to varying degrees. You need to control the rounding yourself and/or COMPUTE the intermediate values yourself to WS variables of sufficient precision. MCM
From: Sergey Kashyrin on 2 Feb 2006 09:07 That's the question which result is incorrect. I don't know Cobol that well but as I can recall 70-s, for PL/I there was an IBM document describing exactly how FIXED operations performing. What I mean is what's the TYPE of the (intermidiate) result of (W-Depr-Mhs /12). Because it's just "12", the result type of this operation should be the type of W-Depr-Mhs, i.e. S9(11)V9(6) In this case IBM result is correct and Microfocus (as well as OpenCobol) are wrong ! If you change "12" by "12.000000000000" you'll get Microfocus result. BTW, AS/400 is giving the same result as 390. Regards, Sergey. "Lee" <flu8349nospam(a)nospam.yahoo.com> wrote in message news:vmiEf.40880$dW3.22826(a)newssvr21.news.prodigy.com... > Fujitsu claims their (incorrect) result is the same as that produced by > IBM COBOL on MVS or whatever they run nowadays. Can someone please > confirm this? TIA > > > > 01 W-Depr-Rem Pic S9(15)V9(2) Comp Value +0. > 01 W-Est-Life-Y Pic S9(11)V9(6) Comp Value +5. > 01 W-Depr-Mhs Pic S9(7) Comp Value +2. > 01 W-Value Pic S9(11)V9(6) Comp Value +14800. > 01 W-Depr-Tot Pic S9(11)V9(6) Comp Value +0. > 01 W-Depr Pic S9(11)V9(6) Comp Value +0. > > PROCEDURE DIVISION. > > Compute W-Depr = (2 / W-Est-Life-Y) * (W-Value - W-Depr-Tot) * > (W-Depr-Mhs /12). > > > > *The result of the above calculation using Fujitsu is > +00000000986.662720 > *The same calculation in Microfocus is +00000000986.666666 > > >
From: Rick Smith on 2 Feb 2006 10:26
"Lee" <flu8349nospam(a)nospam.yahoo.com> wrote in message news:vmiEf.40880$dW3.22826(a)newssvr21.news.prodigy.com... > Fujitsu claims their (incorrect) result is the same as that produced by IBM > COBOL on MVS or whatever they run nowadays. Can someone please confirm > this? TIA > > > > 01 W-Depr-Rem Pic S9(15)V9(2) Comp Value +0. > 01 W-Est-Life-Y Pic S9(11)V9(6) Comp Value +5. > 01 W-Depr-Mhs Pic S9(7) Comp Value +2. > 01 W-Value Pic S9(11)V9(6) Comp Value +14800. > 01 W-Depr-Tot Pic S9(11)V9(6) Comp Value +0. > 01 W-Depr Pic S9(11)V9(6) Comp Value +0. > > PROCEDURE DIVISION. > > Compute W-Depr = (2 / W-Est-Life-Y) * (W-Value - W-Depr-Tot) * > (W-Depr-Mhs /12). > > > > *The result of the above calculation using Fujitsu is > +00000000986.662720 > *The same calculation in Microfocus is +00000000986.666666 Using the directive ARITHMETIC"VSC2" with Micro Focus 3.2.24 gives +00000000986.662720. Without the directive the result is +00000000986.666666. |