HP Forum Archive 20

[ Return to Index | Top of Index ]

34s / 16C Compat. #5 - DBL/ Message #1 Posted by Jake Schwartz on 11 May 2011, 10:13 p.m. Hi, Using 34S build 874, I was attempting the double-word divide example in the 16C manual on page 54. With 2's compliment mode and a word size of 5, an attempt was made to divide the double word representation of -88 decimal in Y and Z by the single-word 11 decimal in X to get the single-word result in X of -8 decimal. All values were entered in binary, so the sequence is: 1000 ENTER 11101 ENTER 1011 DBL/ and the expected result is 11000 binary. Only, on the 34S I get zero as a result with the carry flag set. It does work on my 16C. The followup example in the manual on pages 54-55 does work, however, using unsigned mode and 64-bit word size. Any ideas? Thanks, Jake

Re: 34s / 16C Compat. #5 - DBL/ Message #2 Posted by Paul Dale on 11 May 2011, 11:01 p.m., in response to message #1 by Jake Schwartz Another integer bug :-( This one is broken. There is a problem joining the two halves of the number together when the word size is not 64. DBLMOD is broken too. DBL* will likely have problems with negatives as well :-( - Pauli

Re: 34s / 16C Compat. #5 - DBL/ Message #3 Posted by Paul Dale on 11 May 2011, 11:19 p.m., in response to message #1 by Jake Schwartz And hopefully fixed in the next build :-) - Pauli

Re: 34s / 16C Compat. #5 - DBL/ Message #4 Posted by Angel Martin on 12 May 2011, 1:56 a.m., in response to message #1 by Jake Schwartz That's really thorough beta testing - many SW development firms would pay big bucks for this... well, assuming they're after a real quality product instead of larger profits of course. Cheers, 'AM.

Re: 34s / 16C Compat. #5 - DBL/ Message #5 Posted by Paul Dale on 12 May 2011, 2:29 a.m., in response to message #4 by Angel Martin I'm glad someone is going through the integer mode in depth. Various people found quite a few bugs in the real mode fairly quickly and integer mode really needs similar attention. I tried to test when I coded everything but there was so much to implement and so many special cases. And things have changed quite a bit over time resulting in bit rot in the code. Still, I think we're getting close to a decent calculator. I've actually started using mine for real calculations at work :-) - Pauli

Re: 34s / 16C Compat. #5 - DBL/ Message #6 Posted by Marcus von Cube, Germany on 12 May 2011, 2:48 a.m., in response to message #5 by Paul Dale Pauli, how is your one looking? What type of overlay are you using?

Re: 34s / 16C Compat. #5 - DBL/ Message #7 Posted by Paul Dale on 12 May 2011, 3:29 a.m., in response to message #6 by Marcus von Cube, Germany My overlay hasn't changed from the picture posted here. Sadly, I don't need an overlay, I know what all the keys do :-) - Pauli

Integer Mode Message #8 Posted by Jake Schwartz on 12 May 2011, 11:22 a.m., in response to message #5 by Paul Dale Quote: I'm glad someone is going through the integer mode in depth. Hi, It is my intention to work through the entire 16C manual and report any anomalies....as long as I'm not considered a pain, I'll keep doing it. Thanks for everyones' diligent work, Jake

Re: Integer Mode Message #9 Posted by Walter B on 12 May 2011, 11:36 a.m., in response to message #8 by Jake Schwartz Hi Jake, all, as mentioned we appreciate all testing very much - it's quite obvious we can't do it alone nor will we be so successful as any thorough (wo)man not being part of the development team. So thanks for the checks you've done already - and in advance for those to come :-) Walter

[ Return to Index | Top of Index ]

Go back to the main exhibit hall