Just out of curiosity, how many possible rounding rules can there be? I shouldn’t imagine more than a handful. I don’t remember where I saw it, but some application I used in the past had options to choose which of several rounding rules you wanted to use. (I also don’t recall if this was financial software or not, sorry.)
I would think the solution (if the set of rules is in fact small) is to implement them all and let the user choose the one they wanted/needed. Thus it isn’t ‘impossible’ to solve, just as usual, not as easy as it first looks. Regards, Adrien > On May 23, 2018, at 7:39 AM, Mike or Penny Novack > <stepbystepf...@dialup4less.com> wrote: > > There is no solution to this. Among other things, it would be dependent on > the rules of the jurisdiction whether the tax is supposed to be figured per > item or on the total of all taxable items. So what would fix it in one case > would break it in the other. It simply is true that rounded(a) + rounded(b) > may not equal rounded(a+b). One of the things I used to do in my working days > was to calculate the correct "fuzz" (maximum discrepancy) for comparison > tests << effective equality >> > > It is not only with this that we have a problem. Those of us who keep books > to exact amounts but must file with governmental agencies on a whole dollar > basis know how much "fun" it is deciding which amounts to juggle up or down > (what is the LEAST alteration) as necessary to the book amounts to make the > whole dollar report come out right. This this year, for one organization, > fudging two amounts by a few cents was enough to make the 990-EZ come out > right. > > BTW, there are analogous problems with things like figuring amortization > tables for loans )) how much of each payment is interest and how much > principle). There is no "right" way to do this and so any function used by a > program might not match what the particular bank says. In cases like these we > can't ask that the program be "fixed" to agree. > > Michael D Novack > > > On 5/22/2018 9:11 PM, Matthew Pounsett wrote: >> I've got an off-by-a-cent error on an invoice due to the way taxes are >> calculated. GnuCash (2.6.21) is calculating the tax on each individual >> taxable item, then summing that up, and adding that value to the subtotal. >> This causes an error because each individual tax calculation is rounded >> prior to subtotalling. >> >> Given a tax rate of 13%, GnuCash is doing this: >> >> Item Tax >> 277.50 36.08 >> 92.50 12.03 >> Subtotal: 370.00 + 48.11 = 418.11 >> >> When the real calculation should be: >> >> Item Tax >> 277.50 36.075 >> 92.50 12.025 >> Subtotal: 370.00 + 48.100 = 418.10 >> >> This could be fixed by GnuCash not rounding tax amounts until the final >> total, but typically I believe the calculation done is to add up all the >> taxable items and apply the tax calculation to that subtotal, so that it's >> not necessary to track many decimal places to avoid a rounding error. >> >> I did some searching through the mailing lists and I see a lot of reported >> issues about rounding errors in GnuCash 2.x. I can work around this by >> putting in an extra line item to correct the taxes, but it would be nice >> not to have to do that. Is this something that's been fixed in 3.x? I've >> been avoiding the update because I haven't had time for a careful >> migration, and testing whether I'm affected by any of the various issues >> that have been reported since its release. But, this might be a reason to >> set that time aside. >> _______________________________________________ >> gnucash-user mailing list >> gnucash-user@gnucash.org >> To update your subscription preferences or to unsubscribe: >> https://lists.gnucash.org/mailman/listinfo/gnucash-user >> If you are using Nabble or Gmane, please see >> https://wiki.gnucash.org/wiki/Mailing_Lists for more information. >> ----- >> Please remember to CC this list on all your replies. >> You can do this by using Reply-To-List or Reply-All. >> > > > -- > There is no possibility of social justice on a dead planet except the > equality of the grave. > > _______________________________________________ > gnucash-user mailing list > gnucash-user@gnucash.org > To update your subscription preferences or to unsubscribe: > https://lists.gnucash.org/mailman/listinfo/gnucash-user > If you are using Nabble or Gmane, please see > https://wiki.gnucash.org/wiki/Mailing_Lists for more information. > ----- > Please remember to CC this list on all your replies. > You can do this by using Reply-To-List or Reply-All. > _______________________________________________ gnucash-user mailing list gnucash-user@gnucash.org To update your subscription preferences or to unsubscribe: https://lists.gnucash.org/mailman/listinfo/gnucash-user If you are using Nabble or Gmane, please see https://wiki.gnucash.org/wiki/Mailing_Lists for more information. ----- Please remember to CC this list on all your replies. You can do this by using Reply-To-List or Reply-All.
