[GNC-dev] [GNC] Experimental Balance Sheet report

D sunfish62 at yahoo.com
Wed Sep 16 21:21:48 EDT 2020


Ok. Thanks. I'll leave it lie then. Never mind. 

On Sep 16, 2020, 19:47, at 19:47, Christopher Lam <christopher.lck at gmail.com> wrote:
>Hi David, responses inline. I'm sure you know it's my work, and I am
>(for
>now) happy to defend its state. TBH I do not need this report myself,
>but
>figure out it fills a gap. I also think devel is more appropriate.
>
>In my view it won't come out of experimental anytime soon because of
>https://bugs.gnucash.org/show_bug.cgi?id=797786 -- figuring out price
>from
>stock/foreign-currency into target currency is *HARD*. TL;DR: in some
>price-sources (avg-bal/wt-avg) current balance sheet will tally only
>txns
>UPTO report-date and calculate price from it; in multicolumn report
>there
>are multiple report-dates therefore this strategy would be far too
>slow.
>I'm running out of steam to make pricing behave identically to the old
>balance-sheet.
>
>On Wed, 16 Sep 2020 at 16:42, David T. via gnucash-user <
>gnucash-user at gnucash.org> wrote:
>
>> 1. When initially run, I receive an empty report. This is similar to
>the
>> initial result of the basic Transaction report, except in the latter
>> case, users are given a note that links them to the report options.
>It
>> would seem to me that at the least, this report should do the same. A
>> better solution in both cases would be to have initial parameters
>that
>> return some kind of data--however unrelated to the user's end
>> goal--rather than an empty page. It seems to me that users are better
>at
>> changing something they see that is wrong, rather than trying to
>conjure
>> it up out of thin air. (the famous quote from US Supreme Court
>Justice
>> Potter Stewart regarding obscenity comes to mind here: "I know it
>when I
>> see it. This is not it."). In the Balance Sheet, changing the Period
>> Duration setting from its default "None" to anything else (I'd
>suggest
>> "Year") yields data from which a user can build. (I have no
>suggestions
>> for the Transaction Report.)
>>
>
>This would be worthy of an individual bug report .
>
>
>> 2. The report defaults to including all accounts, including hidden
>and
>> placeholder accounts. While there is an option on the Accounts tab to
>> display hidden accounts, I think the decision to include hidden
>accounts
>> in the report by default is confusing, given that the Account tab
>> doesn't display hidden accounts by default. I think it makes more
>sense
>> for the report either to default to omitting hidden accounts, or to
>> default to displaying hidden accounts in the options dialog. I see
>that
>> the standard Balance Sheet report also includes hidden accounts by
>> default, although this is masked because the standard report limits
>the
>> accounts to 3 levels (the Experimental report defaults to All
>levels). I
>> will note that there is an option on the Display tab to omit zero
>> balance accounts, which may have been seen as a way to skip hidden
>> accounts. However, while it may not be prudent or common practice to
>> hide accounts that have balances, there is nothing in the program to
>> prevent this, which means that hiding zero balance accounts is not
>the
>> same as including hidden accounts. I'll note that if I Clear All and
>> then click Select All on the Accounts tab, the report properly omits
>all
>> hidden accounts, which it should also do by default. I suggest that
>both
>> reports should default to omitting Hidden accounts on opening, and
>that
>> the default level setting should be the same for both.
>>
>
>Also worthy of a bug report. The rationale IMV is that hidden accounts
>can
>carry balances and if they are skipped during report generation, they
>will
>cause undue mismatch between children balances and total parent
>balances.
>
>
>>
>> 3. The report propagates balances all the way up the account
>hierarchy,
>> which makes sense for accounts denominated in a currency. However,
>with
>> accounts denominated in commodities (such as Stock or Mutual Fund
>> accounts), it makes less sense, and in fact renders the report
>extremely
>> difficult to follow, especially with accounts with many different
>> commodities. Seeing a full listing of all 150 different commodities
>(and
>> various subsets thereof) repeated for
>>
>> Assets,
>> Assets:Investments,
>> Assets:Investments:Taxed,
>> Assets:Investments:Taxed:Self, and
>> Assets:Investments:Taxable:Self:Brokerage A
>>
>> gets old and confusing very quickly. It would be better to propagate
>> totals for the book currency  only, and leave the tallying of stock
>and
>> mutual fund holdings to the Advanced Portfolio report. I will note
>that
>> initially, I was going to ask why some entries in the report were
>listed
>> as hyperlinks, while others were not. After some puzzling, I figured
>out
>> that the non-linked entries were summarizations of holdings further
>down
>> the hierarchy. Removing the numerous duplications of commodity
>> summmarizations would at least make it a little clearer what is going
>on
>> in this regard. I will further note that the standard report does not
>> repeat this information up the hierarchy.
>>
>
>This multiplication of commodities only takes place if the report does
>not
>convert to target currency. e.g. if you have US tech stocks and set
>conversion to USD, it will attempt to convert all your AAPL MSFT TSLA
>to
>USD. If conversion is impossible (eg you have TSLA but there's no 
>TSLA/USD
>price) then it will leave TSLA as is and show TSLA + cash+stocks USD.
>For
>that matter, hiding stocks from propagating up the chain isn't a good
>idea
>IMHO.
>
>Moreover, the standard balance-sheet or profit-and-loss report is *not*
>good in forcibly converting all commodities & currencies into target
>currency -- if the conversion cannot take place (e.g. missing pricedb
>entries) the original currency amount becomes zero target currency.
>
>IMHO etc


More information about the gnucash-devel mailing list