Reference dates for net worth barchart

[Wm...]

Thu, 5 Nov 2015 18:26:22 <1883999.Jh6Ha8s1L7 at legolas.kobaltwit.lan>
Geert Janssens <geert.gnucash at kobaltwit.be> wrote...

>The part of the patch I'm interested in here now is the part where you 
>changed code to unconditionally use the start-day-time for the 
>from-date, where it was using either start-day- time or end-day-time 
>depending on whether the report is for income/expense or not. Your 
>patch may be correct, but I don't know because I have no idea why this 
>distinction was there in the first place. I would like to hear if other 
>users/devs know this before I apply your patch.

To state the obvious:

Asset, Liability and Equity accounts are usually reported on at a point 
in time (nominally a moment before midnight, I don't think we need to 
involve time zones here).

Income and Expense accounts are usually reported on over a period in 
time, nominally moment after midnight to moment before midnight.

The Net Worth Barchart is (or should be) balance sheet type reporting, 
in plain terms the sum of a set of accounts (or a single balance sheet 
account) of the users choice at a point in time.  And some other point 
in time.  And another one.  And so on.  But, to be clear, points in time 
for Balance Sheet accounts represented by a bar on a chart.

If my argument holds true the table that goes along with the Net Worth 
Barchart will match balance sheets produced for the same time.

Time passes, I've done some spot checks on weekly data from accounts I'm 
familiar with that change at least weekly and it holds true.

Does the graph match the table?  To test this I hovered my cursor over 
bars in the chart and checked if they matched (date and amount) with the 
table.  Answer: yes, they match.

I'm not seeing a reason for change in terms of what is displayed.

>  So does anyone know a good reason to use another start time for 
>income-expense reporting vs other reports ?

Ummm, if the report was about a period of time (Income Statement or 
Profit and Loss type reports) rather than a point in time then this 
might be something to discuss.  As far as I can tell the Net Worth 
Barchart is a series of Balance Sheet reports.

>It has been like this since 2001 at least when the report was first 
>introduced as a merger between "income-expense-graph.scm" and 
>"net-worth-timeseries.scm" [3] by Christian Stimming. The former used 
>day-start-time, the latter used day-end-time.

Before my time but if appropriate moments are picked there needn't be a 
problem per se.

===

Having said all that I'm wondering what the issue for JD is.  I think 
(but may be wrong) that his moan is about arbitrary human *time periods* 
not always matching up.

We should all know that weeks, months and years can be weird if you are 
dumb enough to believe in a god or follow a particular government and 
prefer their definition of time versus one other humans understand.

Locally, if I use "Step size: week" I get 7 days per step.

Locally if I use "Step size: month" I get a varying number of days.

Why? Because the fuckers that made up months in the centuries before 
anyone reading this was born made months arbitrary numbers of days long 
depending on whether they'd recently killed someone, had sex or won a 
war or similar stuff.  To put this in context we didn't change the 
months of the year for WW1 or WW2 when lots of people died more 
recently.

So, easy way out?  Use the first of the month not the last day as your 
accounting day and if you are a slave use the day before that.

gnc does months well and you get to poke officialdom in the eye because 
you are doing it right  -  there cannot be a more correct measure of a 
month than from 1st of one to 1st of the next, surely, even though the 
number of days vary? :)

Months don't even make sense as investing periods, Feb non leap year vs 
a 31 day month is silly.  Months that include holidays vs months that 
don't, similarly silly.

JD, go for weeks is my advice.

I'm going to need persuading that gnc should accommodate all varying 
length months and strongly suggest it shouldn't even attempt the 
variances because that way is the path to hell [1]

[1] in the sense that as soon as you agree one weird calendar their 
opponents will say you have to include theirs too and before you know it 
the dumb shits will start killing each other and it will all be our 
fault.

>Geert
>
>[1] https://bugzilla.gnome.org/show_bug.cgi?id=757637
>[2] https://bugzilla.gnome.org/show_bug.cgi?id=639049
>[3] https://github.com/Gnucash/gnucash/commit/e8a5a9677cfb0a1eaa7

Please don't block this, Liz.  Mildly intelligent people should read it.

-- 
Wm...