Budgeting prototype

[Derek Neighbors]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

marthter wrote:

| Maybe I'm psycho but I can think of lots of cases where the budget
could have to cut across the accounts heirarchy.  (This sort of goes
back to the 1-dimensional vs. 2-dimensional ideas in account naming.)
|
| Maybe my accounts tree goes
| Expenses:Medical:him
| Expenses:Medical:her
| Expenses:Transport:him
| Expenses:Transport:her
|
| But I want to try a budget based on the him/her distinction.


Generally in accounting principles you have a focus of some sort.  You
generally budget at a higher level than you spend (unless you are
unusually anal)  You would in this case set your accounts up how you
want to see them in relation to that focus.  If your focus is Him vs.
Her as a primary concern.  You probably need to revisit your chart of
accounts to add this emphasis.

Expenses:Him:Medical
Expenses:Him:Transport
Expenses:Her:Medical
Expenses:Her:Transport

Part of the problem is that the way the chart of accounts exists is
one giant hierarchy which falls down.  Most accounting systems let you
define several "elements" to a chart of accounts.  Then you append
those together to form "accounting strings".  Often this is more a
business level usage, but so is what you are proposing to do mapping
of budgets to expense accounts.

| Maybe *I'm* psycho and I have 100 Expense accounts but only to one
level of depth (Expenses:whatever).  And I want to budget 70% of my
money on "accounts that end in vowels" and 30% on "accounts that end
in consonants".  I mean it is an obtuse example, but why should the
design force a decision on the user that makes that impossible?

Why?  To a degree financial software is different.  It isn't a window
manager.  Letting people get "creative" with accounting isn't
necessarily a good thing.  For example maybe I'm odd and like to
change every other entry into a negative number eventhough I enter a
positive number.  Not exactly the best function for an accounting
system.  Just because you can do something, doesn't mean you should.

| Maybe I talk to myself and play my own devil's advocate.  I'm
Spartacus.  No I'm Spartacus.

It is good to have discussion.  My scope is writing budget software
for large governments dealing in billions of dollars.  All
appropriations (spending controls) are done at the fund level (our
focus) and have strong "commitment control" ties to them.  Analytical
forecasting and "what if's" are done out side the maintenance of the
budget.  There is a budget "preparation" and thtere is budget
"maintenance".  I look at what gnucash needs to a large degree is
basic prep (get numbers in, no drivers, no analytics) and strong
budget maintenance.  Plan performance so to speak.

| All four of us like Darin's direction on this point.

Certainly I am not offerring to code a lick of anything.  So if you
have willing and able coders and you all agree on something go for
it.  I am not here to pull the discussion down or shout nasties.  This
just happens to be what I do professionally for a day job so thought I
would interject some real world experience.

- -Derek Neighbors
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQE/WERBHb99+vQX/88RAqQ1AJ9lkes9MMsvrL1eev4FQEw8IzNO6ACfYSl1
XS3YJL4ilSeOKpoQfvdMM+k=
=CMdX
-----END PGP SIGNATURE-----