On Tue, May 15, 2012 at 6:44 AM, Bradley M. Kuhn <[email protected]> wrote:
> I said "accounting-related" in my original email; it's related to
> accounting.  Basically, I'm looking for the toolset that I could handle
> a non-technically-sophisticated bookkeeper and tell him/her to do the
> job, and I could check everything myself by looking in the version
> control repository and see what was done.

Ah... well, my take on that is make scripts to reduce the amount of
work done by a bookkeeper.

One example - I have a OCR/regex contraption that auto-categorizes
receipts ala commercial prodducts like NeatReceipts with minimal human
intervention, and spits out XML (and soon, ledger).

>> It's a document generation task that also requires a database to track
>> and manage documents, that directly affects the accounts in
>> Ledger.

I'm thinking that could be handled with ledger's metadata.
Fundamentally you need a superset of data per document that includes
things that aren't pertinent to Ledger.

>> Storing Ledger data is nice, how do you intend to store the data about
>> the invoices? Just a practical consideration.

In text.  In files that contain things like "work performed", lists of
items sold (I'm toying with ideas of a CLI driven inventory system),
and similar that are a superset of my ledger data.   I just generate
ledger from those, and use it to balance the accounts, but ledger
isn't the center of the world in my use case.

If you want an idea of what my workflow looks like, I did a version
control specific presentation at the MacTech Conference last year, in
which I show using git, ledger, and the rest of my toolset - the .mov
is probably the most interesting bit to this list:

https://github.com/zdw/mtc2011

- Zack

Reply via email to