QIF import and customization

[Bill Gribble]

On Tue, 2001-10-30 at 07:55, Anthony W. Juckel wrote:
> I posted this message on gnucash-user a few days back, but not many 
> people seem to be biting. 

Well, there aren't that many people who know enough about gnucash to
bite :) 

> What it all boils down to is this:  I like extensibility.  It is the 
> main reason that I choose Emacs as my preferred editor, because (at 
> least back when I was choosing editors) emacs' customizability and 
> extensibility were far superior to vi.  I like being able to define my 
> own lisp functions to get run at any step of the way.

Great. I understand that desire, but Gnucash doesn't really support that
style of customization.  Not because anybody thinks it's a bad idea, but
because nobody has demonstrated the desire and willpower to implement
the needed changes.  

We have a 'hooks' mechanism, whereby the user can define functions that
get run at certain events, but there aren't very many hooks defined.  I
believe I saw a reply from Dave Peticolas explaining how to put Scheme
code into your config.user file.  Those are the principal pieces needed
to make gnucash customizable in the way you wish.  What remains is to
identify the places where we need to put user-accessible variables and
hooks, and (as you mention) to document those. 

Right now, development effort on Gnucash *is* focused almost entirely on
customizability, but customizability of a somewhat different type: the
ability to take the pieces that make up gnucash and put together
applications that have nothing to do with "personal finance" (e-commerce
systems, vertical applications of various kinds, command line tools to
manipulate your gnucash data, etc). 

I think it would be great if the kind of user-level customizability you
want also got added, but I don't see much likelihood of the core
developers working on it real soon.  The paid developers are working on
the kind of modularity issues I describe above and on the OpenCheckout
point-of-sale system (see the website, http://www.opencheckout.com); the
active unpaid developers generally have "pet projects" of their own.  

It seems like you are interested and capable of tackling this project
yourself.  I encourage you to do so!  Everybody would (I think) love to
see Gnucash become more Emacs-like in this way.  As a new developer, you
are also probably a good candidate to document the stuff that's
undocumented and confusing as you come across it. 

b.g.