GnuCash-2.0 first impressions

[Ganesan Rajagopal]

>>>>> Neil Williams <linux at codehelp.co.uk> writes:

> Certainly in the short term, yes. I am still hoping to implement the kind of 
> process that an "ordered" collection of objects may require. Such a method 
> has uses beyond the sqlite backend for gnucash.

>> I think we really need something more along the lines of the 'cash
>> objects' XML backend, except for SQL.  Honestly, I think we DO need a
>> Gnucash-specific SQL backend impementation.

> OK. The embedded code will still come in handy in certain areas.

Sorry if I am a bit dense today, what's this "embedded code"?

>> But yes, I would certainly encourage you to come up with a good
>> extensible design for a SQL backend and, once you get the design done
>> (and signed off on -devel) go ahead and start working on it.

Would the existing PostgreSQL backend be a good place to start? The GnuCash
Wiki talks about a Direct SQLite backend and a extensible backend. Do we
need both? I saw some discussions in the gnucash-devel archives about libdbi
and GNOME's libgda. I didn't see the discussion conclude in favor of either
one. Conceptually I find libdbi much simpler. 

Ganesan

-- 
Ganesan Rajagopal (rganesan at debian.org) | GPG Key: 1024D/5D8C12EA
Web: http://employees.org/~rganesan        | http://rganesan.blogspot.com