Re: [Ledger-smb-devel] Question on Addresses and Not Null constraints

2008-05-22 Thread Chris Travers
On Thu, May 22, 2008 at 5:23 AM, Roderick A. Anderson <[EMAIL PROTECTED]> wrote: > > Since I'm just getting started with LedgerSMB (after years away from > SQL-Ledger) I'm not sure what the installation base is nor what is it > going to be. > > If there is a need to send /reliably/ via. a postal s

Re: [Ledger-smb-devel] Question on Addresses and Not Null constraints

2008-05-22 Thread Roderick A. Anderson
Chris Travers wrote: > One thing that is remarkably hard to model relationally is address > data. In particular different places have different semantics. Some > countries have mail codes while some do not. Some places it makes > sense to track states and provinces while in some countries this >

Re: [Ledger-smb-devel] Question on Addresses and Not Null constraints

2008-05-22 Thread Jeff Kowalczyk
On Wed, 21 May 2008 19:00:40 -0700, Chris Travers wrote: > I suppose we could just drop the NOT NULL constraint from state... Is the only reason state is marked NOT NULL is to reference a tax table? If so, perhaps we should have a separate tax_zone, or table matching entity_id to any number of ta

Re: [Ledger-smb-devel] Question on Addresses and Not Null constraints

2008-05-21 Thread Chris Travers
On Wed, May 21, 2008 at 6:13 PM, Jeff Kowalczyk <[EMAIL PROTECTED]> wrote: > On Wed, 21 May 2008 11:44:05 -0700, Chris Travers wrote: >> One thing that is remarkably hard to model relationally is address data. >> In particular different places have different semantics. Some countries >> have mail

Re: [Ledger-smb-devel] Question on Addresses and Not Null constraints

2008-05-21 Thread Jeff Kowalczyk
On Wed, 21 May 2008 11:44:05 -0700, Chris Travers wrote: > One thing that is remarkably hard to model relationally is address data. > In particular different places have different semantics. Some countries > have mail codes while some do not. Some places it makes sense to track > states and prov