postgres backend ???

[Matthew Vanecek]

On Mon, 2003-01-27 at 09:31, Linas Vepstas wrote:
> On Sun, Jan 26, 2003 at 10:50:13PM -0600, Matthew Vanecek was heard to remark:
> > 
> > Accounting uses a very distinct data model. The current PG backend
> > loosely follows that data model, and the rewrite will be even closer. 
> 
> I'm not sure of what the antecedent is.
> 
> When I wrote the pg backend, I made it 'more' correct than the engine,
> and was planning on changing the engine to make it resemble the pg
> backend.  So I'm concerned about the 'even closer' remark.
> 

Don't be too concerned.  The basic data model is very close, but it
needed to be more extensibly designed.  I'm attempting to make the new
data model more portable to other systems, too (e.g., where inheritance
isn't supported).

> On the other hand:
> -- the business modules make new demands, and I don't understand these.

Derek does, and we've talked a bit about it.

> -- neither lots nor books are supported.  Books, partly so, I think.
>    We'll need these.

Derek and I have also talked about Lots, and more recently Books.  Also,
Scheduled Transactions.  We're going for a modular approach, whereby the
business module registers its support of specific types, and other
modules can do the same. Certain stuff will be hardwired in, but still
follow the module API.

It's coming together pretty well.  Just trying to bug-fix a little in
the existing system while I work on the rewrite.

-- 
Matthew Vanecek
perl -e 'print $i=pack(c5,(41*2),sqrt(7056),(unpack(c,H)-2),oct(115),10);'
********************************************************************************
For 93 million miles, there is nothing between the sun and my shadow except me.
I'm always getting in the way of something...
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/defanged-87410
Size: 232 bytes
Desc: not available
Url : /pipermail/attachments/20030127/16ffb7a4/attachment.bin