fix: honor delta from query param

[dnplkndll]

memory error on our SB on the export_salesorders call

[jdetaeye]

Needs updating before this can be merged.

• Ading the delta argument in frepplexml.py is a valid correction
• Using the write date is the correct way to assure we correctly pass any sales order changes to frepple. Using the order date isn't ok for me.
• Odoo skips archived records by default. Why is an extra filter needed?

[dnplkndll]

I am not sure what but in our case. -- count sale_order_line with write: 1412880 create:151940
so memory error @ limit_memory_hard = 6442450944
any interest in a delta_date parameter? perhaps could refactor in general.

great was not aware.

only a few days into this project. so I will clean this up and try increasing the limit but eventually will not want to pay gcp for resources just for a dump

has there been any discussion on more efficient, perhaps sql direct, extractions for large databases?

[jdetaeye]

I am not sure what but in our case. -- count sale_order_line with write: 1412880 create:151940
so memory error @ limit_memory_hard = 6442450944
any interest in a delta_date parameter? perhaps could refactor in general.

Nope.

has there been any discussion on more efficient, perhaps sql direct, extractions for large databases?

We are using the connector with pretty large odoo instances. It synchronizes the data in 5 minutes or so.

For an initial load of a large historical sales data, you can extract it directly from odoo and upload that directly in frepple. After that initial load you load the incremental changes of the sales order data with the delta parameter.

[dnplkndll]

proposed correction to round out floats