The Family Grid

A concept that combines parents & children, column filtering, aggregation and table joining, allowing the user to freely manipulate and analyze data

Data is stored in Openbravo in a header-line relationship, also known as a parent-child relationship. For example a business partner who has two bank accounts and two locations. This data model consistently separates parent and child records which makes it suitable for 1:n relationships where there are few parent records and many child records. It will be less suitable for 1:1 or 1:n relationships where n is small. The user can only work in one view at a time and needs to switch back and forth between parent and children frequently.

The next generation of Openbravo ERP aims to offer a more flexible master-detail GUI where parent & child grids or forms can be shown in the same view. This allows for much faster record creation, editing, comparing and searching. Besides that, different views can be opened on separate tabs, offering multi-tasking.

Reports in Openbravo - and most other ERPs - are produced as one-off documents. They have a pre-defined set of dimensions, layout and data grouping. This works fine in most of the cases but in some cases more flexibility is desired.

Pre-defined views make it easy for the user to create out-of-the-box reports and to work on standard tasks but they do not allow the user to freely experiment with the data. Now imagine a grid view where:

• Parent and child information is combined


• Both attribute values and records can be shown


• Column values can be aggregated (e.g. sum, count, average)


• Complex nested boolean searches across parent and multiple children can be performed on user level by just using column filters


• Grids can be joined


• The user can experiment with data in a flexible way. Questions such as "Give me the sum of all unpaid invoices of over 1000 Euros of customers in the Czech Republic that have an HSBC bank account" can easily be answered by manipulation of a grid


• Most reports can be replaced by this new grid view

The concept that is presented below aims to simplify data by summarizing it, comparable to pivot tables and on the other hand aims to simplify multi-dimensional analytical queries, comparable to on-line analytical processing (OLAP). In the hands of a savvy user, this concept can be a powerful yet easy tool for do-it-yourself business intelligence. For implementation in Openbravo I can imagine that this is an additional view. Not all record types will need it.

I have to admit that I have no idea about potential performance issues. For now I think it is good first to look at whether this idea can make our users more productive, effective and better informed and then assess how to realize it.

This movie presents the idea.

Let me hear your thoughts on the UX Lab forum.