I realize I've been asking you a lot of question lately and will probably get throttled which is fair enough :). But if you've got the time and feel like it: Say some products require reservation on existing stock but some don't. 1/7
So, an order must be accepted or not based on the specific product and stock levels. There's a whole domain around orders and another around stock. They naturally form boundaries for two services: Orders & Stock. Yet the last stage of accepting an order takes input from both. 2/7
It's more complex in fact, as an Order with a value larger than X must be flagged for human review by an Orders Desk. Of course, product pricing does not fall within Orders but rather Sales due to there being a whole different domain around pricing. 3/7
So, an order can be rejected (due to Stock rules) or flagged for review before accepting it (due to it's value derived from Sales prices). Anyway, once accepted, Orders continues processing the order. 4/7
The options I see: 1. Orders is aware of the OrderAccepted and OrderIsLarge (or similar) events published by Stock and Sales respectively. 2. Orders invokes some kind of engine which gets a decision (not raw biz data) from Sales and Stock it uses for subsequent processing. 5/7
Both options have pros and cons so I'm wondering what would you recommend. More biz. analysis can always be done but I'm really interested in dealing with such cross-cutting decisions as I'm sure I'll find one which I cannot model away. 6/7
It's a case of a biz process starting in a service, spanning others, and returning in the initial service. In many cases I see how it can be modeled away with deeper domain insight but I've got something where it doesn't seem the case. Thanks and much appreciated, as always! 7/7
