Intersystems provides External language server connections for various languages which, from a development perspective, is great as I can keep my application code separate but still interact with the IRIS kernel.
It becomes more complex though if you are using Interoperability etc where you might end up with some code in IRIS and some code in another language.
What would be great is having an IRIS to IRIS language server where the application code and stay IRIS based but separate from the database?
I Have some analyzers connecting to a EnsLib.HL7.Service.TCPService.
I can see the client IP and port in the event log but I need access to that ip and port either in the business service or even better the associated business process.
I've been wondering about some code that I have come across a lot over the years.
Let's assume I have class Cinema and class Film. Conceptually the data in these classes are never really physically deleted but only flagged as such, due to business requirements.
What I find is that developers tend to create a 3rd class "CinemaFilms", in a child relationship to parent Cinema, with a reference to Film. Rather than a one-to-many between Cinema and Film.
MessagePack is a binary-based efficient object serialization library. It enables to exchange structured objects between many languages like JSON. But unlike JSON, it is very fast and small.