Death Star architecture is everywhere. It’s a sign of success.
It’s when you’ve got many systems and they interact in many^2 ways leading to increased costs and risk to the business.
Technical polyglot and multiculturalism isn’t the problem.
The real problem is the invisible spaghetti of logical wires that link systems together.
Invisible because each individual wire is inside a programmer’s black box of code.
All devops and coders continuously string together myriad logical wires, like this:
response = request.get("system.bar/some/API")
It works fine on day one - but you can’t run a query and detect all the interactions in your business.
At the core of the OpenSPARKL project code drop is the SPARKL Sequencing Engine (sse).
The sse is a router with a difference. It operates over logical field-sets, not physical data packets.
Clear box workflows arise from this routing being applied over pure tree-structured data.
This data can of course be queried and transformed, for example to generate the canonical OpenSPARKL mix diagram. Or, to produce a list of the systems that interact in a given mix.
Zero Insertion Force
You keep using the same tools, technologies, languages and platforms you already know and love (or hate).
It becomes safer and simpler to introduce new technologies, too.
OpenSPARKL Project Aim
So, the aim of the OpenSPARKL project is to let people kill their own Death Star architectures.
We want to enable this by gradually introducing Clear Box integration to devops and business applications in complex, hybrid environments.
There’s so much already in the project - and with your help I hope a lot more will come!
Please contribute your experiences and opinions, raise issues on GitHub, and if you’re so much as considering contributing code then I’m very grateful!