Thoughts and write-ups on integration architecture and security, written over the years and occasionally revisited as the technology underneath them changes - from SOAP web-services to today’s API-first world. Below is a short summary of every post, newest first, each with the date it was published or last generated.
For practical, step-by-step playbooks from my homelab, see the How To page.
I managed to get my hands on an original Tandy TRS-80 Model I - the machine that first taught me to code, almost 50 years after I bought it as a 16-year-old.
The problem with AI is not that it stops us to produce Open Source projects. The problem is that we develop a new elite with AI enabled specialists and the people best suited to shape the use of AI refuse the do that. Ignoring it not the answer, AI is not going to take note. This way the knowledge and the tools are in the hand of some few organizations, only accessible to the people if they can pay for it. This makes AI now very undemocratic, but the European people could change this. By building their own infrastructure, their own knowledge base, and their own competence. And this way to make the technology available to the normal user. And also to control the business model of the AI providers. But if the Open Soruce Community is only rejecting the use rather than to provide the guidance, it will not change.
This article is an updated version of the 2010 post Integration and Security, rewritten to reflect how integration has moved from SOAP web-services to public, internet-facing APIs. The update, and the research behind it, was produced with the assistance of an AI tool (Claude); all sources are linked inline so the underlying facts can be checked directly.
Why firewalls and DMZs alone cannot secure a service-oriented integration landscape - lessons from the Great Wall of China, applied to SOAP web-services.
What actually makes a service re-useable in a service-oriented architecture - it isn’t the interface technology, it’s designing CRUDS-based basic services and composite business processes on top of a well-defined canonical data model.
The opinions and statements expressed in these blog posts are my own private views and do not represent the positions of my previous or current employers. All posts may be freely quoted and copied, provided the original author is credited and a link back to the original post is included.
