Explainer
What Is a CSDB? The Common Source Database, Explained
The Definition
One Source, Managed
The Common Source Database is the managed repository at the center of an S1000D program. It holds the program's content as uniquely identified objects: data modules addressed by their Data Module Codes, illustrations with their own identifiers and revision lives, publication modules that define what each deliverable draws in, and the planning data — the DMRL — that says what should exist at all. “Common source” is the load-bearing phrase: one authoritative copy of every unit of content, referenced by every publication that needs it, never duplicated into them.
What makes it more than storage is what it does on the way in and the way out. On the way in: identity (no two modules with one code), versioning, workflow state, and — in a well-run program — validation against the business rules, so BREX violations die at authoring rather than at delivery review. On the way out: publication assembly, applicability filtering per configuration, and exchange with partners in the conventions the specification defines. A folder of XML files can hold the same bytes; it can do none of those things unattended.
In the wider content world the CSDB is a component content management system with the S1000D machinery built in — which is why teams from the DITA world recognize it instantly, and why the reverse migration (a generic CCMS pressed into CSDB service) is harder than it looks.
In Practice
An Operational System, Not a License
The most common CSDB failure has nothing to do with the software: it is a program buying one and treating it as a shelf. The CSDB is the single source for a decade or more of sustainment content — it needs a named owner, a backup regime that has been tested, a defined workflow that matches how the program actually reviews and releases, and the BREX wired into the authoring environment so the rules are enforced where content is made. Those four items appear, not coincidentally, on our program start checklist.
One boundary keeps expectations honest: the CSDB is the authoring-side system. Getting published content to the people who use it — filtered, current, and provable — is the delivery problem, and it has its own evaluation criteria. Programs that expect one system to be excellent at both should at least demand proof twice.
FAQ
Questions We Hear
What does CSDB stand for?
Common Source Database — the managed repository at the center of an S1000D program, holding every data module, illustration, and publication definition as uniquely identified, versioned objects. "Common source" is the point: one authoritative copy of each unit of content, from which every publication is assembled.
Is a CSDB required for S1000D?
The specification is built around one — content is authored, managed, and exchanged as data modules with the CSDB as their home — and in practice every serious program runs one. What the specification does not do is name a product: a CSDB is a role, filled by commercial systems of very different depth, which is why 'we have a CSDB' tells you less than 'here is our workflow, our backup regime, and our BREX enforcement.'
What is the difference between a CSDB and a CCMS?
A CSDB is a component content management system specialized for S1000D: its components are data modules, its identifiers are DMCs, and its rules come from the specification and the program's business rules. Every CSDB is a CCMS; a generic CCMS becomes a CSDB only by implementing what S1000D expects — applicability, BREX validation, publication module handling — which general-purpose systems don't ship with.
Can we run S1000D from a file system instead?
For a training exercise or a tiny pilot, files and discipline can imitate one. The imitation fails exactly where programs get hard: enforcing business rules at authoring, tracking which module versions went into which delivered publication, filtering by applicability, and exchanging data with partners who expect CSDB conventions. If the program is real, budget for the real thing.
Get In Touch
Standing One Up?
Selecting, implementing, or rescuing a CSDB is work we do with our hands, not just our slides. Tell us where your program is.