DITA without DITA

---

19 February 2024

The value proposition of DITA hinges around these three things, mainly:

• Enforcing a structure automatically, through strict topic typing

• Reusing content at the topic, block, and element level, by means of direct references or keys

• Generating profile-targetted content, by using attributes that can filter content at the topic, block, and element level

Are these things possible without DITA?

Assuming a docs-as-code pipeline where the content is written in plain text and transformed through a static site generator (SSG), here’s a comparison table for two SSGs that I am familiar with:

DITA feature	Sphinx	MkDocs
Enforcing a structure	No	No
Reusing content	Yes, through variables and includes	Yes, through variables and includes
Generating filtered content	Yes, through the tag object in conf.py together with the include directive in content, and -tflag in makefile (but the implementation isn’t documented well and is buggy)	Yes, through the extra key in mkdocs.yml together with if blocks in content

If this table had had columns for other SSGs too, I suspect that the row values would’ve been similar.

If strict enforcement of a structure is not a requirement, DITA is not needed. One can use the good practices of DITA even outside a DITA environment, even inside a docs-as-code environment.

When is going the DITA way a good option? Let me count the situations.

---