La divitis es una enfermedad antigua, el primer artículo que conozco que habló sobre esta enfermedad es del año 2006: Divitis: what it is, and how to cure it
Ahora he encontrado otro artículo, Stop using so many divs! An intro to semantic HTML, que explica los siguientes problemas de la divitis:
- Accessibility – Many a11y tools are pretty smart, and try their best to parse the structure of a page to help guide users through it in the way the page’s author intends, and to give users easy jump points to navigate quickly to the section of the page they care about. But <div>s don’t really impart any useful info about the structure of a document. The smartest a11y tool in the world still ins’t a human, and can’t be expected to parse class and id attributes and recognize all the weird and wild ways that devs all over the world name their blocks. I can recognize that class=”article-header-level-2″ is a subheading, but a robot can’t. (And if it can, get it out of my computer, I’m not ready for the AGI revolution just yet.)
- Readability – To read this code, you need to carefully scan for the class names, picking them out from between the <div class=”…”></div> boilerplate. And once you’re a few levels deep in the markup, it becomes tricky to keep track of which </div> closing tags go with which <div…> opening tags. You start to rely very heavily on IDE features like coloring different indentation levels or highlighting the matching tag for you to keep track of where you are, and in larger documents it can require a lot of scrolling on top of those features.
- Consistency and standards – It can be frustrating to start a new job or move to a new project and have to learn from scratch all the crazy markup conventions used across the codebase. If everyone had a standardized way to mark up common structures in web documents, it would be much easier to skim an HTML file in an unfamiliar codebase and get a quick handle on what it’s supposed to represent. If only there was such a standard…