Is there a drawback to using non-semantic HTML for nodes?


For example, if I want to use div and span elements instead of h1, p, and ul/li tags?

The reason for this is that they allow more powerful nesting behaviors for commenting and other functionality. But not sure if there’d be any drawbacks such as with SEO, etc.


As for SEO: as long as what you do can be reproduced by means of semantic HTML, you could always have a separate serializer when creating the final version that you publish to the web. If you do that, you can for example to leave out marks or represent them in the form of attributes rather than extra spans, etc. . search engines should only ever get to see what is being saved to a file, not what is being displayed in the dom during an editing session, right?

(that being said, I don’t know if li-elements make any difference to SEO).


The main drawback is that you’ll have to define a separate serializer/parser for the clipboard, because when copying/pasting/dragging between the editor an other applications, your non-semantic tag soup won’t help.