While writing Elli (Erlang Web Server) [Lessons in Semantic Interoperability – Part 1], I got distracted by the realization that web servers produce semantically interoperable content every day. Lots of it. For hundreds of millions of users.
My question: What makes the semantics of HTML different?
The first characteristic that came to mind was simplicity. Unlike some markup languages, ;-), HTML did not have to await the creation of WYSIWYG editors to catch on. In part I suspect because after a few minutes with it, most users (not all), could begin to author HTML documents.
Think about the last time you learned something new. What is the one thing that brings closure to the learning experience?
Feedback, knowing if your attempt at an answer is right or wrong. If right, you will attempt the same solution under similar circumstances in the future. If wrong, you will try again (hopefully).
When HTML appeared, so did primitive (in today’s terms) web browsers.
Any user learning HTML could get immediate feedback on their HTML authoring efforts.
Not:
- After installing additional validation software
- After debugging complex syntax or configurations
- After millions of other users do the same thing
- After new software appears to take advantage of it
Immediate feedback means just that immediate feedback.
The second characteristic is immediate feedback.
You can argue that such feedback was an environmental factor and not a characteristic of HTML proper.
Possibly, possibly but if such a distinction is possible and meaningful, how does it help with the design/implementation of the next successful semantic interoperability language?
I would argue by whatever means, any successful semantic interoperability language is going to include immediate feedback, however you classify it.