A linear amount of data could be saved if we extricated text content from the HTML skeleton that contains it.
I wish Semantic Web had taken off. "Pages with styling" was suboptimal. Web apps are a such a weird evolutionary branch we've descended into that don't relate to documents.
Content instead should have fallen under a type of ontology: news item, blog post, technical reference, comment, status update, ... If we'd adopted such a markup grammar and styled around it, we could parse out meaning, have stronger links in the graph, and compress.
Semantic Web would have happened if commercial web didn't outpace it.
Once you start transforming a document you always risk discarding information that later will turn out to be useful. Besides, I bet images, videos and other non-text data dwarf the space required by HTML markup.
I wish Semantic Web had taken off. "Pages with styling" was suboptimal. Web apps are a such a weird evolutionary branch we've descended into that don't relate to documents.
Content instead should have fallen under a type of ontology: news item, blog post, technical reference, comment, status update, ... If we'd adopted such a markup grammar and styled around it, we could parse out meaning, have stronger links in the graph, and compress.
Semantic Web would have happened if commercial web didn't outpace it.