I guess I don’t quite understand this talk of the semantic web after having read Whatever Happened to the Semantic Web and the Hacker News Discussion about it. I suppose the hardest thing for me to wrap my mind around is just what rdf LOOKS like. I know it’s described as a “triple” of information in this RDF wiki page, that names the relationship between things as well as the "two ends of the link".
And there are practices and treatises and tools for collecting/forming this information like the Web Ontology Language. And I also understand it never got off the ground because- I suppose “tagging” is the complete opposite of a “triple” it’s a “double” and it is pervasive in the way we connect information on the web. I suppose a double is more a immediately efficient way to store and link information and it's going to be hard to deviate from that practice now. At least that's what it sounds like people are saying. I'm not sure why people keep debating HTML and the importance of DIV tags.
I don't really have an opinion on it- I'd say it seems easy enough to add extra links of information and create a tighter "web" of associations that can be more multi-purposes than what we currently have. I suppose when some awesome technology becomes available that motivates everyone to start pushing more for semantic web practices, then maybe this will take off. Like if they make more advanced AI and have home-helping robots that need to make these associations and we are making more humanoid androids yeah, then I could see us trying to cram all of these triples, these RDF structures, in to get the thing working more intelligently.
I read Death By Database and watched the video which gave a good general overview of how I’d organize joins and tables. Some “what not to do’s…” Specifically how to link information in a way that stops data from being compromised using foreign key constraints with relational tables.
Lists get their own table, nouns get their own tables… but you need to put a table in-between them – called a ‘many to many’ relationship. Another tip for keeping track of prices for things within a database- give the item a price in the item table, but also have a specific table for “prices” that is updated and shows the “actual” price
So if you’re moving something from an “abstract” idea to a “concrete” this is what it is now, you’ll need to have another table to represent the “equal values that are not identical”
A database is your first line of defense against “bad data”. Also encouraged naming columns as descriptively as possible. The hackernews discussion about it was not very enlightening however. I guess I wonder where all these people came from who care about such things. Tedious.
I read One Small Step for the Web about soLID and read the Wiki about it. Did not try running my own solid server, I guess I have questions as to what it really does/how it manages to bundle and protect one’s data and how one would use one's own data in transactions using a “pod” …
