Easy RDF For Real-Life System Monitoring – Thomas Passin
Combining Enterprise System Modeling and RDF.
Wow, crazy diagram…shows the connections between RDF and EA. Doesn’t look like a lot of connections from this diagram. But I can’t tell how deep the items that do overlap are.
RDF can handle the evolution of data == scope creep 😉
Data collection must occur before you can create the structure…mostly everything is a one-to-many or many-to-many relationship which means a constant restructuring of data. In the old days this was a database. But if the data is hierarchical then a relational database isn’t the best choice. Especially if you are using duplicates….and might need to be a deeper hierarchy (maybe you thought it was only a year, but in the end it needs more information about the year)….not easy to accomplish in database.
People like lists…graphics are fine but not for large-ish amounts of data. He suggests a text-based structure that is ‘simpler’ than XML….basically it looks like a tag followed by a ‘::’ followed by the content. It doesn’t seem simpler to me…basically it takes out the pointy-brackets and the end tags. Very python-ish. Which might be the impetus behind it because he also wrote a parser for the data in python. I have nothing against Python…I am slowly learning it myself…but do we really need another language structure? I thought the point of XML was that we’d make this standard…that we would understand the basics underlying the language…and what we’d have to learn is new words, not new punctuation. I understand that RDF has lots of options…but do you really have to use all of them?
I think I missed something when I was typing…this live blogging thing is difficult. I’ll be linking the papers to the post so that’ll help me get back to re-read where I lost track.
Overall I see a structure that has no tools to work with it that doesn’t seem to have any advantages over XML (RDF specifically). His last slide seems to be about converting the model to other formats….which it appears is necessary at this point…so maybe this is just a format for understanding the data and looking at data in a different way.
Sid Bowman – Brown University: The markup isn’t as explicit as the markup under the hood. He’s wondering why this looks so much like python and that the explicit ends are lost (wow…something I thought earlier). Thomas says that the content after the colons is run on (I assume until the hard return at the end). He supports the block structure and indentation
Fabio Viali – University of Italia: Compare this approach to the other RDF variants. Thomas says that he was focusing on making this easy for normal people…the indentation matches the way people make lists. RDFa wasn’t an option at the time.