A game of “improve the attention format”
“Now, let’s play a game, try to come up with a better alternative and then we keep improving it.”
I’ll happily play along – in fact I have a format brewing in the back of my head.
Right now thought, all I have time for is a couple of quick comments:
- First, it’s difficult to tell what’s really going on without some knowledge of the metadata surrounding the attributes. Like, which are required and which are optional?
- Second, for my part I’m not sure I want to stuff all the data in attributes. My understanding of XML best practices is to keep metadata in the attributes and data in elements.
- Last (for now) I’m not partial to the parent-cild relationship here. It mean that every time I want to record an item I need to record a feed. But not all of my attention items are in feeds, and I’d also rather just reference a feed for an item when there is a feed (instead of constructing that feed node again), and keep the meta-data about the feed separate. It’ll cut down on the bits going over the wire quite a bit.
Now that’s all negative stuff. Here’s a positive one: I like a lot of the data in this format – it’s closer to what I actually want to capture. One innovative element is the link child of item – that would be a nice bit to capture, and in cases where I’m too lazy to vote for an item an application could use an algorithm to vote for me (like if I clicked on it, vote for it – it would then be my responsibility to vote against stuff I clicked but didn’t like).
Counter-proposal, perhaps tomorrow.
technorati tags: attention data
RSS 2