These are the rules I’ve set for myself on here.
I brazenly reuse old urls and titles when I find that a newer essay seems to cover the same topic as an older one. This is a way to keep the archive from getting too redundant with overly similar posts, and to make sure that old URLs, old links, still point to a currently-canonical expression of the idea there.
Sort of a wiki/blog hybrid. There is a blog view and feed but the idea is also that I can evolve pages over time. Which is bad for the historical record but good for the dream that this li’l homepage will one day become good or at least better.
Because of this, I’ve been pretty happy with the decision to not put dates in the URLs. I don’t update the timestamps (which also repushes it to the feed and to Antenna) for typo fixes and updates but I can for rewrites.
One of the consequences of thinking “wiki-like” for a blog is how it lends itself to breaking out side-topics to their own pages so that I can link to them from other pages that need them. That’s also one reason I have such an unreasonable amount of posts; I can write one thing and while writing realize that “oh yeah, this part should really be its own post” and break it out to another, linking to it from not only the new post but from older posts that might’ve needed it.
Or as I’m writing a new post I can realize that I want to refer to something that’s only a small aside in an older post, so I have to go back and break it out. The refactoring mindset applied to prose.
Speaking of slugs, is it snake_case or hyphen-case? After realizing that I had been pretty inconsistent, I switched to my new policy which is: use hyphen-case if there’s one level, but if there’s two, use snake_case for the lower/tighter level. Like artist_name-song_name. This is an awful policy and not one I would’ve chosen if I had put more thought into it but it’s what gave me the least backwards breakage. Older posts also are not renamed, this isn’t a retroactive policy. It’s just a mess.