TMSU was born out of frustration with the hierarchical nature of filesystems

...what's frustrating about the hierarchical nature of filesystems?

Literally the only ways I can think of historically is mounts/links; which in the POSIX world have been the source of some gigantic piles of academic salt, but in day-to-day (read: relatively sane operation), I've never encountered a case where one can't speed things up for one self by leveraging said hierarchical structure expressively.

In fact, I'd anticipate any type of system that tries to hide paths behind tagging to eventually reduce to "a hierarchical filesystem, where the tag(s), (or combination thereof) represents a folder. It just provides a different mechanism of dimensionality, that I'd assert is more deleterious to the learnability of the entire thing. The desktop/document filing metaphor at least has a first order bootstrap. Drop a person ina situation with a bunch of uncategorized data and something to put it in, and I wager you'll get a hierarchical storage scheme emerging naturally more often than not. "Tagging" requires a db and someone to have already built the index.

It's not like it saves you from having to define where your bloody bits are in the first place.

I don't think the hn title (which I realize is from the link) is very helpful. Better would be the about from github: TMSU lets you tags your files and then access them through a nifty virtual filesystem from any other application.

Could be shortened to "TMSU: tag files and access them through a virtual filesystem from other apps".