This sounds vastly more workable, on a philosophical/ethical level, and is something that I could get behind as long as there was a lot of work on the UI level to make it usable/intuitive. Actually, I could get behind it even if it's not usable/intuitive because large chunks of the Archive already aren't. (Which frankly terrifies me considering how much MORE usable than many other platforms it is. But I digress.)
Because, yeah, I found your previous post very interesting but had nothing like the energy to engage in discussion/debate about why while I could see your very valid performance concerns, I was twitching like hell at the imposition of tag formatting on users.
I will admit I do still wince a smidge and I'm not entirely keen on the idea of wranglers being able to directly affect which tags go on a work even when it's a matter of new tags/canonicals, but this seems like something that could work scalably for the kind of enterprise level platform the AO3 already is/needs to be/will grow into, without involving any kind of leaps of genius. It seems... possible? Practical? And actually in some ways this method could give you significantly MORE control than you currently have over tag formatting. (I'm... bitter about the capitalisation issues I've encountered, perhaps.)
Relatedly, I do think a seperate metadata field for tags that are not meant to be searchable (ie, tumblr style tags) would also go a hell of a way towards lightening the load on both servers and wranglers alike. Those tags are often/usually long and very specific to a given work and it's just absurd to fold them in with the other folksonomy genre/flavour/additional tags even if I'm entirely happy for them NOT to be folded in. (Or in some cases, might prefer them not.)
no subject
Because, yeah, I found your previous post very interesting but had nothing like the energy to engage in discussion/debate about why while I could see your very valid performance concerns, I was twitching like hell at the imposition of tag formatting on users.
I will admit I do still wince a smidge and I'm not entirely keen on the idea of wranglers being able to directly affect which tags go on a work even when it's a matter of new tags/canonicals, but this seems like something that could work scalably for the kind of enterprise level platform the AO3 already is/needs to be/will grow into, without involving any kind of leaps of genius. It seems... possible? Practical? And actually in some ways this method could give you significantly MORE control than you currently have over tag formatting. (I'm... bitter about the capitalisation issues I've encountered, perhaps.)
Relatedly, I do think a seperate metadata field for tags that are not meant to be searchable (ie, tumblr style tags) would also go a hell of a way towards lightening the load on both servers and wranglers alike. Those tags are often/usually long and very specific to a given work and it's just absurd to fold them in with the other folksonomy genre/flavour/additional tags even if I'm entirely happy for them NOT to be folded in. (Or in some cases, might prefer them not.)