silver

OTW approves meta hosting on AO3 announcement

My first reaction was YAY!

But then I realized that there is already a really big problem that AO3 has no resolution to, and this decision is sure to aggravate it: the fact that the filtering system is not good, and adding meta to the mix will make it worse.

In addition to all sorts of analysis springing up on various journals, here's FFA's roundup of links and discussion: http://fail-fandomanon.livejournal.com/50237.html?thread=232275517#t232275517

I confess I love these discussions about AO3. There's something in the discussion of how one should manage a project/resources that really fascinates me - some of it's venting, but there are lots of interesting comments that talk about the lack of existing infrastructure and how that's a problem. It's a great read if you're interested in that sort of thing, and they've helped crystalize a lot of my thoughts.

One of the core problems I actually have with the announcement, right off the bat, is that they said yes, meta is allowed! but they're not ready. They are so unready that they do not have a definition of meta. I know that they want to be more transparent and let us know, but telling us that 'something is approved!' while being unsure of what 'something' is is not actually helpful.

Leaving aside the issue of whether meta should be included*, there are major problems in implementation. The underlying problem of categorization on AO3 is becoming increasingly apparent: the listing of fandom names is unbelievably inconsistent, the anime/manga & comics vs visual media vs sequential art vs I-don't-even-know category name is still undecided, the inability to filter out anything using the tag system, the tag system's essential deception (telling users that they can label with whatever while doing a top-down decision on which tags to connect). All of this contributes to a site which is hard for the reader to navigate.

The thing that AO3 does really well is the writer's side of things: the posting interface is a dream, editing is easy, deleting saves you a copy, it's easy to mark up your fic as this and that. But as a reader, filtering is a nightmare. I'm lucky that I can literally read everything in the archive for my fandom, because it's small; scrolling past all the crossovers is doable. But that's an inelegant brute-force solution, and so is manually filtering out tags by search - faceofcathy had an interesting post on search vs browse: filtering using search means knowing exactly what terms are used and listing them all. Manually. And this is not to mention that filtering had to be taken offline, it was hogging so many resources before. Adding meta to this mess - meta, which is often tagged with every single fandom that the author can think of - will basically spam each fandom in turn, and make it even worse.

Given AO3's reluctance to make users do anything, it seems unlikely that the meta will be findable or filterable. AO3 does not require users use a consistent tag or even to mark their work as anything separate - you can embed videos and art, but there's no way to find them without using some kind of search if and only if (IFF!) the author has marked their work "vid" or "art". If they haven't marked their work as such, good luck wading through! Vids and art are clearly both fanworks. But fic is clearly a separate category from art which is separate from vids, and right now, we don't even have a way to clearly differentiate different mediums. It's all up to the user, and there isn't a clear way to even distinguish that - do you tag? Put it in the title? Put it in the description? Which of these will be indexed? If fic/art/vids are hard to distinguish on the archive now, how will meta/fic be filtered for? They're both text-based; at the moment, a crude way of telling is to filter by wordcount (since a vid or a piece of art won't be 50K), which isn't doable for meta.

There needs to be some way to make them findable, and making that infrastructure to support all this - and it's a fine thought, I don't deny - needs to come before the cheering.

*seperis's post here talks about meta and whether or not it even has a place on the archive, given its status.