Following recent examples by via_ostiense
, here's what I'm doing right now as part of my OTW work, in broad strokes.Translation
: I think I finally mastered context rules
. GO ME. What's context rules? Simply put, they're an integral part of the translation interface for the Archive of our Own
that the inimitable rebecca is building for us. (It's beautiful. It's for translating the interface only, though, not for stories.)
In more detail, it's grammar nerdery*: How many potential versions of a phrase/word are there in language_x, depending on context? We have to determine that to teach it to the archive so that it produces as many versions as we need. (*Actually, it's information science slash grammar which makes for a slightly brain-breaking combination -- hence the taking a bit to wrap my mind around it...)
An example: in Polish, the possessive form varies depending on whether the subject is male of female. Now...this means we need a way to tell the gender of a name. Now imagine pseudonyms in their infinite variety. It's not possible. So what do we do? Add a "gender" setting? Uuuh...You see this is getting into volatile territory. There are a couple of other possibilities, thankfully, but this is the kind of eventuality we have to consider.
So we've been gauging the limitations of the system for a while by testing for as many languages as we could get our hands on. If we don't make context rules flexible enough for just about *any* language and script system, we'll get into real trouble once it's implemented on the live site and it turns out that, say, new language_x makes a distinction between animate/inanimate objects that we hadn't considered.
We've already tested to varying degrees for Finnish, Italian, German, Korean, Spanish, Dutch, Norwegian, Romanian, and Polish; we're hoping to test Chinese and Hindi soon, because as you can see, we have tons of Indoeuropean languages that are relatively similar, and desperately need other script systems and language families. And yes, we totally want you for this! We'll make an open call for testers soon, I'll signal-boost once it's up.
Other than that, I signed up two Italian translators last night, one an old friend. I always enjoy these introductory chats because of the high energy and (often) wacky cheer that inevitably develops! :> So now I'm dropping website texts for translation into Gdocs and assigning them via GQueues (our new to-do-list tool). Go go Italian team! *\o/*
For Internationalization & Outreach
(eh, I&O), meeting up with Claudia from my committee later to work on Internal Survey. Lots of food for thought from the recent discussions, thinking about how to integrate that meaningfully into the survey. Later this week, I need to work on our term evaluation (what have we done, what are the loose ends in need of following up?) and a collaboration with the Archive peeps. Oh, and do some wiki gardening because of the rename. >_>;