The Unconference idea upends that model. Instead of starting with vendors and solutions like most conferences, the idea was to start from the problems (which are numerous) and just talk things through. The format had a few things going for it:
- Everyone -- tool vendors, LSPs, customers, consultants -- was on equal footing.
- The agenda was set first thing in the morning, based on what the attendees wanted to talk about.
- No PowerPoint allowed.
- The SalesForce.com cafeteria makes a pretty good sandwich.
One of the things I really enjoyed was that the Unconference put front and center the discussions that are usually relegated to hallways and post-conference trips to the bar: what things don't work, why things don't work, and why making them work is tricky.
This is something that the localization world has been grappling with for a long time, and our progress hasn't been tremendous. There's an obvious, growing realization that localization is increasingly technology-driven, yet most people still feel like they're underserved by technology. And process, which is so critical, can vary wildly, due to a combination of fragmentation in the tools market, lack of standard practices, and because so many specific industries require so many things to be done their own way.
So what did we talk about? I sat in on several sessions:
- "Towards a Unified Localization Process." (Note that none of these sessions had official titles, so I'm making them up.) Tex Texin of Xencraft was informally driving this discussion, which was an attempt to abstract away some of the twists and turns of business-specifical localization workflow into something more generic. Towards the end, Tex plugged in his laptop to show a mindmap he had made of the things that might be involved in a localization workflow. It looked like a spider with a thousand legs. Simplifying this to something reasonable is a tall order for a single hour, and I am sorry to report that we didn't succeed. But it's absolutely the sort of discussion that people need to be having.
- "Localization tools: Why is everything terrible?" (Again, that is merely my name for it.) At the start of this, Shawna Wolverton asked for a show of hands from people who were happy with their current localization tools. I believe a single hand was raised. This was the session I was most directly involved in, since I've spent some time working on tools, and the difficulty of creating broadly-useful tools is something I've thought about a lot. I'll probably make a follow-up post about some of this at some point, since a lot of it was interesting.
- The obligatory Machine Translation session. The discussion ended up being largely about how to tell whether MT is actually a viable alternative from an ROI perspective. This was interesting in part because it indicated (to me at least) that we're moving past talking about whether MT is good or bad for the industry, and on to talking about when and how it should be adopted. There was also a lot of good discussion about what other things need to be done to actually make MT effective -- in particular, improving communication throughout the entire authoring and localization process.
- "Localization and Agile Development Processes." This started as vent session from people whose teams have switched to agile software development and left them holding the bag for localization. There are some really interesting questions here, and few obvious answers. As in the MT session, a recurring theme was the importance of communication between doc and loc teams.
Anyways, there's another one coming up in Dublin on May 20, which I will unfortunately not be attending due to its being very far away. There is talking of making this an annual event, which I hope happens.
Oh, lastly, there was one notable accomplishment:
See you there.
0 comments:
Post a Comment