It was nice to be back at No Time To Wait. It's such a fantastic gathering of archivists and developers and the atmosphere is always great. It's nice to feel welcome although I am not an archivist and know little of all the things they have to go through in their job. Although every time after NTTW I feel like I know a bit more and thus can help more.
The work they do with Matroska (combined with FFV1) is exactly one of the main use Matroska was designed for. Except neither me nor the other creators of Matroska ever dreamed it would ever be used professionally for prestigious archival like at the BFI (British Film Institute) and possibly at the Library of Congress in the USA. Of course there was already traction in the "corporate" world as MKV is used for all kinds of video sharing as files and also the basis of WebM. It's officially supported in OSes like Windows 10 or Android. But the archival world is a different thing. It's not only about sharing the latest movie//TV show you got, but it's keeping important content for a long time. IMO it has a deeper impact in the long term and some historical, political and artistic value that can't be beaten. This is also what makes it so special to me as I always try to find an extra bit of "soul" in whatever I do. It's not just about writing some code or documentation. It's also a great motivation knowing it will benefit something bigger.
It's amazing what a hobby project started (forked) 17 years ago has become. In my mind (and probably all other involved) we were driven by the same spirit that was on the Internet during that time. Creating something great for free and possibly challenge the "corporate" world. One of the U in my robUx4 nickname stands for Utopia. That was always part of the goal.
Just talking about why some details of Matroska and how we came to the conclusion of that detail always remind me of the amount of work we put in this, all the challenges we faced (like trying to be as good as ogg for streaming). It's fun having to go back to all these memories when we're doing the IETF specifications and realize the things we got right as complete n00bs and some we got wrong (nanoseconds non rationale timestamps that we called timecodes, to show how n00b we were, the clock was even in floating point because in the analog world clocks aren't perfect). Matroska was designed to last 10 years in a time were there was a new video codec every 3 months. It was still hard to predict the full evolution of things (no VR for example). The challenges posed by long term archival is also interresting. Here we have to support all kinds of sources (analog and digital) with very specific characteristics (and keep everything as does RAWcooked thanks to attachments). There's still plenty of areas to improve (Timecodes, Bayer support). After 17 years, Matroska still has room to grow.
I am very grateful to have meet all kinds of new people who care deeply about the work we do on Matroska and see again familiar faces who care at least as much. You have no idea how great is to have all of your support and that you took a leap of faith choosing Matroska when we just started making proper specifications. From NTTW1 it was not very clear whether people would actually use Matroska at all.
I would like to thank in particular the organizers of the event: Dave, Jerome, Ashley, Alessandra and Zsuzsa. It's a privilege to be welcome in your community.
👓👓
