- From: Glenn Maynard <glenn@zewt.org>
- Date: Thu, 7 Jun 2012 22:21:53 -0500
- To: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
- Cc: Philip Jägenstedt <philipj@opera.com>, public-texttracks@w3.org
- Message-ID: <CABirCh-sTpVxxNh=sPW1c_S-LFv2Nr6BhxwxTXZoYWc3+uacLw@mail.gmail.com>
On Thu, Jun 7, 2012 at 9:44 PM, Silvia Pfeiffer <silviapfeiffer1@gmail.com>wrote: > No it's not, but it's easier to put DRM on muxed files than on plain > text files. > (Sure, but that's not why people mux captions into .MKVs.) It was at one stage. I doubt they will do it again - with or without > muxed files. If they were, I would indeed hope it to be muxed. I > would, however, not expect browsers to implement automated muxing from > video+track markup (it would be nice, but I'm not hopeful, because it > creates a different content to the one that was published). So, when > downloading (right click "save video as") you will continue to get > individual files for the foreseeable future. > I don't think there are any hard problems in the way of browsers making "save as" automatically mux everything together, but I agree it's more likely to be a server-side feature for a while. TL;DR: we should probably move on to say: no matter our motivation, we > agree that we need metadata in vtt files. Let's design the solution. > I think we mostly have a basic proposal from the previous thread. To sum it up without repeating all the details, it would look like: Key: value or Key: | value value . Key include "kind", "srclang", "label" and "default", with the provision that if a value conflicts with the value specified by a container (eg. <track> or WebM track data), the container's value takes precedence. -- Glenn Maynard
Received on Friday, 8 June 2012 03:22:29 UTC