On 28 March 2013 03:18, David Lyon <span dir="ltr"><<a href="mailto:david.lyon@hackerpads.com" target="_blank">david.lyon@hackerpads.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im">On 28.03.2013 05:39, Tim Ansell wrote:<br>
> If you have watched my talk at<br>
</div>> LCA2012 <a href="http://www.youtube.com/watch?v=rCoCRmcrPlM" target="_blank">http://www.youtube.com/watch?v=rCoCRmcrPlM</a> [1] - you'll<br>
<div class="im">> already know the best way to get high quality video is to do live<br>
> streaming.<br>
<br>
</div>Nice Tim,<br>
<br>
Forgive me asking, but in that video back then you said you wanted to<br>
get it going.<br></blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Fast forward to now, I assume it's pretty much sorted out? I know you<br>
know<br>
that I have seen it working nicely, but I'm wondering how many things<br>
are<br>
on the todo list.<br></blockquote><div><br></div><div>I know you're trolling, but here is an actual detailed response for anyone who is interested.</div><div><br></div><div><div>LCA2012, PyCon US 2012 and 2013 where all examples of where the system are/were working. Keep getting lots of great feedback. Things are moving fast - FYI Just in the last year, webm has gone from <5% of the streams to around >75% (go chrome!). Working with James to eventually make this a single button setup using Amazon Cloud Formation. <b>Would love help!</b> Code at <a href="https://github.com/timvideos/timvideos">https://github.com/timvideos/timvideos</a> and mailing list at <a href="https://groups.google.com/forum/?fromgroups#!forum/timvideos">https://groups.google.com/forum/?fromgroups#!forum/timvideos</a> (moved from <a href="https://groups.google.com/forum/?fromgroups#!forum/streamtime">https://groups.google.com/forum/?fromgroups#!forum/streamtime</a>).</div>
<div><br></div><div>The "single room, single talk, user group" user case is still being worked on. The first revision was just not reliable enough to allow someone other than myself to operate (mainly due to dodgy cabling and placement of parts). I have a <i>revision two</i> of the portable system and I'll be using it as the year goes on. Still waiting to produce something that is easily reproducible, might not be possible with this hardware setup.</div>
<div><br></div><div>The TwinPac device which is the central system of the current solution has been EOL and nothing was released to replace it. We are now working on Open Source hardware to replace it and provide features that were previously impossible. Code at <a href="https://github.com/timvideos/hdmi2usb">https://github.com/timvideos/hdmi2usb</a> and mailing list at <a href="https://groups.google.com/forum/?fromgroups#!forum/hdmi2usb">https://groups.google.com/forum/?fromgroups#!forum/hdmi2usb</a></div>
<div><br></div><div>dvswitch which is also central to the current system only supports a "standard definition" workflow. We are interested to upgrading to high definition and multiple streams, hence the gst-switch is being developed. Code at <a href="https://github.com/timvideos/gst-switch">https://github.com/timvideos/gst-switch</a> and mailing list at <a href="https://groups.google.com/forum/?fromgroups#!forum/gst-switch">https://groups.google.com/forum/?fromgroups#!forum/gst-switch</a></div>
<div><br></div><div>New technologies have opened new avenues, hangouts and webrtc are coming. Things like automatic speaker tracking using OpenCV and face detection are on the cards, it's an exciting time to be involved with this stuff.</div>
<div><br></div><div>The whole point of this group is to help evaluate these systems and get them ready for LCA without it being a burden on LCA organisers. A/V discussions for 2014 and 2015 should already be happening now.</div>
</div><div><br></div><div>Tim</div></div>