summaryrefslogtreecommitdiffstats
path: root/2019
diff options
context:
space:
mode:
authorAmin Bandali <bandali@gnu.org>2020-05-13 02:45:54 -0400
committerAmin Bandali <bandali@gnu.org>2020-05-13 02:45:54 -0400
commit9ac9c1a1c2f3ae55e28de64c7feccf0e3e77df9b (patch)
tree130d3d3c0340564ce173b0e8caee71bc8d3b31ae /2019
parent19ac3821537ff8b19283e2a7569e838f6444a894 (diff)
downloademacsconf-wiki-9ac9c1a1c2f3ae55e28de64c7feccf0e3e77df9b.tar.xz
emacsconf-wiki-9ac9c1a1c2f3ae55e28de64c7feccf0e3e77df9b.zip
add 2019 planning page
Diffstat (limited to '')
-rw-r--r--2019/planning.md82
1 files changed, 82 insertions, 0 deletions
diff --git a/2019/planning.md b/2019/planning.md
new file mode 100644
index 00000000..ad2b7a34
--- /dev/null
+++ b/2019/planning.md
@@ -0,0 +1,82 @@
+[[!meta title="Planning"]]
+
+The organizers and other volunteers hang out in the `#emacsconf` IRC
+channel on `irc.freenode.net`. If you would like to get involved and
+help us out with the planning, come by the IRC channel and say hi!
+
+We will be collecting various ideas and plans for organizing the
+conference and adding excerpts from `#emacsconf` logs to this page.
+
+## Current plans for infrastructure
+
+The event will probably be comprised of two main parts: the video-call
+part, and the live-stream part. We are separating these two, both for
+speed/overhead/bandwidth concerns (if a large number of people decide
+to "join the video-call" to watch), as well as moderation and
+distraction concerns (having to moderate a large number of people in a
+video-call, where only one of them is supposed to be speaking at a
+time).
+
+So, we will use one series of software for the actual video-calls,
+whereby speakers would take turns joining the call and deliver their
+presentation, and another series of software for streaming the
+video-call for everyone else to watch live.
+
+1. *Video-calling:* we currently have two candidates for the
+ video-call software, namely [Jitsi Meet](https://meet.jit.si),
+ [Jami](https://jami.net); under Apache-2.0 and GPLv3+ respectively.
+ I have not yet tried Jami, but Jitsi Meet seems to work fine in
+ Firefox and Trisquel's Abrowser (without LibreJS). Sadly, Jitsi
+ Meet currently doesn't have explicit license headers on their site,
+ so it seems that LibreJS blocks their JS.
+
+ I will be opening a bug report for Jitsi Meet asking them to add
+ license information to their hosted version of Jitsi Meet so that
+ LibreJS would allow their JS to run. If they decide not to, we
+ could try and run a self-hosted instance, though then the challenge
+ would be the inter-continental latency: I'm fairly sure Jitsi
+ Meet's hosted version at meet.jit.si uses an array of servers
+ across the globe to help reduce latency, and that's something that
+ we certainly can't afford if we were to self-host.
+
+2. *Live-streaming:* this is the software most people (i.e. watchers)
+ will be facing. We will likely use a combination of OBS
+ Studio+Nginx+RTMP to capture the video call, and upload/stream it
+ to our server, where people would then be able to either watch the
+ stream in their web browser (similar to LibrePlanet's embedded
+ video player) or point their media player (such as VLC or mpv)
+ directly to the stream and skip the browser entirely.
+
+ Greg Farough helpfully provided me with a series of [config
+ files](https://git.sr.ht/~ggoes/obs-nginx-rtmp) they had previously
+ used for setting up live-streaming as I described above. I will be
+ trying to set it up on our (well, my) server and test streaming.
+ Since my server is a fairly small (virtual) machine, I'm thinking
+ of asking the FSF sysadmins to see if they could kindly help out
+ with extra computing power and/or bandwidth on the day of the event
+ if needed.
+
+## Older notes for infrastructure
+
+We need to decide on a good way to host the conference. An important
+priority here is to use as much Free Software for this as possible,
+ideally 100%.
+
+At the moment, it seems Jitsi Meet is our best bet. Let's see if we
+can self-host it and if it's usable with limited resources. If not,
+we might have to use the flagship version hosted at
+<https://meet.jit.si>.
+
+Bandwidth-wise, it seems that Jitsi's videobridge may not be a
+[bottleneck](//old.reddit.com/r/emacs/comments/cjam80/emacsconf_2019_call_for_proposals/evd0n8e/).
+It seems using Jitsi could be feasible.
+
+At some point, we should probably contact the admins of meet.jit.si
+and verify that it is okay to run a conference on their site. We
+should also obviously try it out thoroughly well in advance of the
+conference.
+
+Other candidates?
+
+- maybe ask Nextcloud to sponsor some talk/video hosting?
+- maybe ask the people at <https://streaming.media.ccc.de> for advice?