Extensible Web Summit Feedback

Now that I've had a chance to catch up and catch my breath, I wanted to
share some feedback on the summit.  I'm extremely excited that it was held
in the first place, and it was great, but there were a few bumps in the
road it would be great to improve upon in the future (in my opinion).  I'll
name them here as well as some ideas on how we might improve on them (where
I have any).  Keep in mind, I criticize because I <3 :)

Mostly, it was too short for the arrangement.  I feel like you could have
extended it and had fewer sessions simultaneously or kept the same length
and just tweaked how it was run and it would have been better.  The
strictly bar-camp approach, as it turned out, seemed pretty tough to me.

It seemed like very few people participated in actually setting topics
 Seems to me a lot of devs who attended were not comfortable throwing out
ideas in the given company.  I will come back to that - but at least as far
as ideas for topics goes - I'm willing to guess that that would be much
easier when you are not under pressure or in the company of people you are
perhaps a little intimidated by.  So one idea would be to collect session
ideas via www-tag or online upfront, from people attending.

The second bit is related, which is that it was really tough to arrange
them on the spot and not run into two problems: a) There were clear winners
and losers of sessions - some that were jam-packed and others which didn't
get enough attendance to really be viable.  That's a shame because b) the
'losers' weren't so much sessions that no one was interested in as much as
they were up against the real 'winners' and people had to make a really
hard choice.  Personally, I missed at least 2 that I really wanted to
participate in quite a bit because they were up against others I was also
interested in and thought I might benefit from more/be able to contribute
something to.  Gauging this sort of thing in the moment seems excessively
hard - what is a more optimal layout of schedule topics so that you
guarantee a good distribution?  If there were a collection of topics online
(as suggested above) in the days leading up to the event, it makes sense
that you could also gauge interest, narrow them down and generate a few
variant schedules.  Then on summit day, you could simply ask people to vote
on possible schedules.

My final comment is just that it seemed there were two kinds of "interest"
those where someone had specific things to talk about and those where
someone just wanted to hear more about something.  It wasn't really clear
which was which or who was going to present or whether people actually had
questions/comments/ideas, etc.  As a result - I saw two entirely different
kinds of sessions.  The first where there was a scramble of "who can talk
about this" and then not really a lot of actual discussion.  Sometimes we
didn't have someone to present and getting started was slow - and then once
presented it was just sort of a lot of awkward "ok... now what".  The
second where there was a clear thing to be discussed, but people really
disagreed on what it was.  In those a lot of contentious discussion drawing
things off-topic and that ate too much time and prevented a lot of the
stuff most of us really wanted to discuss.  It seemed to me that providing
or electing a facilitator (not necessarily the same as the presenter or
whatever we called them) to help run the discussion and keep it on track up
front might have been helpful.  In the former kind of scenario, they could
have asked leading questions and encouraged participation (from developers
too, who, as I mentioned sometimes seemed to have a hard time speaking up)
when it slowed down.  In the later they could have tried to remain
impartial and help keep it on track.

Perhaps that is too much structure, I'm not sure - it still feels pretty
good to me - but take it for what it is worth: Just one developer's opinion.

-Brian