[H-GEN] Talk at today's Humbug meeting

[Arjen Lentz]

Hi Russell

On 21/03/2009, at 1:04 PM, Russell Stuart wrote:
> We have done the bid twice now, and so the bid document is already
> pretty polished.  After finishing it last year I recall saying if we
> lost it wasn't obvious what we could do to improve it, should we  
> loose.
> No doubt looking ideas like those you list below and reading at
> Wellington's bid will prompt a few changes, but still I suspect it  
> will
> be more a case of tinkering around the edges.

I don't think that angle is a winner. If the bid was the best, it'd  
have won; so we may conclude that it wasn't.
And in the mean time there's one additional LCA to compare any bids  
with. It'll have to differentiate, and in some cases possibly improve  
- not just on itself but also compared to anything at LCAs that can do  
with improvement.


>> For instance, I wouldn't mind seeing lunches included.
>
> Yes.  Having everyone in a "common" lunch place would encourage
> schmoozing.  Including lunches in the ticket price would encourage  
> that.
> It would have to pass the budget test though.

Ditching the speaker event/dinner would cover this abundantly, I would  
think.
So you now have a bit of money to spare for other things.


>> In Hobart, the BoFs were a scattered mess; rather than accepting  
>> this,
>> see if it can be improved; it'd be nice if more people went to BoFs
>> and didn't lose out, and if things didn't conflict with 'em.
>
> In hindsight, when looking back at previous LCA's there are always
> things that could have been improved.

Exactly. So let's.

> That said, Hobard had an unusually large number of things I would have
> done differently, the BoFs being but one.

Yet I do think it was one of the best LCAs.
Anyway, inventorise the grant topics, the details in those that were  
particularly wrong, and then we might get an idea of what can be  
improved.
It's not merely a fact. It's an opportunity.


>> I personally have to say that while the speaker dinner is nice, it
>> separates a group from the rest for no particular reason. The actual
>> benefit is quite minimal, in fact it just means that some peope can't
>> talk with some other people that night. It's nice as a thank-you for
>> speakers, but I'm pretty certain that if it came down to it speakers
>> don't actually give a stuff and wouldn't not submit for LCA if there
>> weren't a speaker event.
>
> Agreed.  But you would have to replace it with something that did
> facilitated people talking to socialising.

Organised events are actually a hindrance to people socialising, since  
there's logistics of getting to the venue, it's often a limited set  
(speakers, prof.delegates, etc) and the venue arrangements themselves  
will impose restrictions. The prof delegates dinner in Hobart was  
cramped and thus no good for chatting, given that it was raining  
outside. Sometimes chatting might mean getting a laptop involved, and  
again that is unpractical at organised events. There's no abundance of  
space to sit in little groups and talk or do stuff.

I think it's important to realise that not all days at LCA have  
evening events now.
Ditching a particular limited-subset event does not require  
replacement as such.
The non-subset part of LCA delegates were already doing their own thing!

In Hobart, various subgroups cooked dinner; this was a great success  
in terms of people getting together creatively and doing stuff that's  
not computer-related. At the same time, using laptops was easy as  
there were couches and chairs in the units, and wireless of course.

I think that what LCA can do on some nights is perhaps enable certain  
things (as in make possible), and maybe facilitate.
But it's the over-organisation that in my opinion may actually be part  
of the problem. (mind you, OSDC'07 Bris didn't get this quite right  
either!)


Example... I still think that the Melbourne conf dinner was among the  
best, because it was at the Melb markets.
At other organised dinners you get fed and sit at big round tables...  
having only that just doesn't lend itself to mingling and creativity,  
it becomes quite awkward with moving around and so on.
LCA can set up the framework, and the rest will sort itself out. No  
need for a tight program in those events.

I'll give you an *example* of how this does work. At OSDC 2007 we did  
something that I'd seen done before at an AUUG conf, but it just ended  
up so much better: we put a few jars of Playdoh at each table, after  
the pre-conf talk we gave each table the task of "depict a bug". This  
forces people around a table to interact if they were a bit shy, and  
work together on something. All kinds of ideas and stuff get  
discussed, and any social awkwardness is forgotten. It's not  
smalltalk, and not forced. It just works. Also the most amazing things  
came out of it, including a complete claymation movie with audio/music  
that ended up on YouTube. Who would've thought!

Another thing we did at that dinner was randomise (part of) the  
seating. The base idea was that you tend to huddle together with  
people you already know, but that kinda defeats the purpose of a conf  
where you'd want to meet more people. Some of the organisers reckoned  
many would not like this, and so we ended up with randomising half of  
each table. We put numbered tickets into a hat and people walking in  
had the option of picking a number and thus let their table be chosen  
for them. Guess what... we ran out of numbers!!
Lesson there: never underestimate the ability of a geek to embrace a  
new idea, you just need to get 'em out of the automated flow (the  
natural tendency is to cluster with known people). The hat did that  
and the result was brilliant!


>> I also think there's an overload of sessions, but a lack of space
>> (both in time and physical arrangements) for just meeting and talking
>> with people. Everybody agrees that's a key aspect of a conf, but most
>> confs don't actually keep that in mind when organising a venue and
>> timetable (to name just two aspects that would influence this). If  
>> you
>> can make this a key factor, then other things might change and that
>> will certainly differentiate a Bris LCA bid from others. And it'll
>> make sense.
>
> Did you have any specific ideas on how to do this?

Ohyes ;-)

I think there should be an abundance of couches and tables/chairs  
around the conf venue, of course with powerboards.
No rectangles, but circles. Order in pizza for one night, and you have  
a perfect social/hack evening. LCA can fund that pizza.

I wouldn't suggest to replace the conf dinner with this, but if we try  
it once we may well end up ditching the conf dinner ;-)
So, the above kind of arrangement can replace the speaker dinner  
night, and be inclusive for everybody.


Perhaps freeform session time early in the afternoon (like BoFs), and  
groups potentially showing results at the lightning talks.
The amount of creativity present... it just needs time to jive, and  
confs are so full this becomes a problem.

I think a better conference might have fewer fixed presentation  
sessions, and more freeform stuff and workshops.
The appropriate infrastructure and arrangements need to be provided,  
allowing magic to happen.


Everybody always agrees that the best bits about a conf happen outside  
the main sessions.... let's take that feedback on board, and make the  
main conf pick up some of that difference that makes a conf more  
worthwhile.

Cheers,
Arjen.
-- 
Arjen Lentz, Director @ Open Query (http://openquery.com)
Affordable Training and ProActive Support for MySQL & related  
technologies

My blog is at http://arjen-lentz.livejournal.com
OurDelta: free enhanced builds for MySQL @ http://ourdelta.org