[Linux-aus] LCA2014 - Next steps to get it done?

[Anthony Towns]

On 4 September 2012 16:08, James Polley <jamezpolley at gmail.com> wrote:
> I'm not sure where you get this idea from - I'm hoping it's not actual
> experience, because that would make me Very Sad(TM)

It varies -- I don't think the Brisbane bid team for 2010 had a very
good idea why Wellington's bid was chosen over theirs; and IIRC the
main challenge with Ballarat's bid versus Brisbane's for 2011 was
getting everyone to Ballarat, which probably could have been resolved
before the 2011 decision was made. More recently, aiui there haven't
been unsuccessful bids for either 2012, 2013 or 2014 for this to be an
issue... :)

> http://wiki.linux.org.au/Linux.conf.au_Bid_Guidelines#Bids encourages teams
> to contact the Council early, so that we can provide feedback as they're
> developing their bid.

Doing that in public might help both in the "more eyes" sense, and in
the "oh, right, that's how the team that won it last time did things,
let's do what they did" sense.

> If the main feedback is "Surprise, you win!", our problems are bigger than
> I'd realised.

For the Brisbane 2011 bid, it was a "surprise, you win!" in the sense
that making the decision between Ballarat and Brisbane didn't happen
on time, in what I presume is a similar way to the call for
expressions of interest not going out on time this year. "Surprise,
*foo* won!" traditionally happens for everyone not on the winning bid
team up until the end of the preceeding conference.

>> Some ideas:
>>  - don't include the "surprises" in the bids at all; ...
> Mikal is uncomfortable sharing the bid document too widely now - I'm sure
> his opinion will have changed by February next year, when it's time to call
> for bids for 2015.

Right, but how does that help the people bidding for 2014, who are not
only missing the Canberra bid docs but also apparently Brisbane and
Ballarat?

> Bid documents for 2008-2010 are available at [...]

> There are some surprises that I think *do* need to be part of the bid - eg,
> LCA2013 dropped to just one day of miniconfs, which was a fairly substantial
> change.

Well, I hope that wasn't meant to be a surprise, because if so you
just spoiled it? :)

(Also, uh, one day of miniconfs? With ten already announced? Augh...)

(Ah, thanks for the clarificiation Mike :)

> On the other hand, LCA08's main surprise was a completely novel way
> to do the penguin dinner.

For the sort of thing I was meaning, LCA08's main surprise was handing
out OLPCs iirc. LCA03's main surprise was getting Linus to attend.
Surprises are things that might require budget and organisation, but
that are (theoretically at least...) way more fun if nobody knows
about them in advance. "PDNS in a castle!", "Speakers dinner on a
boat!", "Balloon rides!", that sort of thing.

> Since the venues aren't always locked down by the
> time a winning bid is picked, *that* probably wouldn't have been vital to
> include.

I think by their very nature surprises shouldn't need to be in the bid
document -- if the only way a conference is going to be any good is
something you can't tell people about until the conference is
underway, that's probably not a recipe for confidence. And get rid of
the surprises, and there's no reason the bid documents can't be public
from the start that I can see?

>>  - make the "expression of interest" phase a bit more useful to
>> bidding teams, like helping them get introduced to speakers/ghosts, or
>> other folks who might be able to help out organsing. "we are
>> interested in running lca in 2014 in Broome; would interested folks
>> mind joining us on irc on friday at 6pm WA time to ask us any
>> questions and/or offer help? thanks!";
> Yep, that's the (intended, at least) current process. I know that we've been
> in touch with all the teams who've expressed interest this year - if they've
> felt any lack of support, we haven't been aware of it.

Well, the main thing is that the council isn't getting the number of
bids it wants, so it needs to do something to make bidding easier/less
costly and more valuable/beneficial to potential organisers.

>>  - have the preliminary round of bids be a bit more interactive -- two
> The preliminary round of bids (according to
> http://wiki.linux.org.au/Linux.conf.au_Bid_Guidelines#Timeline) runs for
> *three months*, rather than just a few weeks - plenty of time for the teams
> to get feedback from the council/community/other teams if they want that.

Depends. If the tradeoff is "we get our ideas stolen and maybe flamed
vs we keep it secret and hope we're the best" that might not get as
much feedback as "okay, we have to make it public, and do the best we
can".

Public bids would've raised the "uh, nobody's bid for lca 2014" a bit
sooner when it might have been a little less desperate to try to come
to a solution. (And would've made the "uh, only one bid for 2012 and
only one bid for 2011 -- have we exhausted the interested teams?"
issue something that could be a bit more publicly addressed too)

> I've heard a couple of variations on this theme. For instance, if we could
> pick a winning bid 2 years in advance,

FWIW, we tried having the next LCA location announced in advance of
the preceeding LCA being held once (2008's location was announced
prior to LCA 2007), and the 2007 team found that distracting. It might
be worth trying it again, potentially handling it differently, though.
(If all bids were public, and there was only one bid, this would of
course happen automatically...)

Cheers,
aj

-- 
Anthony Towns <aj at erisian.com.au>