Re: [Debconf-team] DC13 Venue constraints

[Gaudenz Steinlin <gaudenz@debian.org>]

Hi 

Thanks to Richard for the detailed breakdown of attendee statistics. 

Richard Darst <rkd@zgib.net> writes:

> I remember for DC10, at first some localteam people wanted to reserve
> 400 rooms because it would be huge and DC7 had 400 people.  But I
> looked at the detailed breakdown and it wasn't so big actually, even
> though 391 people had "arrived" according to penta, the day with the
> most people (according to penta) had only around 250 people.  (I
> think) I eventually convinced people to reserve only 300, and we still
> had less than 200 used any given day.
>
> I just went and found that old script (I would have had it in my
> standard dumps, but the God table I join with only exists for the
> current debconf so I didn't extend to old years).
>
> Here's the data:
>   http://rkd.zgib.net/http/debconf/historical-room-statistics.txt
>
> for every debconf, it lists number of people requesting accom, and
> total number of people in parentheses, per day.  You can compare to
> total attendees at the top.

do I understand this correct, that the number is paranthesis are all the
participants, including those who arranged their own accomodation?

>
> Some effects I can think of:
>
> - "biggest" conference per total attendees is still dc7, but it isn't
> that much bigger on any given day.  This means lots of people came out
> for just a subset of days, non-overlapping.  (implication: dc13 needs
> fewer beds than expected)
>
> - The conferences that have more appealing surrounding city (7, 10)
> have a smaller proportion of people in accom.  (implication: dc13 needs
> more beds, counterbalancing the above)
>
> - it could be the dc13 attendees won't be as high as expected because
> if it's more remote, fewer people will come out for just a few days.
> Alternatively, if that isn't the case, it means the population won't
> be so high as expected based on the big-city debconfs.
>
>
> Basically, my main point is people should think not only in terms of
> total conference sizes, but per-day attendees, and how that affects
> things.  I don't know the solution, and I don't have time or desire to
> debate the finer points, so I'll leave it to you all to see.
>

Or another calculation attempt: I think our estimations should be based
on the actual attendee count and not on those requesting accomodation. I
expect the percentage of people arrangeing their own accomodation to be
about 10%.

capacity of DC13 (326 + 10%):          358
max attendants on a single day at DC7: 267
difference:                             91 (= 34% of 267)

What I want to say with this is even if we have 34% more attendees than
DC7 or the DC7 numbers are wrong by up to 34% we still have enough beds
at "Le Camp". I know this is a bit theoretical and has many if's. But
this calculation makes the issue much less pressing than it seemed to me
before.

In summary: I'd still like to have an option to extend the number of
beds at Le Camp with a camping option if we need it. But if this is not
going to work I think that an alternative locaiton needs to be at least
as good on the main other cirteria than raw number of beds (distance to
airports, hacklabs and talk rooms, price, ...)

Gaudenz

-- 
Ever tried. Ever failed. No matter.
Try again. Fail again. Fail better.
~ Samuel Beckett ~