[OSGeo-Conf] FOSS4G selection 2014 - consideration for new method

[Eli Adam]

Hi all,

Given the quality of both proposals for 2020, I've been thinking a lot
about the best criteria to make a decision.  Since about 2014 and possibly
before, I think that the FOSS4G selection process does not serve our
community or the conference as well as it could [1].  The selection process
may also have harmful side effects.  Due to my personal involvement with
2014, I'll keep most of my comments oriented towards 2014 but it has been
applicable to other years as well.

What are valid criteria for selecting the FOSS4G LOC?  The criteria I
personally have used are that FOSS4G is OSGeo's primary source of income
and thus very important.  The conference should have a high probability of
success and low risk.  I look at the budget, how reasonable I think the
numbers are, and if there are any objectionable contracts (usually hotel
block commitments).  I look at the LOC members and their experience.  I
also look at the geography of past conferences and value bringing FOSS4G to
a new region.  Beyond that, I have not been able to come up with additional
selection criteria that I consider valid.  What do others think?  I'd like
to add to this list.  Recapping the criteria, that is:
1) High probability of success
2) low risk
3) reasonable budget
4) absence of objectionable contracts
5) LOC experience
6) FOSS4G geography and history

(I also have personal preferences like where I might have a free place to
stay, what's a cheaper travel option, who I know, etc but don't consider
those valid criteria.  And purposely don't vote on those items.)

Given those valid criteria, I often evaluate all the FOSS4G proposals as
extremely good.  Each having extremely high probability of success and
relatively low risk.  In many years, I've not really found valid reasons to
select one proposal over another.  I found that to be the case even when I
was on the LOC of one of the proposals!

While a member of the 2014 LOC during the bid process, I could not honestly
assert that the PDX proposal was any better than the DC proposal.
Obviously as a member of the PDX LOC, I was in favor of ours, but that
self-serving interest is not a valid basis.  Both proposals would have led
to great conferences with high probability of success, low risk, realistic
budgets, no objectionable contracts, great LOC experience, and FOSS4G
geography.  I've found this near-equivalence of proposals to be the case in
more than one subsequent year.

With proposals of near-equivalence, I see no point in voting and selecting
one.  This leads to putting two spatial centers of great OSGeo and FOSS4G
enthusiasm into opposition.  This competing is not the typical
collaborative OSGeo and FOSS4G way.  It is in fact perhaps contrary to the
manner in which we build software together.  With the FOSS4G selection
method we use now, we invariably greatly disappoint one of the proposal
groups.  We also are creating a lot of waste and wasted effort.  I'd like
to see a conference selection method that more closely matches the
collaborative spirit in which we approach other endeavors.

How our current selection method fails to best serve the conference or our
community and possible harmful side effects:
1. Makes something trivial overly important.
2. Creates divisions
3. Zero-sum competition (as opposed to the competition of the old WMS
shootouts which were beneficial to all the softwares and users of the
software).
4. Does not mirror our collaborative approach to software development and
other collaborative activities.
5. Disappoints a group and region
6. Fails to make use of great potential.
7. Does not make a better conference based on the above criteria

I take FOSS4G selection more seriously than anything else that OSGeo does.
FOSS4G selection is more important than anything that the Board will do in
the next year.  OSGeo's (financial) existence depends on the FOSS4G
selection. Therefore I'd like us to re-examine how we make the selection.  I'd
like to consider a new FOSS4G selection method.  Would you like to see a
new FOSS4G selection method?  What would that look like?

This is an off-handed critique I leveled in private conversation which I'll
quote: "If we were a competent organization, we would recognize that there
is demand for TWO successful conferences in Canada.  We would on the basis
of costs and other advantages, select one for 2020 and the other for a 2021
regional conference (the 2021 "regional" conference may actually be
"better" by following after the other and building on the enthusiasm and
having another year of planning.)"  I've not been involved with the
FOSS4GNA organizing but perhaps these efforts could be harmonized in some
manner?  I'm not really knowledgeable on this topic, so someone
knowledgeable should talk about this.  While I'm straying from 2014
commentary, I'll also comment that these two 2020 proposals for a North
American year were strikingly similar.  Both are in Canada (I would have
expected at least one US entry before two from Canada), both are taking the
novel approach of in-housing the PCO services, and both rate well on the
above valid criteria.


[1] Previous thoughts about ties but similar to these thoughts.
https://lists.osgeo.org/pipermail/board/2014-February/006720.html


Best regards, Eli
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/conference_dev/attachments/20181218/453573bc/attachment-0001.html>