[Foss4g2013] presentation selection [SEC=UNCLASSIFIED]

Thu May 23 02:58:52 PDT 2013

Hi Bart,

I didn't want to be fierce, but explaining my experience. I especially
felt like replying as you gave me a reality check quite often on topics
where I had a completely different view in the past.

My problem with the community voting is, does it really reflect what the
community wants? I'm not saying the community is too stupid to know what
they really one and someone else needs to decide what's best. I think
the problem is the open voting. It's easy to get an bias in there. The
people that actually vote is a small subset of the people that will be
at the conference, but the conference should please the whole audience.

I for example prefer developer centric talks. I don't care much about
talks that are about "I've used this and that open source technology to
do x and y" or about INSPIRE. Though there are probably quite a few
people from institutions that don't yet use an open source stack or want
to learn how to leverage open source when they need to meet the INSPIRE
goals. It would be valuable to have such presentations. This is what the
LOC is for, they can make the call to include those as well.

Another example which is a bit artificial, it's about popular
presenters. Let's take Paul Ramsey as an example, he's one of the best
speakers I've ever been to at conferences. If he would submit 10 talks,
probably all of them would get voted by the community (being it due to
great abstracts or to know that Paul is presenting). But of course you
don't want to have one person doing to many talks.

And finally the problem of people trying to abuse the public vote (or
have friends that try it). You can filter those out sometimes, but would
you then publish the filtered results or the raw data?

Though I think it's good to have this discussion. These thoughts have
previously only in my brain and never written down. So it hopefully
helps for future conferences to improve the process.

Cheers,
  Volker

On 05/23/2013 06:51 AM, Bart van den Eijnden wrote:
> Given the fierce responses, I will think twice about ever making a suggestion on a selection process for FOSS4G again. Sorry to have spend my time on this.
> 
> Bart
> 
> Sent from my iPhone
> 
> On May 23, 2013, at 1:59 AM, Bruce Bannerman <B.Bannerman at bom.gov.au> wrote:
> 
>> Thanks Volker.
>>
>> Agreed.
>>
>> Can I suggest that if someone believes that they have a better process, that they volunteer for the LOC of the next international FOSS4G conference and try it then?
>>
>> Bruce
>>
>> ________________________________________
>> From: foss4g2013-bounces at lists.osgeo.org [foss4g2013-bounces at lists.osgeo.org] On Behalf Of Volker Mische [volker.mische at gmail.com]
>> Sent: Thursday, 23 May 2013 1:04 AM
>> To: Bart van den Eijnden
>> Cc: foss4g2013 at lists.osgeo.org; <conference_dev at lists.osgeo.org>; Barry Rowlingson
>> Subject: Re: [Foss4g2013] presentation selection
>>
>> Bart,
>>
>> I second the approach that was used by the LOC. It's similar to what was
>> done in 2009 (when I was part of it).
>>
>> Barry described how they made the selection in detail. It is important
>> that the way the decision was made is transparent, not the decisions
>> themselves (it would take way to much to give a reason for every not
>> accepted abstract).
>>
>> The LOC should make the final call and normally it's pretty close to
>> what the community voted for (at least that was the case in 2009).
>>
>> Cheers,
>>  Volker
>>
>>
>> On 05/22/2013 04:58 PM, Bart van den Eijnden wrote:
>>> Barry,
>>>
>>> does this mean you don't have enough trust in the community voting that
>>> they will filter out anything inappropriate?
>>>
>>> I see this as an unnecessary and confusing step.
>>>
>>> Best regards,
>>> Bart
>>>
>>> --
>>> Bart van den Eijnden
>>> OSGIS - http://osgis.nl
>>>
>>> On May 22, 2013, at 4:50 PM, Barry Rowlingson
>>> <b.rowlingson at lancaster.ac.uk <mailto:b.rowlingson at lancaster.ac.uk>> wrote:
>>>
>>>> On Wed, May 22, 2013 at 2:49 PM, Bart van den Eijnden
>>>> <bartvde at osgis.nl <mailto:bartvde at osgis.nl>> wrote:
>>>>
>>>>> But apparently the selection committee filtered out abstracts based
>>>>> on the
>>>>> words open and or free, which seems a weird and error-prone approach
>>>>> to me.
>>>>
>>>> We did *not* purely filter out based on words.
>>>>
>>>> We looked at the title, short abstract, and long abstract. If from
>>>> those items we could not see a  free/open-source, open-data, or
>>>> geospatial angle, we *thought carefully* about whether that should be
>>>> included in the conference.
>>>>
>>>>> My second talk was about GeoExt and since I thought since everybody knows
>>>>> GeoExt is about open source, I did not mention those words explicitly
>>>>> in my
>>>>> abstract.
>>>>
>>>> Yes, we have enough expertise on the panel to know our open source
>>>> packages. Anything we didn't know, we looked up. However we can't look
>>>> up something omitted from an abstract...
>>>>
>>>>> Someone had a great abstract on big data, but it wasn't selected
>>>>> because it can be used with both open source software and closed source
>>>>> software, and it's not about open data specifically.
>>>>
>>>> An abstract that doesn't mention any open geospatial technology could
>>>> well be about doing analysis in ArcGIS or Oracle Spatial. Its not the
>>>> committee's job to second-guess the presenter or ask the presenter for
>>>> clarification - the abstract is space enough to provide clarity and
>>>> full details.
>>>>
>>>>> My personal opinion is
>>>>> that if the general public wants to see this talk, it should not
>>>>> matter if
>>>>> the abstract contains the words free or open.
>>>>
>>>> Again, we did not filter on the words. We took the totality of the
>>>> submission and checked appropriateness for the Free and Open Source
>>>> for Geospatial Conference, amongst the other criteria.
>>>>
>>>>> Also, if this is filtering would be done, it should be done *prior*
>>>>> to the
>>>>> community voting phase IMHO.
>>>>
>>>> Personal opinion: there's no point - the outcome will be the same, it
>>>> will just require a committee to review everything before and after
>>>> the community voting. There were very few inappropriate submissions.
>>>>
>>>>> Can the selection committee elaborate on the approach they used?
>>>>
>>>> I think we've discussed this at great lengths on this and other
>>>> mailing lists. Basically: First pass: include community vote top 100.
>>>> Second pass: include committee vote top 100 (giving us ~130 included).
>>>> Discuss, eliminate anything inappropriate. Next pass: include lower
>>>> ranked community votes. Next: lower ranked committee votes. Check for
>>>> multiple submissions, similarities with workshop sessions, and make a
>>>> decision on near-duplicates (which may include rejections, choices, or
>>>> mergers). Keep going until coffee runs out or all slots filled. We did
>>>> not run out of coffee.
>>>>
>>>> I think fuller details will be posted to the lessons learned/cookbook
>>>> wiki pages.
>>>>
>>>> Barry
>>>
>>>
>>>
>>> _______________________________________________
>>> Foss4g2013 mailing list
>>> Foss4g2013 at lists.osgeo.org
>>> http://lists.osgeo.org/cgi-bin/mailman/listinfo/foss4g2013
>>
>> _______________________________________________
>> Foss4g2013 mailing list
>> Foss4g2013 at lists.osgeo.org
>> http://lists.osgeo.org/cgi-bin/mailman/listinfo/foss4g2013