[Qgis-psc] Voting for grants

Thu Jun 4 04:44:19 PDT 2020

Hi all,

We discussed this issue in this week's PSC meeting and have come to the
following decision:

Instead of using the old system of ranked voting for top 6 proposals (6
points to the highest ranked to 1 point to the lowest ranked), this year
each voting member will be asked to select his/her favorite 6 proposals
and each of these proposals gets 1 point.

Concerning ideas to limit who can vote on which proposal (e.g. no voting
on proposals by "own company"), we have decided against implementing any
such rules. The reasoning follows many of the concerns raised in this
tread: it is highly complicated to define who gets to vote on what and
this could be a potentially big source of conflict every year.
Furthermore, the PSC thinks that what we have seen so far shows that
voting members can be trusted to vote in the best interests of the QGIS
project.

We already shared the results of the PSC's review of formal proposal
requirements, as well as the decision to fund two documentation
proposals directly from the documentation budget in an earlier email.
The remaining eligible proposals result in a total requested budget of
EUR 45,400 while the grant programme budget has EUR 40,000 available.
Therefore, we can expect a high accept rate.

Regards,

Anita

On 02.06.2020 12:46, Andreas Neumann wrote:
> Hi all,
>
> Interesting discussion - and difficult to find a good solution.
>
> I agree with Martin that the voting process where the first selected
> entry gets 6 points, the next 5, etc. is broken. Just like the PSC
> voting (I mean the mechanism to decide who will be on the PSC) is
> broken (which has a very similar setup). A person, or a proposal isn't
> (at least normally) not 6 times as useful to the project than others.
>
> Both need to be overhauled and made more transparent. I like the idea
> where all proposals could be scored and then we will accept the ones
> with the highest scores as long until the grant proposal budget is
> allowing acceptance.
>
> I would also favor the idea that the PSC could select out some of the
> proposals and finance them through the regular budget - e.g. we have a
> documentation budget that is never fully used. If we get entries in
> the QGIS grant proposal around documentation, PSC should be able to
> accept these right away through the documentation budget, thus leaving
> more room toward other grant proposals. Infrastructure is another
> category where we already have a budget for.
>
> We have a PSC meeting tonight at 8pm (middle European Summer time).
> This topic is certainly on the agenda, if someone would like to join
> the discussion you are welcome to join the PSC meeting.
>
> Greetings,
>
> Andreas
>
> Am 02.06.20 um 11:35 schrieb Martin Dobias:
>> Hi all
>>
>> Very good discussion and good points are being made.
>>
>> Like Denis, I also have various concerns about grants this year:
>> - which grants are actually features (which we wanted to exclude)
>> - what is the overall value that
>> - how good is the plan
>> - whether the budget looks appropriate
>> - how much trust we have in the developer
>>
>> Voting members need to somehow consider all of the above (and more!)
>> for all proposals and then create a list of 6(?) which seem to be the
>> best. Which is incredibly hard, especially with so many projects we
>> have this year. I feel uneasy that as a voting member can only assign
>> some rating to 6(?) proposals, and with very limited options - the
>> proposal I choose as the first will get 6 points, the last one gets
>> just one point - it feels like as if the first proposal was six times
>> better...
>>
>> What I would like to suggest is whether we could move towards a
>> scoring process that is common in public bids... QGIS.org would define
>> criteria that would be scored 0 to 10 and voting members would score
>> those criteria separately, for example:
>> 1. fit to the grant call - something that looks like a feature if we
>> don't want features could be scored low
>> 2. value to the project - is this something really important that will
>> help lots of users / developers? Or something more obscure that few
>> users would benefit from?
>> 3. proposal quality - is it clear what would be done and how it would
>> be done? Or is the proposal brief and missing important details?
>> 4. price - is the expected budget reasonable?
>> 5. author's track of record - this could be scored low if the author
>> is unknown to the project or it is unclear if they have the right
>> skills
>>
>> The criteria could have either equal weights or adjusted according to
>> what we think is the right balance.
>>
>> This system would place a greater burden on voting members to analyze
>> the proposals and score them, but it may help fairness of the whole
>> process and various of the biases voting members have would dissolve
>> as there would be more data coming from each voting member.
>>
>> Or if we skip the criteria, but at least allow voting members to score
>> _all_ proposals, each with a single number 0-10 ?
>>
>> Regards
>> Martin
>>
>> On Tue, Jun 2, 2020 at 10:06 AM Denis Rouzaud
>> <denis.rouzaud at gmail.com> wrote:
>>> Hi all,
>>>
>>> It's indeed an interesting discussion and I feel very well concerned…
>>> I have been personally influenced by the corporate voting during
>>> last session.
>>>
>>> While I understand this concerns, I would remind that the voting
>>> member is something personal and not tight to a company.
>>> I have mixed feelings. As far as I can see, elected voting members
>>> have proven real engagement towards QGIS project.
>>>
>>> Again, voting rights are personal. But the money allowed money for
>>> grants will generally go to the company rather than the individual.
>>> Thereby, I would rather try to restrict the number of allowed grants
>>> per entity rather than restricting the voting rights.
>>>
>>> But we'll fall into the same point Matthias was mentionning earlier:
>>> what's an entity. Some devs are more or less coupled, without being
>>> hold literally in a company.
>>>
>>> So, my order of preference of action would be
>>> 1: try to make voting members aware of the conflict of interest and
>>> trust them
>>> 2: if needed, restrict per company
>>>
>>> Now, looking at this year proposal, I am far more concerned about
>>> what we propose to vote on rather than the voting process:
>>>
>>> 1. Many of the grants are features, while we call for a refactoring
>>> / infrastructure round
>>> 2. Some are made by unknown people from the community, and I have
>>> mixed feelings about this (openess vs rewarding of involvment)
>>>
>>> I don't have a perfect solution to propose but I think we should
>>> maybe re-think the process more globally. What is the goal of the
>>> grants proposal?
>>> * Is it to fund projects that are difficult to fund otherwise? But,
>>> then shouldn't it be a kind of ongoing budget?
>>> * Are we trying to attract devs?
>>> * Are we trying to replace a crowd-funding operations for nice-to-have?
>>>
>>> Best wishes,
>>> Denis
>>>
>>> Le mar. 2 juin 2020 à 09:16, Matthias Kuhn <matthias at opengis.ch> a
>>> écrit :
>>>> Hi all
>>>>
>>>> On 5/30/20 1:21 AM, Nyall Dawson wrote:
>>>>> In general, I'd propose that we consider introducing (in the official
>>>>> statutes) a limit of one-voting-member-per-organisation. This would
>>>>> bring the community voting membership into line with the user group
>>>>> membership, where user groups have one single voting member who
>>>>> represents the group's view as a single vote. This would also limit
>>>>> the potential for (god forbid) a "hostile takeover" situation,
>>>>> where a
>>>>> coalition of organisations (commercial or user group) could dominate
>>>>> voting. (I think it would be wise to apply the same
>>>>> one-member-per-org
>>>>> limit to PSC/board membership too!).
>>>> A very good discussion.
>>>>
>>>> How would a one-member-per organisation rule work in reality?
>>>>
>>>> If someone is a very active and respected member of the community with
>>>> all the skills required to judge and discuss a QEP (or other
>>>> motion) and
>>>> has voting rights. And then gets employed by a company which
>>>> already has
>>>> someone with voting rights, will his voting rights be withdrawn? I
>>>> would
>>>> expect that this will have an effect on his choice of employer and
>>>> could
>>>> impact his engagement within community processes.
>>>>
>>>> Also, how would we deal with loosely coupled groups of developers
>>>> which
>>>> are legally not an organisation/company but still share business,
>>>> communication and opinions?
>>>>
>>>> Best regards
>>>>
>>>> Matthias
>>>>
>>>> _______________________________________________
>>>> Qgis-psc mailing list
>>>> Qgis-psc at lists.osgeo.org
>>>> https://lists.osgeo.org/mailman/listinfo/qgis-psc
>>> _______________________________________________
>>> Qgis-psc mailing list
>>> Qgis-psc at lists.osgeo.org
>>> https://lists.osgeo.org/mailman/listinfo/qgis-psc
>> _______________________________________________
>> Qgis-psc mailing list
>> Qgis-psc at lists.osgeo.org
>> https://lists.osgeo.org/mailman/listinfo/qgis-psc
> _______________________________________________
> Qgis-psc mailing list
> Qgis-psc at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/qgis-psc