Re: selecting Debian mentors for the mentor summit

[Daniel Pocock <daniel@pocock.pro>]

On 09/08/18 16:47, Lucas Kanashiro wrote:
> 
> 
> On 08/09/2018 06:16 AM, Daniel Pocock wrote:
>>
>> On 09/08/18 00:57, Lucas Kanashiro wrote:
>>> I do not know if you understood what I was saying... I am not arguing
>>> that people that do not maintain packages do not worthwhile (I know the
>>> Debian constitution). I was claiming that even mentors can be newcomers
>>> in our project, and if want to keep them tighten to our community, a
>>> conference such as Debconf is the right place.
>>>
>> Yes, I fully agree we need to make it easy for people to attend Debian
>> events and events where Debian has a stronger presence (reminder: the
>> DPL offers[1] USD 100 to anyone for travel to a BSP, contact[2] him for
>> details)
>>
>> For the summit, however, we still need to decide whether to use some or
>> all of the criteria that have come up or just to fall back on the
>> previous algorithm: randomly pick from those people who never attended
>> the summit before.
>>
>> >From what I can see, there are now 6 candidates (Dashamir, Lucas,
>> Jaminy, Urvika, Chirayu, Milena) and none of them attended before, is
>> that correct?
>>
>> To tweak the previous algorithm to ensure diversity, we could start by
>> randomly selecting one of the 3 women and then make a random selection
>> from the remaining 5 candidates.  As Molly is the only remaining person
>> named in the last delegation, perhaps we can ask her to roll the dice?
> 
> You can ensure diversity for sure (1 slot for women). But again, pick
> randomly people that may not understand what Debian is to represent the
> project might not be a good idea. I think that I already made my point,
> I'll leave this decision up to admins.
> 

I'm not saying this is the best approach, only that if we don't get a
lot of feedback from people it may be easier to simply fall back on the
default (or something very close to it) from previous years.

If somebody wants to volunteer to take the possible criteria and work
them into a new selection procedure now is the time.

For example, this could be a nice little script in the outreach-admin
repo on Salsa:

1. putting all the candidates in a table, with boolean values for each
of the criteria

2. making a set of all candidate pairs that satisfy the selection goals
(every mentor should appear in at least one pair)

3. randomly choosing one of the pairs from the set

Step (2) is the part of the process that may need some thinking about.

Regards,

Daniel