[email protected] writes:

> On Tuesday, September 27, 2016 10:26:50 PM Arne Babenhauserheide wrote:
>> [email protected] writes:
>> > On Tuesday, September 27, 2016 11:43:26 AM Arne Babenhauserheide wrote:
>> >> Don’t bother too much. I have all I found in csvs which are much easier
>> >> to work with.
>> >> 
>> >> I uploaded a tarball with the votes and cost-estimates along with an
>> > 
>> >> evaluation tool:
>> > Arne,
>> > 
>> > it is not fun for me to volunteer days on something while neglecting my
>> > personal duties only to wake up to someone saying "throw away your
>> > existing
>> > work, I finished it already!"; especially considering I had asked you to
>> > please not do this.
>> 
>> You asked me not to do this while the poll is open.
>> 
>> I asked explicitly whether it’s closed and only posted the results after
>> you said that the poll is closed.
>
> Well it seems you misunderstood me then, because that is not what I said or 
> at 
> least intended to say:
>
> [Tuesday, September 20, 2016] [10:43:07 PM]
> <xor-freenet> ArneBab: i slightly agree with nextgens. it'd be nice if you 
> could wait for the main evaluation of poll results which i'm working on 
> before 
> you publish your own evaluation.   i mean you're free to speak whatever you 
> want but it's sort of derailing things if we constantly get threads on devl 
> which want to push the poll off its railways... we all know that those 
> threads 
> end up in huge discussions which block people from reaching real decisons

Oh, yes.
I misunderstood that, otherwise I’d have already answered
differently back then. Sorry for that.

However a discussion about the results will be necessary, because the
process was ill-defined from the beginning, and attempts to fix that
were met with hostility. My evaluation is not a try to make that
discussion harder, but to make it easier because it actually gets data.

Some of the things which were proposed and would have more or less fixed
the process:

- clearly limiting the tasks to what can be done with 25k and allowing
  only 1 point per task (no fractions), so people would have to declare
  tasks as not valuable for the 25k to increase the weight of something
  else. That would still have allowed for strategic voting (the first
  two or three items might be bogus), but the downvotes would have been
  meaningful.

- declaring the process as a means to assess the wishes of the
  developers and community, not as a vote. That would have removed most
  incentives for strategic voting. It would have also allowed people to
  add tasks which were missing due to the initial definition of the poll
  as a means to spend the 25k, not as a means to create a larger
  roadmap.

- declaring up-front that the process is a means to create (or update) a
  roadmap. Then it needs refined non-overlapping tasks (without missing
  important tasks like fixing the installers or transport plugins), not
  just a list of things which could be done with 25k, and it means a
  well-defined voting method which is at least somewhat resistant
  against strategic voting.

If you want to get an idea of strategic voting: Imagine an FMS user
giving high value for short node references because he or she does not
want the money spent on part two of WoT and thinks that many other
people will give it high value, too - even though he or she does not
think that short node references have any real value.

Essentially strategic voting means that people choose the lesser evil or
*give up* what they actually want to *prevent* something they really do not
want.

So, to end this on a positive tone (and since it fits right now):

https://www.youtube.com/watch?v=Dv6-hc4pI5M
https://www.youtube.com/watch?v=RIiW2j4lugg

  Why vote for the lesser evil?
    Cthulhu for president!

Best wishes,
Arne
-- 
Unpolitisch sein
heißt politisch sein
ohne es zu merken

Attachment: signature.asc
Description: PGP signature

_______________________________________________
Devl mailing list
[email protected]
https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl

Reply via email to