Standard of approval for library submissions (Was: Proposal: Add singleton function to Data.List module)

[Bertram Felgenhauer]

Taylor Fausak wrote:
> It has been a week since I submitted my proposal. During that time, 28
> people voted, with 16 expressing approval and 12 expressing
> disapproval. To everyone that voted so far: Thank you! You made for
> interesting discussion.

I thought the standard of approval for library proposals was a broad
consensus for adoptation, which we clearly do not have here. It was
supposed to be a reasoned decision based on arguments put forward by
proponents and opponents of the idea. There has always been a tendency
to express agreement or disagreement numerically, but recently these
agreements and disagreements have been treated as a voting system, and
I believe that's an unhealthy trend.

With a simple majority-based straw poll on the mailing list we risk
ending up with a ton of questionable additions to the core libraries
that have a handful of outspoken proponents and nobody who really cares;
the end result will be, predictably, a library full of functions that
nobody ever uses.

In the case of a dispute (which we have here), the decision should be
made by the maintainer(s), i.e. the Core Libraries Committee in this
case. As maintainers, they are also in a better position to address
concerns like consistency (which is hard to obtain by incrementally
adding small functions.)

Perhaps we should also consider that voting takes almost no effort,
whereas composing a coherent argument takes much time and effort.

Overall I believe we should go back to weighing arguments rather than
voting for library proposals.

Cheers,

Bertram