Proposal #3339: Add (+>) as a synonym for mappend

[Stephen Tetley]

On 14 August 2011 15:25, Yitzchak Gale <gale at sefer.org> wrote:
> Thomas Schilling wrote:

>
> The only reason Monoid is currently so much more common is
> because the semigroups package is fairly new. Semigroups are
> simple and quite ubiquitous. We should be strongly encouraging
> their use, not discouraging it by creating an awkward namespace
> clash.

Semigroups seem a lot less useful than monoids.

They might be more ubiquitous, but there seems to be less you can do
with them - e.g. sconcat in the Semigroup class is quite convoluted to
avoid emptiness; a Writer-like thing without zero would be very
strange (presumably it would actually have to be a state monad?).

If semigroups don't support an operationally useful API, are they
worth having from the documentary perspective - is it valuable to
identify that you have semigroup "things" in your program when you
have to use type specific operations to do anything useful on them?