Re: PL/pgSQL 2

On Wed, Sep 3, 2014 at 10:07 AM, Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com> wrote:
> When you use name plpgsql2 you say, so plpgsql2 is successor plpgsql. It is
> very hard to accept it. So any other name is not problem for me - like
> plpgsql-safe-subset or something else

plpgsql2 *is* the successor of plpgsql, that's why it has a "2" in the name.
Anything which is very different from plpgsql should instead get a
different name.

For all new users, having a convenient shorthand (plpgsql2) for
enabling what ever the
project think is the best-practice collection of pragmas, is a simple
and efficient way
of helping new users to get the best possible behaviour of the
language, when starting
from scratch.
It also simplifies communication among developers, when they talk
about code written in plpgsql2,
they will all eventually know what they mean, instead of having to
describe what collection of pragmas
they use in their code. That also simplifies code examples, but most
importantly, one does not have
to declare all the pragmas for each function, or worry about the
pragmas in the config file will ever change.

Once we have agreed upon plpgsql2, then it will be a constant, and
never break compatibility,
and that's a good thing. Then we can all write new code according to
the updated specs and
hopefully we will not need a plpgsql3 until year 2030.