[zeromq-dev] Moving to 0MQ/3.0, yes or no?

[Pieter Hintjens]

On Sun, Apr 10, 2011 at 8:34 AM, Martin Sustrik <sustrik at 250bpm.com> wrote:

> In the recent month, the discussion about 0MQ/3.0 have shown that the
> radical "parting with the past design flaws" goal of the endeavour
> cannot be achieved. Every tweak has its user that will be hurt if its
> removed. That makes the 3.0 version more or less useless: The flaws
> won't be fixed while at the same time the backward incompatible changes
> will hurt everybody.

This seems like a false dichotomy.

The point about changing 0MQ and breaking compatibility is not that it
can't happen. Of course it can happen, and IMO most users are positive
about change in this project.

The key thing is to document and discuss up front, so that we make the
changes *accurately*. It's really painful to see designs that are
essentially good but simply haven't been properly discussed and
polished, being published, and then becoming difficult to fix.

We have a stable 2.1 branch, with space for some evolution in 2.2, and
it makes perfect sense to make brutal changes in 3.0, but only if
these are done right. And IMHO that means up front discussion, design
that involves users, and specs that can be critiqued before there is
released code.

Please do write specs up front, Martin, and get your users involved in
the thinking process, and then you'll know exactly how much change you
can make, how fast, and where.

-Pieter