[zeromq-dev] subports

Pieter Hintjens ph at imatix.com
Wed Jul 27 15:48:42 CEST 2011


On Wed, Jul 27, 2011 at 3:41 PM, Martin Sustrik <sustrik at 250bpm.com> wrote:

> Ok. What's the exact problem with the proposed solution? AFAICS it's just a
> more generic version of what you are describing, thus it should cover all
> associated use cases. Am I missing something?

Here's my immediate reaction as user: too complex, not complete, kind
of seems to miss the target.

That could be mitigated by wrapping it up as an inprocess bridge but
that seems incomplete too. The problem will come down to mapping
socket patterns over this. That leads either to "reinvent the wheel"
like VTX is doing, or "do it in the core", which seems the most
accurate solution.

This problem hits all tunneling solutions afaics, unless you know better.

-Pieter



More information about the zeromq-dev mailing list