(Migrated) Near-future development roadmap

(This message has been automatically imported from the retired mailing list)


here’s a couple of things that will be implemented or refactored in
upcoming months.

Our development happens in GitHub and once ready the features below will
be announced in this list and made available in source code form:


  • Making Redis an optional dependency - we will still support it but it
    will not be a core requirement anymore

  • Plugins

  • ZeroMQ MDP http://rfc.zeromq.org/spec:7/

  • Guaranteed delivery

  • WebSockets

  • Making ZeroMQ, AMQP and WebSphere MQ connectors run in greenlets
    instead of connector processes (already done for ZeroMQ)

There will be also more installation options in Amazon - Beanstalk and
AMI - added to https://zato.io/docs/admin/guide/install/index.html

There are more ideas and they will be seen to as well, such as LDAP,
SAP, SMSC, channel caches and others but the above will be the focus of
the next few months.

In particular, plugins will let one create new channels and outgoing
connections without waiting for new Zato releases and share them through
a repository of plugins.

In fact, some new connection types and a few existing ones will be first
rewritten as built-in plugins, that is, sharing the same underlying
codebase that user-defined ones will use.

A patch 2.0.8 release is also planned to cover issues raised since 2.0.7
was put out.

If you have any particular wishes or ideas, please speak up so they can
be discussed and taken into account.

best regards,

On 14/06/16 15:13, Samuel Rose wrote:

Thank you Dariusz and Zato team. BST is likely slated for further down
the road, yes?

Hi Sam,

BST is definitely slated for delivery in 3.0 and I think the only core
feature that is still to be added looks are timeouts, i.e. perform a
forced transition if some other transition did not happen with N

There is also the idea to add a Visio/OmniGraffle-like GUI to BST so
that processes could be designed in web-admin but I’m not clear at this
point how much effort it would need, basically what the state of
JavaScript libraries required for it is.

Is there anything else you were thinking of?