On Fri, Mar 18, 2011 at 02:13:56PM -0700, Iuri G. wrote:
> 
> So I was able to narrow the problem. It does not appear to be just
> ruote sequel. I have been able to reproduce both: process in undefined
> state (between two participants) and nil.[] error I posted couple
> posts above.
> 
> Please take a look at those 2 files: https://gist.github.com/876827

Hello Iuri,

sorry for the late reply, I was in some kind of "road trip".


> Could you please let me know if it is a bug in ruote or maybe I am
> configuring it wrong.

If you subclass StorageParticipant, you'd better do something like

---8<---
def consume(workitem)
  p :hello
  super(workitem)
end
--->8---


> Also, I am still interested in whether reply and reply_to_engine are
> same things?

Ruote::StorageParticipant#reply != Ruote::LocalParticipant#reply

The first one makes sure to remove the workitem from the storage before 
replying to the engine

https://github.com/jmettraux/ruote/blob/9e960f1a8b9ede4233d3e7c2efc2cda6a742c69f/lib/ruote/part/storage_participant.rb#L118-136

Your consume method is shadowing the consume method of StorageParticipant, when 
calling reply(workitem), delete fails because the workitem has not been stored.

It's not a ruote bug, it's a misuse of StorageParticipant.


> Note: we are not using yajl gem. Instead we are using json and json-
> pure because yajl was conflicting/resulting in false positives when
> used with cucumber and capybara.

Too bad, yajl is so much faster than json (and of course json-pure).


I hope this helps, best regards,

John

-- 
you received this message because you are subscribed to the "ruote users" group.
to post : send email to [email protected]
to unsubscribe : send email to [email protected]
more options : http://groups.google.com/group/openwferu-users?hl=en

Reply via email to