ruote log_2010-11-02

2010-11-02 06:45:04 utc jmettraux codebeaker: hello
2010-11-02 07:03:23 utc jmettraux codebeaker: you usually don't need to ask the workflow where you are, it tells you by giving you some work
2010-11-02 07:03:26 utc jmettraux it's a workflow engine
2010-11-02 07:04:15 utc jmettraux but if you want to know anyway, you can do engine.process(wfid).position
2010-11-02 07:04:48 utc jmettraux it will list the participants to which it has attributed workitems
2010-11-02 08:14:37 utc codebeaker thanks jmettraux (sorry, was sleeping)
2010-11-02 08:14:45 utc jmettraux you're welcome
2010-11-02 08:14:59 utc codebeaker … that makes sense, also then the right way is to see when something requires work from the "user" and show it to them
2010-11-02 10:32:38 utc hassox good evening
2010-11-02 10:52:01 utc hassox hey guys
2010-11-02 10:52:14 utc hassox is anyone here who could help me work out a storage_participant query?
2010-11-02 10:59:47 utc tosch_le hassox: i'll be back in about an hour.
2010-11-02 10:59:53 utc hassox kk
2010-11-02 11:00:02 utc hassox I'll prolly be gone by then... but I'll remember it ;)
2010-11-02 11:00:06 utc tosch_le perhaps i may help quickly
2010-11-02 11:00:07 utc hassox thanx
2010-11-02 11:00:14 utc hassox kk lemme gist
2010-11-02 11:00:58 utc hassox tosch_le: just pm'd the gist
2010-11-02 11:00:59 utc hassox so
2010-11-02 11:01:07 utc hassox there is a participant incidents
2010-11-02 11:01:25 utc hassox and when I launch, I pass it :incident_id => 1234
2010-11-02 11:01:51 utc tosch_le ok
2010-11-02 11:01:56 utc hassox lets say I have a Rails controller action that accepts the witness reports
2010-11-02 11:02:04 utc hassox and when all witnesses have reported
2010-11-02 11:02:08 utc hassox I want to reply to the engine
2010-11-02 11:02:12 utc hassox so
2010-11-02 11:02:29 utc hassox I'm guessing the first thing I need to do is to find the work item for that participant with that command
2010-11-02 11:02:50 utc hassox engine.storage_participant.query(:participant => "incidents", "${f:params:command}" => "/gather_witness_reports}", "${f:incident_id}" => 14)
2010-11-02 11:03:03 utc hassox sorry.. the incident_id should be 1234
2010-11-02 11:03:08 utc hassox copy / paste ;)
2010-11-02 11:03:23 utc hassox would that query find the right thing that I could then reply to the engine with?
2010-11-02 11:03:41 utc hassox ACTION is a bit lost with long running processes
2010-11-02 11:03:46 utc tosch_le no, dollar substitution is for process definitions only afair
2010-11-02 11:03:52 utc hassox so
2010-11-02 11:04:08 utc tosch_le let me have a look at the query method
2010-11-02 11:04:22 utc hassox engine.storage_participant.query(:participant => "incidents", "f:params.command" => "/gather_witness_reports", "f.incident_id" => 14)
2010-11-02 11:04:32 utc hassox engine.storage_participant.query(:participant => "incidents", "f.params.command" => "/gather_witness_reports", "f.incident_id" => 14)
2010-11-02 11:04:35 utc hassox I meant that one
2010-11-02 11:06:39 utc tosch_le engine.storage_participant.query(:participant => "incidents", "incident_id" => 14)
2010-11-02 11:07:04 utc tosch_le i don't know if you may check for an entry in params with that method
2010-11-02 11:07:22 utc hassox but, that could grab either the witnesses sequence, or the employments_involved sequence couldn't it?
2010-11-02 11:08:05 utc tosch_le yes. you'll have to select the appropriate items yourself
2010-11-02 11:08:42 utc tosch_le hmm. i'm not happy with my answer. please drop a post in the ml, that could be a feature request
2010-11-02 11:08:53 utc hassox sure
2010-11-02 11:08:57 utc hassox thanx :)
2010-11-02 11:09:30 utc tosch_le anyway, perhaps it's more easy to put the fei into the form
2010-11-02 11:10:02 utc tosch_le s/fei/wfid
2010-11-02 11:10:37 utc hassox but, how do I get hold of it?
2010-11-02 11:10:40 utc hassox oh
2010-11-02 11:10:41 utc hassox wfid
2010-11-02 11:10:48 utc hassox I was talking to john the other day
2010-11-02 11:11:16 utc hassox and he said that there could be many fei's per wfid (one for each point in the workflow)
2010-11-02 11:11:32 utc tosch_le yes, sure
2010-11-02 11:12:03 utc tosch_le let me repeat for better understanding: you want to show a form?
2010-11-02 11:12:14 utc tosch_le for gathering reports
2010-11-02 11:12:28 utc hassox so
2010-11-02 11:12:47 utc hassox I'm not really sure how to deal with long running processes (i.e. that don't respond immediately)
2010-11-02 11:12:59 utc hassox so the idea is that when the incident occurs, a process is launched
2010-11-02 11:13:19 utc hassox at some point, witnesses are identified
2010-11-02 11:13:42 utc hassox that's the record_witnesses command
2010-11-02 11:13:54 utc hassox (I don't know how to get that done yet but it's the same answer)
2010-11-02 11:13:55 utc tosch_le ok, i'd recommend to name your participants after the actions
2010-11-02 11:14:14 utc hassox so, when I get to the gather_witness_reports command
2010-11-02 11:14:21 utc tosch_le if you use the catchall, each will point to the storage participant
2010-11-02 11:14:30 utc hassox I will email the witnesses asking them to follow a link, to complete a report
2010-11-02 11:14:33 utc tosch_le will make querying easier
2010-11-02 11:14:43 utc hassox I don't understand
2010-11-02 11:15:14 utc tosch_le let me re-start
2010-11-02 11:15:27 utc tosch_le incidents :command => "/gather_witness_reports"
2010-11-02 11:15:38 utc tosch_le will block until all reports are gathered?
2010-11-02 11:16:36 utc tosch_le you have to functionalities here: 1) send mails to the witnesses and 2) wait for (and process/save) answers
2010-11-02 11:16:56 utc tosch_le 1) is done immediately, 2) blocks
2010-11-02 11:17:29 utc hassox well, until all witness reports are gatherd
2010-11-02 11:17:35 utc hassox correct
2010-11-02 11:17:36 utc tosch_le i'd use two different participants for that
2010-11-02 11:17:57 utc tosch_le 1) sends the mails, 2) is the storage participant
2010-11-02 11:21:08 utc hassox that makes sense
2010-11-02 11:21:13 utc hassox I guess I've overloaded them
2010-11-02 11:21:45 utc tosch_le i forked your gist
2010-11-02 11:22:00 utc tosch_le just to give you the idea what i meant with renaming the participant
2010-11-02 11:22:31 utc tosch_le the storage participant may be used as a catchall, it will catch all workitems not caught by another participant registered
2010-11-02 11:22:46 utc tosch_le you may use the participant name in the query method
2010-11-02 11:23:07 utc tosch_le engine.storage_participant.query(:participant => "incidents", "incident_id" => 14, 'participant' => 'witnesses')
2010-11-02 11:23:41 utc tosch_le but querying after the 'command' param should work, too. i just don't know how atm
2010-11-02 11:24:07 utc hassox looking
2010-11-02 11:25:03 utc tosch_le (i mean: it should work not in a sense that it already does ;-) )
2010-11-02 11:25:08 utc hassox so
2010-11-02 11:25:21 utc hassox don't provide a participant and just let it go through to the keeper so to speak?
2010-11-02 11:25:33 utc hassox then query and reply manually via the storage participant?
2010-11-02 11:25:43 utc tosch_le sorry, have to leave now. will be back in an hour
2010-11-02 11:25:44 utc hassox ACTION brain melt
2010-11-02 11:25:46 utc hassox np
2010-11-02 11:25:52 utc hassox would love to continue this tomorrow :)
2010-11-02 11:25:57 utc tosch_le but yes, that's the way i'd go
2010-11-02 11:26:02 utc tosch_le your welcome!
2010-11-02 11:26:54 utc hassox tat's really interesting
2010-11-02 11:26:58 utc hassox hey jmettraux
2010-11-02 11:27:10 utc jmettraux hassox: hello !
2010-11-02 11:27:11 utc hassox I didn't realise that's what a catchall is used for
2010-11-02 11:27:15 utc hassox how are you mate?
2010-11-02 11:27:21 utc jmettraux fine and you ?
2010-11-02 11:27:38 utc jmettraux is tosch_le helping you out ?
2010-11-02 11:27:53 utc hassox indeed
2010-11-02 11:28:01 utc hassox ACTION just made my first post to the ml :D
2010-11-02 11:28:07 utc hassox talking about this one
2010-11-02 11:28:30 utc jmettraux just "always allow"ed you
2010-11-02 11:28:30 utc hassox trying to wrap my head around processes that don't respond immediately
2010-11-02 11:28:36 utc hassox thanx :D
2010-11-02 11:29:11 utc jmettraux OK, dinner time for me, let me answer your email right after it
2010-11-02 11:29:37 utc hassox thanx heaps dude :D
2010-11-02 11:29:41 utc hassox there's no great rush
2010-11-02 11:29:45 utc hassox ACTION is about to o to bed ;)
2010-11-02 11:29:59 utc jmettraux ok, great
2010-11-02 11:30:03 utc hassox first working day in 4 days tomorrow ;)
2010-11-02 11:30:24 utc jmettraux :-)
2010-11-02 12:52:31 utc jmettraux tosch_le: hello, many thanks for helping hassox
2010-11-02 12:53:10 utc tosch_le never mind, had some troubles figuring out what to say in what way, but it was fun
2010-11-02 12:54:11 utc tosch_le he wanted to query the storage participant for a field within the params field. is this possible?
2010-11-02 12:54:26 utc tosch_le after having a look at the code, i'd say no.
2010-11-02 12:55:57 utc jmettraux depends on the storage
2010-11-02 12:56:32 utc jmettraux IIRC, we've arranged only for querying top level fields
2010-11-02 12:57:24 utc tosch_le it's ok for me, but i'll have to tell hassox
2010-11-02 12:57:48 utc jmettraux he fired an email to the mailing list, IIRC it includes that question
2010-11-02 12:58:16 utc tosch_le yep.
2010-11-02 16:20:21 utc jmettraux answered to hassox : http://groups.google.com/group/openwferu-users/browse_thread/thread/2b76b1b7ac024b52