ruote tmp/log_2011-05-23.html

2011-05-23 08:01:48 utc biv hey jmettraux , you awake?
2011-05-23 08:02:05 utc jmettraux it's 1702, yes :-)
2011-05-23 08:02:10 utc jmettraux hello
2011-05-23 08:02:15 utc biv ha, sorry
2011-05-23 08:02:25 utc biv it's just an irc expression, you know
2011-05-23 08:02:31 utc biv hrm, can I ask a question?
2011-05-23 08:02:42 utc jmettraux just teasing you
2011-05-23 08:02:43 utc biv well, do you have time to answer one? :)
2011-05-23 08:02:44 utc jmettraux please do
2011-05-23 08:02:54 utc biv ok, let gist it
2011-05-23 08:05:18 utc biv https://gist.github.com/986379
2011-05-23 08:05:48 utc biv ok, currently the problem is that the participant can't add more items onto the field which the cursor iterates over
2011-05-23 08:06:00 utc biv the items there when the process def runs are downloaded, but not new ones
2011-05-23 08:06:16 utc jmettraux it's like it's getting "cleaning" after each rewind ?
2011-05-23 08:06:31 utc jmettraux "cleaned" ?
2011-05-23 08:06:40 utc biv hard to tell, but do you think the principle Im following is sound?
2011-05-23 08:07:15 utc biv it's like the extra fields added to the array
2011-05-23 08:07:33 utc biv on line 47
2011-05-23 08:07:49 utc biv either arent being added to the right variable or the cursor is ignoring them
2011-05-23 08:07:51 utc jmettraux looks OK, I'm wondering if ruote is not cleaning up the fields when doing the "rewind"
2011-05-23 08:08:09 utc jmettraux making them like they were when the workitem reached the cursor
2011-05-23 08:08:13 utc jmettraux double-checking
2011-05-23 08:08:42 utc biv yeh, it might be a state issue
2011-05-23 08:08:57 utc biv this is on 2.1.11 btw
2011-05-23 08:09:20 utc biv it's also my first use of the cursor statement, so I may have done something stupid
2011-05-23 08:10:01 utc jmettraux ah, I'm wrong, rewind does no resetting
2011-05-23 08:10:05 utc jmettraux checking your code again
2011-05-23 08:10:16 utc biv me too
2011-05-23 08:11:26 utc jmettraux are you using the master branch ?
2011-05-23 08:12:03 utc biv nah, just the gem
2011-05-23 08:15:02 utc jmettraux the concurrent_iterator is probably merging the workitems in a way you weren't expecting
2011-05-23 08:15:51 utc jmettraux by default, the first branch to reply wins
2011-05-23 08:16:46 utc jmettraux http://ruote.rubyforge.org/exp/concurrence.html you could use :merge_type => :isolate or :stack to prevent the overwriting
2011-05-23 08:17:27 utc jmettraux too bad, on master I just added a "union" merge_type https://github.com/jmettraux/ruote/commit/aa913d90eb0786c3a25550efad078f8ab61abd78
2011-05-23 08:18:23 utc biv sorry dude
2011-05-23 08:18:28 utc biv someone knocked on the door
2011-05-23 08:18:29 utc biv reading now
2011-05-23 08:18:56 utc biv hrm, well I tried upgrading to master but it broke, remember?
2011-05-23 08:19:03 utc biv so I'm just picking my battles one at a time right now
2011-05-23 08:19:13 utc biv reading about merge_type
2011-05-23 08:19:15 utc jmettraux I fixed the issue
2011-05-23 08:19:24 utc biv o really?
2011-05-23 08:19:35 utc biv ok, let me see if upgrading works then
2011-05-23 08:19:50 utc jmettraux https://github.com/jmettraux/ruote/commit/dee9ed6f9ebb2fdc2ef0140be7b2e00a2fd0898c
2011-05-23 08:20:20 utc jmettraux https://github.com/ngty/sourcify/issues/11
2011-05-23 08:20:56 utc biv cool
2011-05-23 08:23:03 utc biv ok cool
2011-05-23 08:23:12 utc biv ah an error
2011-05-23 08:23:15 utc biv after upgrading
2011-05-23 08:23:49 utc biv https://gist.github.com/986379#gistcomment-32020
2011-05-23 08:25:39 utc biv I have a bit more info, Ive run it again with --trace
2011-05-23 08:26:55 utc jmettraux could I have the error in raw ? The github autoformatting hurts my eyes :-(
2011-05-23 08:27:31 utc biv https://gist.github.com/986398
2011-05-23 08:27:43 utc biv it's borking on to_json
2011-05-23 08:27:44 utc jmettraux aah thanks
2011-05-23 08:27:55 utc jmettraux json_pure
2011-05-23 08:28:09 utc biv yup
2011-05-23 08:28:18 utc biv I might switch back to 2.1.11
2011-05-23 08:28:21 utc jmettraux always using that one ? not json or yajl-ruby ?
2011-05-23 08:28:31 utc biv I just use what the default is
2011-05-23 08:28:53 utc biv we have yajl-ruby in the gemfile
2011-05-23 08:28:57 utc biv I think ruote recommends it
2011-05-23 08:29:00 utc jmettraux yes
2011-05-23 08:29:02 utc biv but json_pure is being loaded anyway
2011-05-23 08:29:29 utc biv anyway, Im moving back to 2.1.11 and the original problem
2011-05-23 08:29:50 utc jmettraux too bad
2011-05-23 08:29:51 utc jmettraux ok
2011-05-23 08:30:07 utc biv ill come back to the upgrade
2011-05-23 08:30:14 utc biv but it's another effort and Im on the clock for this
2011-05-23 08:30:31 utc jmettraux ok, remember to require yajl before ruote for that one
2011-05-23 08:30:42 utc jmettraux so, the original issue
2011-05-23 08:30:53 utc jmettraux by default , when doing a concurrence or concurrent_iterator
2011-05-23 08:31:04 utc jmettraux the first branch to reply
2011-05-23 08:31:15 utc biv hrm, Ill just remove concurrence there for the timebeing
2011-05-23 08:31:18 utc biv simplify things
2011-05-23 08:31:20 utc jmettraux sees its workitem be used when the flow resumes after the concurrence
2011-05-23 08:31:25 utc jmettraux ok
2011-05-23 08:33:04 utc biv if it
2011-05-23 08:33:21 utc biv s running late for you I might pop in tomorrow morning, your 10 am or so, and work on this
2011-05-23 08:33:25 utc biv it's 6:30 pm here
2011-05-23 08:33:32 utc jmettraux it's not late for me
2011-05-23 08:33:39 utc biv ok
2011-05-23 08:34:00 utc jmettraux no worries
2011-05-23 08:37:20 utc biv ok, still no rewind (or change of state to the archive_catalog_integrations field from that line I pointed out)
2011-05-23 08:37:43 utc biv https://gist.github.com/986405
2011-05-23 08:39:39 utc biv let me check something here
2011-05-23 08:43:07 utc biv ok, so it is adding these extra ids to the field I want to rewind on
2011-05-23 08:43:28 utc biv let's pick this up tomorrrow, Im tired and Im going to grab dinner and spend more time pulling this aparr
2011-05-23 08:43:41 utc biv but you think it should be good to go without concurrence with the current strategy?
2011-05-23 08:50:32 utc jmettraux argh, sorry, scroll stuck
2011-05-23 08:50:38 utc jmettraux reading up now
2011-05-23 08:51:02 utc jmettraux the simpler the better
2011-05-23 08:51:09 utc jmettraux in doubt, add "p" statements
2011-05-23 09:08:45 utc biv no worries
2011-05-23 09:08:48 utc biv chat tomorrow mate
2011-05-23 09:08:50 utc biv im crashing now
2011-05-23 09:09:00 utc biv thanks, as always, for your advice
2011-05-23 09:10:30 utc jmettraux you're welcome, ciao !
2011-05-23 13:09:53 utc Defusal hi everyone
2011-05-23 13:10:20 utc Defusal argh, i dont suppose anyones around
2011-05-23 21:02:29 utc cloder hola
2011-05-23 22:50:52 utc jmettraux cloder: hello and welcome to #ruote
2011-05-23 22:51:02 utc cloder thanks
2011-05-23 22:51:35 utc cloder we're looking at ruote for a project of ours
2011-05-23 22:51:40 utc cloder just thought i'd hang out here :)
2011-05-23 22:54:16 utc jmettraux please make yourself comfortable
2011-05-23 22:54:29 utc cloder ACTION stretches out and grabs a beer
2011-05-23 22:57:03 utc jmettraux :-)