ruote tmp/log_2012-11-21.html

2012-11-21 01:59:37 utc jmettraux typedef: hello and welcome to #ruote
2012-11-21 01:59:46 utc typedef hi.
2012-11-21 01:59:53 utc typedef are you around or is that automated?
2012-11-21 02:00:15 utc jmettraux I'm around
2012-11-21 02:00:29 utc jmettraux ACTION gets ready for a Turing test
2012-11-21 02:00:35 utc typedef haha
2012-11-21 02:01:04 utc typedef i have a question about using "await"
2012-11-21 02:01:09 utc jmettraux ok
2012-11-21 02:01:27 utc typedef is it possible to "await" for multiple participants to complete? i'm having trouble getting that to work.
2012-11-21 02:01:57 utc typedef specifically, i'm trying to wait for all of them (as opposed to any of them)
2012-11-21 02:01:58 utc jmettraux isn't that rather best served by a "concurrence"?
2012-11-21 02:02:43 utc typedef sometimes maybe.
2012-11-21 02:03:09 utc typedef ultimately i'm trying to represent arbitrary DAGs, so maybe you can stop me right now if that isn't possible.
2012-11-21 02:04:13 utc jmettraux yes, it's going to be difficult
2012-11-21 02:04:27 utc typedef for example if I had a graph like A -> C <- B -> D, where task C depends on A and B, and D depends only on B. I'd like to be able to launch D as soon as B completes without waiting for A, but block C until both A and B complete.
2012-11-21 02:04:45 utc typedef (but C wouldn't have to wait for D)
2012-11-21 02:05:10 utc jmettraux feels like a job for a petri net
2012-11-21 02:07:31 utc jmettraux maybe
2012-11-21 02:10:15 utc typedef that could be useful.
2012-11-21 02:10:21 utc typedef let me see if i can use forget to get closer.
2012-11-21 03:01:14 utc jmettraux typedef: indeed, await could be help:
2012-11-21 03:13:16 utc typedef no, checking now
2012-11-21 03:14:13 utc typedef err, thanks :)
2012-11-21 03:14:37 utc typedef it would still help me if i could use multiple awaits, is that possible?
2012-11-21 03:15:20 utc jmettraux let me look at the doc
2012-11-21 03:15:56 utc typedef e.g., maybe this task d has to wait not only for b, but also some other task e completely unrelated to the rest
2012-11-21 03:18:04 utc jmettraux doing left_tag => [ 'a', 'b' ] does a OR
2012-11-21 03:18:09 utc jmettraux still looking
2012-11-21 03:19:04 utc jmettraux no, sorry, out of the box it only supports 1 thing or ORs the things
2012-11-21 03:19:13 utc jmettraux maybe you can craft a regex
2012-11-21 03:19:38 utc jmettraux argh, that wouldn't work
2012-11-21 03:19:42 utc jmettraux well
2012-11-21 03:21:25 utc jmettraux this is how you could do an AND:
2012-11-21 03:21:53 utc jmettraux quite a piece...
2012-11-21 03:23:58 utc mburnett wow, interesting
2012-11-21 03:24:29 utc mburnett i would not have tried that
2012-11-21 03:34:48 utc typedef that works and gets me much closer to my goal.
2012-11-21 03:35:30 utc typedef domo!
2012-11-21 03:44:05 utc typedef yep that seems to do it. participants can block on an arbitrary list of tagged parents by using await in a concurrent block. thanks again :)
2012-11-21 03:53:16 utc jmettraux you're welcome, douitashimashite
2012-11-21 11:42:52 utc lmarburger whoa. using concurrent awaits for dependencies is pretty awesome.
2012-11-21 11:44:23 utc lmarburger jmettraux: do you have anything published demonstrating some advanced things you can do with ruote? the examples in the documentation are great and now i'm really curious to see some larger ruote processes used in the real world.
2012-11-21 11:49:07 utc jmettraux lmarburger: hello, let me look up somethign
2012-11-21 11:49:18 utc jmettraux s/thign/thing/
2012-11-21 11:50:42 utc jmettraux lbt_ and team have put up
2012-11-21 11:51:13 utc lbt_ hey, jmettraux :)
2012-11-21 11:51:30 utc jmettraux there's
2012-11-21 11:52:03 utc jmettraux lbt_: oh cool, where is "La Mer" ?
2012-11-21 11:52:29 utc lbt_ jmettraux: inside the device and driving the Sailfish OS
2012-11-21 11:52:40 utc jmettraux ah, excellent
2012-11-21 11:53:10 utc lbt_ that phone UI was built with ruote driven QA processes
2012-11-21 11:53:32 utc lmarburger jmettraux: thanks!
2012-11-21 11:55:06 utc jmettraux you're welcome
2012-11-21 11:55:08 utc lbt_ lmarburger: I'm busy today/tomorrow but feel free to catch me if you need more info about that link
2012-11-21 11:56:02 utc lmarburger lbt_: thanks! do you have the raw ruote processes listed there or is it just the overview?
2012-11-21 11:56:57 utc lbt_ sec
2012-11-21 11:58:03 utc lbt_
2012-11-21 11:58:13 utc lbt_ explore that area
2012-11-21 11:58:37 utc lbt_
2012-11-21 11:58:56 utc lbt_ is a live view of the process used to promote middleware
2012-11-21 12:01:26 utc lmarburger that's massive. thanks!
2012-11-22 00:27:53 utc jmettraux mburnett: yes, I'd be very interested by your slides, my email is jmettraux at, I'll keep them for myself