18:00 <cookieboyfan> http://stackoverflow.com/questions/22117856/datamapper-into-string we need help anyone????

22:14 <cvkline> Hey, question... if I have a class Schedule that includes "belongs_to :carrier" then how do I create a new schedule that includes which carrier it belongs to?

22:15 <cvkline> Schedule.create(name: "northbound", carrier: { name: "Local Bus Company" }) doesn't work; it creates a new Carrier rather than find the existing row with that name.

22:16 <cvkline> And c = Carrier.first(name: "Local Bus Company") ; Schedule.create(name: "northbound", carrier: c) just blows up.

22:17 <cvkline> I guess I'm unclear on how I create new child objects of a parent class that refer to an existing parent.

22:21 <cvkline> I just answered my own question...

22:21 <cvkline> I was doing c = Carrier.all(name: "Local Bus Company") which returns an array of Carriers, not a single one, even if only one result happens.

22:22 <cvkline> So after that, Schedule.create(name: "northbound", carrier: c[0]) works.

22:22 <cvkline> And of course so does using .first instead of .all

22:52 <mbj> cvkline: carrier = Carrier.find ...

22:52 <mbj> carrier.schedules.create(.....)

22:52 <mbj> Should be the "most clean" version.

22:55 <cvkline> Ohhh, I understand.

22:55 <cvkline> Starting at the top of the tree.

22:55 <cvkline> I'm porting a "traditional" SQL-based application over, and it's hard to keep thinking in this object-oriented way when it's all been SQL select statements.

22:55 <cvkline> Thanks

22:57 <mbj> cvkline: np, have fun!

22:58 <mbj> cvkline: When working with DM (Or any other ORM) you should think in terms of the domain, not in terms of the SQL. This holds for 99% of the cases, till the ORM abstraction leaks.

22:58 <cvkline> hm...

22:58 <cvkline> irb(main):023:0> ctest.schedules.new(name: "1100", capacity: 11)

22:58 <cvkline> NoMethodError: undefined method `schedules' for #<Enumerator: Carrier:find({:abbrev=>"ISH"})>

22:58 <cvkline> from (irb):23

22:58 <cvkline> from /usr/bin/irb:12:in `<main>'

22:58 <cvkline> mbj: I get that concept, it's just hard to keep the mental discipline. :)

22:58 <mbj> cvkline: Carrier.first

22:58 <mbj> cvkline: not Carrier.find, sorry. My dm-1 days are "long" over :D

22:58 <mbj> Carrier.find returns an enumerator (as you just saw).

23:00 <cvkline> Ha, yup, thanks.

23:00 <cvkline> I'm intrigued by what i'm seeing about dm-2 or "rom" I guess it's called now. But it seems far from ready for prime time.

23:01 <mbj> cvkline: heh

23:01 <mbj> cvkline: We are all to busy commercially.

23:01 <mbj> OSS work is a chicken egg problem.

23:01 <cvkline> I totally get how that goes.

23:01 <mbj> ROM would help my clients a LOT.

23:02 <mbj> If ROM would exist, I'd probably have more time for OSS.

23:02 <cvkline> As much as I've kept trying to learn "modern" web frameworks like this, I kept never having time. Now that I have a consulting gig to redo a crappy old PHP+MySQL spaghetti code application, I am way more motivated!

23:02 <mbj> ;)

23:02 <cvkline> Anyway, thanks again for your help.

23:02 <mbj> k, have fun!

23:03 <mbj> If you are interested in ROM, join the #rom-rb channel.

23:03 <mbj> Sometimes there is activity ;)

23:03 <cvkline> heh ok