better explanation for optional fk values in rels

[davewood]

No description provided.

[ribasushi]

The original discussion on this was:

<davewood> can the "problem" be simplified to: "if your foreign key column is not nullable you want join_type => left for complex resultsets to work correctly"?
<ribasushi> davewood: itym s/not nullable/nullable/
<ribasushi> and "for relatonship traversal to work consistently in all situations" (nothing complex about the resultset I gave as example)
<ribasushi> davewood: but yes, that'd be much better - convert it to a final patch and show again?

The current patch however simply rewords two sentences, while keeping the original confusing semantic structure intact. Is there a reason you didn't go for the much simpler explanation as in the IRC log above?
I could apply the patch as-is, but it imho makes no difference whatsoever wrt the problem you were originally trying to solve.

[ribasushi]

So, all patches in this PR have been incorporated except for davewood@5911039b

<ribasushi> davewood: about the cookbook entry
<ribasushi> davewood: it needs no changes, however on first read it seems to imply that DBIC will take care of the nulling and all that jazz, which is not true
<ribasushi> that is - it will work as described *only* if you deploy() from DBIC
<ribasushi> davewood: I think something needs added to this effect
<ribasushi> davewood: on a tangent - I actually have correctly working code to do dbic-side on_X emulation, but it has no tests (part of a larger thing)
<ribasushi> davewood: if you want to spin it into a separate dist - I would be very very happy
<ribasushi> davewood: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?p=dbsrgits/DBIx-Class-Shadow.git;a=blob;f=lib/DBIx/Class/Relationship/Cascade/Rekey.pm;hb=HEAD
<davewood> ribasushi: I wouldnt mind helping out with unit tests, will get back to you in 2014.

<ribasushi> davewood: so what do you think wrt the cookbook thing?
<davewood> ribasushi: the least would be to add a note saying that this only works if deployed from DBIC and I can also add the SQL needed for those working the other way around.
<ribasushi> davewood: cool, I will keep it in the branch for the time being then, poke me when you have a writeup

Currently commit is in https://github.com/dbsrgits/dbix-class/tree/topic/nullable_belongsto_recipe

[ribasushi]

@davewood I will close this PR for the time being. The branch is still available https://github.com/dbsrgits/dbix-class/tree/topic/nullable_belongsto_recipe for further improvement as described in #37 (comment)

Looking forward to another PR with the discussed changes in place.

Cheers and thanks!