Handle VoteEvents that address the passage of _multiple_ Bills

[mileswwatkins]

There is a legislative procedure known as "consent agenda" or "consent calendar," that allows for quick passage of innocuous legislation. In this system, a single VoteEvent can be on the passage of multiple Bills; a 1:M relationship that Pupa doesn't currently allow for.

One solution would be to add related_entities to the modeling of VoteEvents, analogous to Bill.actions objects' related_entities. This feels slightly off, since a more native solution would perhaps allow VoteEvent.bill to accept both a Pupa ID/pseudo-ID (ie, current behavior) or an array of these IDs; however, that may be too big of a modeling change for an edge case such as this.

Real-world examples exist in SD and PA, at least: https://github.com/openstates/openstates/issues/2100

[fgregg]

Almost all the votes in the Chicago City Council are like this.

[mileswwatkins]

@fgregg, how do you currently model those? VoteEvent.extras? Or is there a Pupa-native method that @jamesturk and I missed?

[fgregg]

They get modeled as separate vote events.

[mileswwatkins]

But ideally they would be a single VoteEvent, correct? (Just not a blocking issue for either of our uses.)

[fgregg]

Sorry, yes. I meant to be supportive of this modeling change by giving example of this pattern.

I would definitely support a more accurate voteevent.

[jamesturk]

This should probably be an OCDEP not a pupa issue as we'd have to make the change there & in py-ocd As for the issue at hand, it'd be a pretty big breaking change unless we kept around VoteEvent.bill and had an optional VoteEvent.bills or similar. I am conflicted on that one as it'd let us keep backwards-compatibility & allow the 90% case to be optimized but would quite possibly get confusing.

…

On Mon, Jan 29, 2018 at 12:43 PM, Forest Gregg ***@***.***> wrote: Sorry, yes. I meant to be supportive of this modeling change by giving example of this pattern. I would definitely support a more accurate voteevent. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#308 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAAfYr9_Ir9BFdMVP5di7Ha_iraxOCLPks5tPgMogaJpZM4RxCx5> .