[WIP] Dark crystal

[m4gpi]

See #952

[ameba23]

@KGibb8 in terms of changing custody.js to use the query forwardRequest.pull.forOwnShards, i think we might have to make some changes to the data structure you were putting together (store and records). i think it would maybe be easier if they were indexed by 'secret owner' rather than the id of the shard message. but i wanted to check with you before changing it. see commented out code in the last commit.

[m4gpi]

We can't do it like that if we want to display all patchwork shards you've been asked to be the custodian of.

First we have to display a list of shards you hold, that contain an attachment (relevant to Patchwork), then we need to search for forwardRequests that match up to those specific shards (we cross-reference using secretOwner on the request field, with the shard author).

Otherwise this view: https://user-images.githubusercontent.com/17786217/54701493-05893380-4b2d-11e9-9b5d-e5158f50a97c.jpg won't work

UPDATE: So looking at the forOwnShards, it does exactly that, nice work! :) But the data has to be rendered shard by shard for the view I've designed... My main concern is making sure we've got an array at the end of it that looks like the fake dataset I've generated. We can easily reduce the data returned by scuttle.forwardRequest.pull.forOwnShards() using transform in the collect callback, so I'm fine to index by feedId. But we need to make sure that scuttle.forwardRequest.pull.forOwnShards() returns the relevant shard data too, unless we're opening multiple pull streams....