Re: Call for consensus for the pagination design (ISSUE-42)

Hi elf

>If you use URI of the view as subject of hydra:member
></an-issue/comments?page=3> hydra:member <some_member_21>,<some_member_22> 
>.
>Where would information below come from?
></an-issue/comments> hydra:member <some_member_21>, <some_member_22> .
Either by calling it directly or or indirectly by implying that if 
</an-issue/comments?page=3> isViewOf </an-issue/comments>, view's members 
are members of the collection the view presents. Collection is a union of 
all it's views.

>I guess hydra:isViewOf would need to define it somehow...
Exactly
>, or it doesn't
>even matter to have those triples if each view describes the very same
>resource? eg. foaf:Group
I'm not sure if I understand. All views are depending on it's "parent" 
collection, which is same resource in this case.

>As long as we can clearly recreate underlying relations between
>resources which hydra:Collection describes (e.g. foaf:member of
>foaf:Group) and know how to stitch the graph together from all those API
>responses IMO we stay on the right path.
Actually I never wanted to build a whole graph from all API requests - I 
wasn't considering such a case. In ReST API usually I don't need whole 
graph - I receive a resource, modify it's state and send it back. Stiching 
something from various API calls may be considered a violation of the ReST 
approach in my opinion.

>Maybe we all can take few days to write code which will page members of
>Hydra CG and this way compare various approaches based on actual
>implementation experience?
Interesting - I'm in the middle of a POC angularJS client application that 
will utilize a Hydra API description, but I'm still not at the pagers yet.

Regards

Karol 

Received on Monday, 26 October 2015 21:26:57 UTC