users@jersey.java.net

[Jersey] Guidance for shallow/deep resource GET

From: <thomas.black_at_accenture.com>
Date: Thu, 24 Feb 2011 16:06:14 +0000 (GMT)

We have several resources that are composed of 20+ fields that we store
as name value pairs. Two of which are large email cut-and-paste jobs,
which make thier size up to 500K in size.

The idea is we want to drive a grid style list with only a core set of
properties, maybe 5 or so, enough to identiy the resource. This
shallow fetch would keep our UI and network happy. Only after a user
drills down to a specific resource would we do a deep fetch to retrieve
most of the other name/value properties. A specific extra click would
fetch the 500K field.

Our initial idea is that each name value pair has a numeric level where
"0" mean the shallow fields, "10" means all properties of that level or
less, and "99" is a byName fetch of a known large field.

Does this sound reasonable? Any guidance or completed projects that
took a different approach to keep performance snappy and make the
expensive GETs just-in-time?