[TUSCANY-3670] Review use of Domain URIs/APIs

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Won't Fix
• Affects Version/s: Java-SCA-2.0-M5
• Fix Version/s: Java-SCA-2.x
• Component/s: SCA Java Runtime
• Environment: All
• Assignee: Unassigned
• Reporter: Simon Laws
• Created: 07/Sep/10 13:33
• Updated: 23/May/11 09:57
• Resolved: 23/May/11 09:57

Description

Primarily a placeholder against which to track issues related to the way the we now use Domain URIs and the APIs that allow us to configuration and access domains/nodes in 2.x.

Initially I'm looking at whether we followed up with what we said we were going to do. For example, http://www.mail-archive.com/[e-mail]/msg11205.html

Also there are some things that I don't understand about how domain URIs are used. For example,

NodeFactory.newInstance(String configuration)
SCAClientFactory.newInstance(URI domainURI)

Seems odd that they take different things

Also we have the original Node API which is Tuscany specific and focused on initial configuration of nodes. We also have the domain-node API which is spec based (not spec provided) which focuses on incrementally configuring a node. They both seen useful and it would be good to combine in some way so we don't have two node implementations. Or at least if we do they follow some sensible naming/module name pattern so that people can work out what they're for.

Sub-Tasks

• domain registry URI not parsed into properties and hence don't configure Hazelcast registry: Closed