[
https://issues.apache.org/jira/browse/JCLOUDS-172?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14202341#comment-14202341
]
Adrian Cole commented on JCLOUDS-172:
-------------------------------------
one last major change.
Currently, for at least machine types and instances, we set the jclouds id to
encoded zoneName + instanceName.
This has created a ripple of complexity including types that save off this
information (ex. MachineTypeInZone, InstanceInZone) as well make it harder to
construct server side filters for operations like listByIds.
I will make a change that converts the ids used in the compute service to just
be the selfLink. I expect that between this and using aggregatedList, it will
save a lot of code and complexity. Plus be cheaper on the api.
Feel free to update non-computeservice types while I'm working on this, or help
convert expect tests to mock tests, etc.
> "Graduate" GCE to core
> ----------------------
>
> Key: JCLOUDS-172
> URL: https://issues.apache.org/jira/browse/JCLOUDS-172
> Project: jclouds
> Issue Type: Task
> Components: jclouds-labs-google
> Affects Versions: 1.8.0
> Reporter: Andrew Bayer
> Assignee: Adrian Cole
> Priority: Critical
>
> We should really get GCE to a "finished" state and include it in 1.7.0.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
