Hi, Replies are in-line.
On Wed, Aug 26, 2015 at 1:19 PM, anthony shaw <[email protected]> wrote: > Hey > Just a couple of quick ones, I'm working on a few features for the > dimension data driver namely support for a loadbalancing driver and > extending the compute to support more functions and the newer API. > > 1. I understand the branching requirements are referencing a Jira ticket. > Do you create a ticket per feature and a feature branch? I've gone down > that path with 736 and LIBCLOUD-7372. Yes, we follow "feature branch" approach. In your case it would be ideal to submit two PRs - one for new loadbalancing driver and one for compute driver improvements. In case those features are related and share a lot of code (e.g. you are going to refactor some shared code in a common module), then it's OK to just open a single PR. > Unit tests. The existing tests were mock http results, which I've extended > coverage for. Do any of the other drivers utilise a real http request? What > is the expected level of coverage on the driver? > None of the drivers right now use real HTTP requests since testing against a real provider API is quite hard (especially emulating all the edge cases, etc.). As far as the expected coverage level goes - we have a no fixed number requirement, but the more the better :) > 3. What is the plan for a networking driver? Along the lines of the > openstack systems > I started to work on this in the past ( https://github.com/apache/libcloud/pull/345), but sadly I didn't have the time to finish it. It would be great if someone could pick it up. > > Ant > Sent from Outlook
