Thanks for the pointer John. Why is Blob in core and Compute in Labs? What is the criteria for leaving labs?
Ross Microsoft Open Technologies, Inc. A subsidiary of Microsoft Corporation -----Original Message----- From: John Kew [mailto:john....@socrata.com] Sent: Wednesday, February 4, 2015 3:47 PM To: dev@jclouds.apache.org Subject: Re: Contributing Azure support There is also an existing azureblob implementation which we currently use in production: https://github.com/jclouds/jclouds/tree/master/providers/azureblob I took Adrian's original implementation and extended it for larger blob types. I don't particularly care about the implementation, and the notion of formal support pleases me (in fact I think I also pushed MS to add JClouds on roadmap since we were on the azure board), but it would be wise to check it out to verify compatibility. -John On Wed, Feb 4, 2015 at 3:11 PM, Ross Gardler (MS OPEN TECH) < ross.gard...@microsoft.com> wrote: > Hi JClouds folks, > > We'd like to contribute Azure support to JClouds. We already have a > fairly complete implementation of compute and storage and plan to work > further on this implementation. However, this implementation is built > against an the Azure Java SDK rather than the REST API. From > conversations with some folks familiar with JClouds I understand that > this is not the usual way of doing things here. > > There are good reasons why we've implemented it this way. We can go > into those if you want, but it will make no real difference since > that's the working code we have available. It is feasible that we'd > consider moving this over to a REST API if active community members > are going to help with the work, but if it's just our ongoing > contributions then continuing with the work already done is our > preference. Rather than discussing it endlessly I'd propose a code talks > approach. > > Having poked around a fair bit it looks to me like your model is to > have new code worked on in a lab at jclouds-labs. Given the SDK vs > REST API approach it would seem sensible to start this effort as a lab > anyway. That is using a lab would give time for both code maturity and > for the community to evaluate the approach. > > I also note that there is some existing work on Azure compute in > https://github.com/jclouds/jclouds-labs/tree/master/azurecompute it > would make sense to have our code in the repo so that those working on > the two different implementations can avoid duplicating effort > wherever possible and community members have access to both > implementations in order to provide feedback. > > How do you want to proceed? Shall we go ahead and make the > contribution via that repository? This will provide an early > opportunity for feedback and code commentary. > > (NB both myself and Eduard Koller are ASF committers so the legal side > of the process is clear, just asking about the community contribution > process here in JClouds land) > > Ross > >
