Hi Andrew and Ignasi, Thanks for your thoughts. As Andrew said, I think the much safer approach would be to handle this at the vcloud level. So, I've implemented a request filter for VCloudVersionsAsyncClient and it works fine.
I hope this way is ok? On Wed, Jan 1, 2014 at 9:09 PM, Ignasi Barrera <ignasi.barr...@gmail.com>wrote: > I think it is a safe patch. According to the spec [1], not sending the > Accept header means the client accepts all media types, and the */* value > for the header means the same. > > I also added that in the OkHttp driver pull request [2]. Nirmal, could you > open a pull request, as suggested by Andrew, with a method similar to the > one I added in my pull request, so the behavior can be overridden/extended > if needed? > > Ignasi > > [1] http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html > [2] > https://github.com/jclouds/jclouds/pull/232/files#diff-0669ed207d502fa8adc5114f1c4e8167R89 > El 01/01/2014 16:17, "Andrew Phillips" <andr...@apache.org> escribió: > > I think this is of no harm, what do you think? >>> >> >> I think the easiest way to check this is to open a pull request and see >> if we get any test failures ;-) If everything looks OK and there are no >> concerns from committers, we could consider this. >> >> But unless "*/*" is the 'standard' default for the 'Accept' header if >> nothing else is specified, I suspect the correct fix for this will be to >> ensure that the vcloud provider sets this property appropriately. >> >> Have you been able to try a fix along those lines (see e.g. [1]) >> >> ap >> >> [1] https://github.com/jclouds/jclouds/blob/master/apis/ >> vcloud/src/main/java/org/jclouds/vcloud/internal/VCloudLoginApi.java#L42 >> > -- Thanks & regards, Nirmal Senior Software Engineer- Platform Technologies Team, WSO2 Inc. Mobile: +94715779733 Blog: http://nirmalfdo.blogspot.com/
