Copied user alias as well.
-Jeffrey
On 7/9/14 3:17 PM, "Jeffrey Nguyen (jeffrngu)" wrote:
>Hi Devs,
>
>I'm looking for neutron API documented here
>http://jclouds.apache.org/reference/javadoc/1.7.x/org/jclouds/openstack/ne
>utron/v2_0/NeutronApi.html. I checked the code in jClouds release 1.7
Jeremy, thank you and congratulations on completing the new
openstack-swift provider and I look forward to jclouds promoting it as
the preferred Swift interface in 1.8. We duplicate effort by continuing
to fix bugs and add features on the legacy provider and I hope we can
cease, or at least signif
Hi Devs,
I'm looking for neutron API documented here
http://jclouds.apache.org/reference/javadoc/1.7.x/org/jclouds/openstack/neutron/v2_0/NeutronApi.html.
I checked the code in jClouds release 1.7.3 but didn't find it. Is this
only available in jClouds-labs or future release of jClouds?
Th
I was more concerned about domain objects that might contain large numbers of
optional elements. It seems such cases would need either Optional or @Nullable?
From: Andrew Phillips [aphill...@qrmedia.com]
Sent: Wednesday, July 09, 2014 3:39 PM
To: dev@jcloud
Also, while 3 valued logic is great, it isn't supported that well in
Java (until Some/None/Option in Java 8, which helps some).
jclouds is quite fond of Guava's Optional, but at the API level we
mainly use it to allow access to extension APIs that are, well,
optional.
I'd agree that we d
Howdy all,
Also, while 3 valued logic is great, it isn't supported that well in Java
(until Some/None/Option in Java 8, which helps some). It is also common to use
checked exceptions as a 3rd value, but that also has its perils, in my opinion
[1].
As mostly a user of JClouds, I'd prefer a runt
I generally prefer using primitives where possible, and even more in return
types (what does a returned null mean?). Using primitives is in general
easier for client applicarions and provide more clear semantics.
I really think that avoiding nulls [1] is an extremely helpful practice and
we should
and... there is a such thing as " Three-valued logic" which can be
handful some cases, so I think API should care about semantics...
On 07/09/2014 07:59 PM, Jeremy Daggett wrote:
No particular reason, just for consistency. ;)
On 7/9/14, 9:40 AM, "Andrew Phillips" wrote:
We need the APIs to b
I think there are some edge cases where this might be problematic.
Say you get a resource JSON back from a service and deserialize it into an
object, where certain domain object fields have not been returned by the
service (someField not included in the response).
Then boolean getSomeField() shou
No particular reason, just for consistency. ;)
On 7/9/14, 9:40 AM, "Andrew Phillips" wrote:
>> We need the APIs to be consistent.
>
>Any particular technical reason for this? Personally, I would be
>tempted to say "boolean" when it can't be null, and @Nullable Boolean
>otherwise, but that's just
We need the APIs to be consistent.
Any particular technical reason for this? Personally, I would be
tempted to say "boolean" when it can't be null, and @Nullable Boolean
otherwise, but that's just a somewhat context-free style preference.
ap
Hi devs,
I have noticed a mix of the types returned by many of the APIs in the jclouds
codebase and need your help. How do others feel about standardizing on
returning primitive types, versus object types for all APIs:
boolean someMethod(); ? <- Most APIs return primitives
Boolean someMethod();
12 matches
Mail list logo