On 16/04/15 16:54, Sergio Fernández wrote:
Hi everybody,
give the latest discussion, I think we have lost a bit the focus we had for
our first release. Therefore I'd like to start talk about it. Looking to
the mailing list and issue tracker, I'll try to summarize the current
status, and see what we can afford for 0.1*:
While we get GH/PR sorted out as a project process, let's discuss via
email/JIRA except for specifics of code review for specific PRs (i.e.
the programming details, not API design).
* I think we can continue with the plan an use the current
groupId:artifactId, which should solve COMMONSRDF-2.
+1 except see below for naming confusion.
* The big issue remains with blank nodes. For me it's fine to go out with
an api defining a simple contract, then is up to each implementation to
take care of special details. That means we shift COMMONSRDF-6 and
COMMONSRDF-14 to version 0.2.
+1 (go as is) for blank nodes
+0 for IRI and literals : ideally, define hashcode and equals as default
methods based on getters. Not a blocker.
(This is saying IRI and literals can pass between implementations, blank
nodes do not provide the necessary guarantees in 0.1.)
* I guess immutability requires further discussion, so also
shifting COMMONSRDF-7 to 0.2.
OK (the issue seems to mix contract requirements and implementation,
though any likely implementation has not very much room for manoeuvre as
far as I can see)
* Current version of the simple implementation is enough for testing the
api and showing some ideas to actual implementators.
OK - other than "simple" is becoming a misnomer!
In addition I'm working to get the site up (see INFRA-9260 for details),
but I'm confident to find a solution with the infra people in the next few
days.
Great - I think this is a blocker for a release, or at least announcing it.
The, I ask: is there any other thing that stop us for preparing release 0.1?
Thanks.
Best regards,
(*) with 0.1 I generically refer to both commons-rdf-api-0.1-incubating
and commons-rdf-simple-0.1.0-incubating artifacts
This is the right naming scheme but the new POMs disagree and have
commons-rdf-api-incubating-0.1
("-incubating" in the artifact name, not the version)
I guess that was not intended?
