Santiago,
it is hard to tell whether hypermedia is not used because the API is not complete, or whether it is not complete because it is not used… anyways, it was just stats from my side. Personally (not speaking for my audience now) I'm good with the plan, but would love to see CDI improved, too, if there is time for that. On the other hand I doubt that it makes big sense to write a spec and THEN add an implementation to all products; it might be the more wise solution to write the spec if at least two products already HAVE such a feature (independent of which feature we talk about).
-Markus
From: Santiago Pericasgeertsen [mailto:santiago.pericasgeertsen_at_oracle.com]
Sent: Mittwoch, 7. Oktober 2015 21:16
To: jsr370-experts_at_jax-rs-spec.java.net
Subject: Re: [jax-rs-spec users] Proposed Plan
Hi Markus,
Just one clarification, the MVC/JSON-B bullet is just an “alignment”, which typically means very little work for us. It’s quite different from CDI and Hypermedia in that sense.
Our data (mostly coming from Evangelists) suggest that hypermedia is not that popular in real systems (unfortunately!), and given that basic support is already available, we feel it is a feature whose development can wait until more feedback is collected from real-world users.
Thanks for the feedback.
— Santiago
On Oct 6, 2015, at 2:11 PM, Markus KARG <markus_at_headcrashing.eu> wrote:
Dear Specification Leads,
Dear Experts,
I am glad that finally there now will be at least _some_ progress on our JSR after months of apparent stagnancy and unclear status, and that the RI will provide at least _some_ of the proposed features!
This said, I need to tell you that after several presentations about JAX-RS 2.1 at JUGs and conferences in Europe, I noticed that the features proposed in the original specification request are clearly noticed and widely expected by the audiences -- all of them, not just numbers 1 through 4. My personal statistics is different from your plans:
Particular high interest (i. e. significantly more questions) had been shown on the following topics:
* CDI
* Hypermedia (i. e. "full" REST)
Particular low interest (i. e. significantly less questions) had been shown on the following topics:
* MVC
* JSON-B
(If you are interested in more community stats I could do short ballot after my talk on JAX-RS at JUG Saint Gallen On October 29).
This not necessarily means that MVN and JSON-B is expected to be low priority, but it indicates that CDI and Hypermedia should not get discarted from the plan.
To sum up, I would say "yes" for a shorter plan, but "no" for only that particular four items.
-Markus
From: Santiago Pericasgeertsen [mailto:santiago.pericasgeertsen_at_oracle.com]
Sent: Montag, 5. Oktober 2015 15:31
To: jsr370-experts_at_jax-rs-spec.java.net
Subject: Proposed Plan
Dear Experts,
After a discussion with Marek, and mainly due to the slowness of the JSR thus far, I’d like to propose that we prioritize on what we consider the main features of the 2.1 release (in hindsight, we may have included too many features for a dot release!):
(1) Reactive API
(2) Non-blocking I/O
(3) SSE
(4) Alignment with MVC and JSON-B
We feel that if we focus on these features, we should be able to produce an EDR before the end of the calendar year, and possibly deliver a complete spec within 9 months; alignment with the other specs will naturally depend on their schedule. Features not on this list would be considered for a future version of JAX-RS, either 2.2 or a 3.0.
It is time that we show some progress to the community and the JCP, and we feel this more focused plan will get us there. We hope you will support us. We have actually done some work over the last weeks around Reactive and NIO that we plan share shortly.
— Santiago & Marek