Thanks everyone for engaging in this I deeply appreciate ya’ll’s time and effort.
Replies in line.
On Wednesday, February 21, 2024 1:27 PM rest-dev
rest-dev-bounces@xxxxxxxxxxx On Behalf Of Markus Karg MK via rest-devsaid:
Sent:
Subject: [EXTERNAL] Re: [rest-dev] EE11 Platform call update
MK> Speaking just for myself, I need to say that I do have a problem
MK> with *removing* JAXB. I have to maintain a software that on one
MK> hand shall use latest features, but on the other hand MUST stick
MK> with JAXB for backwards compatibility. It is OK for me that JAXB
MK> would become *optional* so I can simply drop it into the mix on my
MK> own. But *removing* instead of *optional* is a no-go for me and
MK> completely WAY off Java's long-term backwards compatibility
MK> benefit. Having said that, it is not our charter to necessarily do
MK> exactly what the platform people dream of just because they would
MK> like so, in particular not *forcefully remove* things just because
MK> they want to spare maintenance costs for their paid products or
MK> whatever benefit they like to have from *removing* it.
HI Markus, I'm glad you brought this up. In my role as release co-coordinator for Jakarta EE 11, I assert that one artifact delivered by the community served by the rest-dev
list is a component spec and ratifying compatible implementation for Jakarta EE 11. For discussion, label this as ASSERT_REST_EE_11.
Other artifacts delivered by this community are not of interest to me in that role.
Can someone please validate ASSERT_REST_EE_11?
MK> Due to that I am -1 for *forcefully removing* JAXB, but I am +1
MK> for having it *optional* in the sense of "application vendors can
MK> put it on the classpath if needed". As a side effect, there is not
MK> justification for a 4.0.
In my role as release co-coordinator for Jakarta EE 11 I want to draw your attention to the decision made in the Jakarta EE specification committee (aka "the platform people
dream of") that predates my accepting the role of release co-coordinator for Jakarta EE 11.
On 2023-02-08, the Jakarta EE Specification Committee (not "the platform people") SpecCmte: wrote:
SpecCmte> "Draft resolution for optional features in the Platform
SpecCmte> specification (also see issue #54), to be put forward for
SpecCmte> ballot in the 02/08 call:
https://github.com/jakartaee/specification-committee/issues/54
SpecCmte> "An individual specification can have optional features,
SpecCmte> however when a component specification is included in the
SpecCmte> Platform and Web Profile, and Core Profile an optional
SpecCmte> feature must be explicitly declared as required, otherwise
SpecCmte> it is not required. This requirement is noted in the
SpecCmte> Platform specification.”
SpecCmte> The resolution was put forward by Scott Stark and seconded by Tom Watson"
SpecCmte> Kenji Kazumura - Fujitsu [+1]
SpecCmte> Tom Watson - IBM - Emily Jiang [+1]
SpecCmte> Ed Bratt - Oracle - Dmitry Kornilov [+1]
SpecCmte> Andrew Pielage - Payara - Petr Aubrecht [+1]
SpecCmte> David Blevins - Tomitribe - Jean-Louis Monteiro, Cesar Hernandez [absent]
SpecCmte> Ivar Grimstad - PMC Representative [+1]
SpecCmte> Marcelo Ancelmo - Participant Member - Abraham Marin-Perez [0]
SpecCmte> Werner Keil - Committer Member [absent]
SpecCmte> Scott Stark - Red Hat - Scott Marlow Enterprise Member [+1]
SpecCmte> Zhai Luchao - Shandong Cvicse Middleware Co. - Enterprise Member [+1]
SpecCmte> 7 +1 / 1 0 / 0 -1 / 2 absent
SpecCmte> This resolution passes
Source:
https://docs.google.com/document/d/1e6FNVgcKJfVz4TZ6qSPs4QbcOZvkIdhu6jGwgk7txlw/edit
Now, coming back to what Markus said about the charter of this community (of which I am not a member, but I am a Java Champion and have a passionate stake in the success of this
community.)
MK> it is not our charter to necessarily do exactly what the platform
MK> people dream of just because they would like so, in particular not
MK> *forcefully remove* things just because they want to spare
MK> maintenance costs for their paid products or whatever benefit they
MK> like to have from *removing* it.
Where is it documented what is in and not in this charter?
On Wednesday, February 21, 2024 1:45 PM rest-dev
rest-dev-bounces@xxxxxxxxxxx On Behalf Of Jim Krueger JK via rest-devsaid:
Subject: [EXTERNAL] Re: [rest-dev] EE11 Platform call update
JK> My [wording] was perhaps unfortunate. JAXB itself is not being
JK> “removed”. It will still be optional. It just is not part of the
JK> platform any longer. Liberty, for example, will still support
JK> JAXB with EE11. What we’ve been asked to do is to remove the
JK> dependency that Jakarta Rest currently has on JAXB. This will
JK> just involve removing the Link.JaxbLink and Link.JaxbAdapter inner
JK> classes that were deprecated in Jakarta Rest 3.1.
On Wednesday, February 21, 2024 3:25 PM rest-dev
rest-dev-bounces@xxxxxxxxxxx On Behalf Of Arjan Tijms AT via rest-devsaid:
Subject: [EXTERNAL] Re: [rest-dev] EE11 Platform call update
AT> I think the major issue is that JAXB cannot be in Jakarta REST,
AT> since Jakarta REST is the core HTTP engine in Jakarta EE (in the
AT> Core Profile). The core profile does not have JAXB included. This
AT> is the same reason that Jakarta _expression_ Language had to be
AT> removed from CDI.
I believe this is also technically correct. When I imagine sitting down to churn out the
ENTIRELY NON NORMATIVE EE 11 M2 maven artifacts, I am pretty sure this will blow up in my face.
AT> Newer products like Quarkus can't support those older things like
AT> Jakarta _expression_ Language and JAXB (at least not out of the box,
AT> I know there are various third party add-ons that bring it back
AT> from the Quarkiverse).
AT> Traditional servers like GlassFish, Open Liberty, WildFly etc
AT> should have no problems with supplying JAXB, and build-your-own
AT> stacks using e.g. Jersey and Tomcat should also not have any
AT> problems to continue support for JAXB.
On Wednesday, February 21, 2024 5:17 PM rest-dev
rest-dev-bounces@xxxxxxxxxxx On Behalf Of James Perkins JP via rest-devsaid:
Subject: [EXTERNAL] Re: [rest-dev] EE11 Platform call update
JP> The removal of XML Binding was already done in the release-4.0
JP> branch [1]. I wouldn't think we'd be doing anything beyond
JP> removing the module dependency and the Link.JaxbAdatper and
JP> Link.JaxbLink [2]. Is this what you mean or do you mean
JP> specifically with the specification document removing the optional
JP> XML Binding specification?
JP> FWIW my understanding is we only want to remove the module
JP> dependency and the two deprecated types.
JP> [1]:
https://github.com/jakartaee/rest/commit/ae1f6ef7f4db9957079b8c383041a4537f33f35b#diff-82115d28404407d63171c556f5b3f2ad04c1e63624d9bda468795757176667cc
JP> [2]:
https://github.com/jakartaee/rest/commit/ae1f6ef7f4db9957079b8c383041a4537f33f35b#diff-bdc4aba48dff256058507610393cbb181c1acccfe46ca35a5dce0b314017e314
From: rest-dev
rest-dev-bounces@xxxxxxxxxxx On Behalf Of Santiago Pericasgeertsen SPG> viOn Thursday, February 22, 2024 11:12 AM est-devsaid:
Subject: Re: [rest-dev] [External] : Re: EE11 Platform call update
SPG> That’s the only compile-time dependency that I remember, so that
SPG> is basically it AFAICT.
That is a relief.
Thanks
Ed
Release co-coordinator for Jakarta EE 11
| edburns@xxxxxxxxxxxxx | office: +1 954 727 1095
| Calendar Booking:
https://aka.ms/meetedburns
|
| Please don't feel obliged to read or reply to this e-mail outside
| of your normal working hours.
|
| Reply anonymously to this email: https://purl.oclc.org/NET/edburns/contact