Speaking just as a user of JEE/Jakarta, we saw the risk of using MP in production as it looked like its initial goals were accomplished with the Jakarta transition (preventing the stall of innovation). We figured any further innovation would likely go back into the Jakarta spec and development would slow down, which it did.
All to say is, I really appreciate everyone that worked on MicroProfile. I remember sitting in DevoXx UK when the idea was announced. I don't want anything to sound like I'm not thankful for all of the work and passion people put into the project. But as a user, I'd be fine with moving out without MP included in TomEE. I'm only one voice of many however, but I wanted to share this single data point. thank you, On Thu, Oct 24, 2024 at 6:25 PM David Blevins <david.blev...@gmail.com> wrote: > > On Oct 22, 2024, at 11:37 AM, Markus Jung <ju...@apache.org> wrote: > > > > Actually I’m a bit confused now how is TomEE 8.0.0-M3 the only version > listed on start.microprofile.io <http://start.microprofile.io/>? How can > that be? As far as I know TomEE 8 was never Java EE 8 certified. Also did > TomEE 9.x never claim MP 5.0 compatibility? > > Reasonably so. > > MicroProfile used to be a very lenient project with the aim of > accommodating everyone in the name of rapid experimentation and forward > movement. After the Eclipse Foundation forced the project to create a > Working Group, things just got slower, more strict and started to lose > participation. That loss of participation is why start.microprofile.io > is now so derelict. The strictness of the rules is also why we are not on > that site after TomEE 8.0.0-M3 and why we are not certified. > > TomEE 8.0.0 still used the Geronimo implementations of some MicroProfile > specs which were never up-to-date (our lack of resources) so it prevented > any certification. > > TomEE 9.0 switched to the SmallRye implementations that other vendors > use. We passed the Jakarta EE 9.1 WebProfile TCK and certified. We wanted > to pass the MicroProfile 6.0 TCK which was not yet final and would have > allowed us to claim certification on the day of the spec release which > would have been a first for us and a huge boon to morale. > > The other vendors in the MicroProfile group however wanted to update the > MP 6.0 spec to require Jakarta EE 10 Core Profile certification, despite > the fact that 1) none of the MP specs were leveraging EE 10 and 2) we had a > several month discussion and successful vote saying we would not increase > the required Java or Jakarta EE version unless there was a technical > feature in an MP spec that necessitated it. We attempted to make the case > that this would needlessly exclude TomEE who was just 9.1 certified and > could be ready for MP 6.0 certification in time for the final MP 6.0. The > unanimous and overwhelming response from the other MP WG members (IBM, > Microsoft, Oracle, etc) was that it would be negative for MicroProfile and > they grouped together to ensure we could not certify with MP 6.0 on EE > 9.1. Again, the MP 6.0 TCK fully ran and could be passed on EE 9.1. > > The MP 6.0 spec went final with EE 10 Core Profile certification as a > requirement, despite no technical dependence, and so we are blocked from MP > certification till we certify on EE 10 or later. Red Hat also was not able > to certify as MP 6.0 as they had some issues with the Metrics spec and felt > it should be optional. We agreed and wanted to see them certified as they > were the implementation used ship every MP spec to date and it would seem > only reasonable to accommodate them, but the project now overall more > strict said no. So no TomEE and no Quarkus implementations of MP 6.0... > for the good of MicroProfile. > > That's what it looks like when you go from community-over-code to > code-over-community. > > The whole thing was a tough pill to swallow. We worked so hard on TomEE 9 > with so few resources. > > We later did go and file a certification request for TomEE 9.0 on MP 5.0, > but as you note start.microprofle.io is basically abandoned without an > owner, so we're still not there. > > > -David > > -- Jonathan | exabr...@gmail.com Pessimists, see a jar as half empty. Optimists, in contrast, see it as half full. Engineers, of course, understand the glass is twice as big as it needs to be.
