[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [jakartaee-platform-dev] Platform popularity
|
Hi,
On Saturday, December 3, 2022, Vano Beridze <
vanuatoo@xxxxxxxxx> wrote:
Jakarta EE 10 is a great achievement but if you compare it to Java EE 8, there are hardly any revolutionary changes. Java EE 8 was released 5 years ago.
It’s undoubtedly true that EE 8 was a long time ago, and that the recent EE 10 release is the first real release since. No doubt indeed this long transition period has hurt the platform.
That said, the last major Spring release was a long time ago too. And, EE did continue in those years under the MP name (in a way). Between EE 8 and EE 10 a bunch of new APIs were released such as Config, Rest client, JWT etc.
Now that EE itself is “back”, it’s technically a matter of merging that MP work back. In my humble opinion, and I appreciate not everyone shares it, MP has served its purpose and the EE / MP split doesn’t do anything else than confuse users.
What we mostly do is to make sure platform changes don't affect existing user base, instead of thinking mostly how to become the popular again.
That’s not entirely the case. One of the major themes in EE 10 was to remove everything deprecated. Servlet, CDI and Faces did exactly that, and it will break the existing user base. Faces for instance completely removed JSP support, completely removed managed bean support, and removed dozens of classes, methods and fields throughout its API.
What I propose to do is to have just two profiles:
1. Full profile - which is there to support customers who plan to use legacy applications.
2. Core Profile - which is where the exciting staff happens.
Full profile had indeed been proposed to serve that purpose, along with a Legacy Profile.
We need to do identify specs which does not have future and completely stop investing in them.
Quite a number of specs have indeed been identified as such. The term we use for them is “stable specs”. The
jakarta.ee/specifications page however does an exceptionally poor job of making it clear which specs are stable and which are active.
This is a know concern, but very hard to address (the page in question is extraordinarily hard to edit).
We need to combine forces with MicroProfile community, identify the set of specs which are critical to our success and bring them to life under one name as soon as possible.
Actually, the EE and MP communities already largely overlap. It’s more like joining forces with ourselves, and be honest to ourselves about a split that’s not needed.
We need to start thinking how to create good onboarding environment for developers and stop relying on vendors to do that. Jakarta EE starter is the good place to expand, add more documentation, real use cases.
We need to stop thinking that because Spring uses small number of specs, Jakarta EE is standard. Spring will never be Jakarta EE compatible and they could easily remove Jakarta EE reliance in one year if they wanted to.
Not sure if that’s the case. I think there were some efforts to move away from Servlet, but customers and users don’t want that to happen, seemingly. I may be mistaken here though.
Kind regards,
Arjan Tijms
Other popular frameworks also are not Jakarta EE compatible and if we want them to be, there should be drastic changes in how we do things. With the current process and dynamics I'm afraid in 5 years there will be proprietary frameworks and Jakarta EE as the failed attempt to have standard in Enterprise Java.
Hi
I hardly imagine a situation where Jakarta EE will be chosen over Spring Boot or Quarkus if somebody is going to create a microservice.
It might help us a lot if you were able to give some specifics about the reasons why you can't imagine that.
Kind regards,
Arjan Tijms
jakartaee-platform-dev mailing list
jakartaee-platform-dev@eclipse.org
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/jakartaee-platform-dev
