When should we remove Javax support? Current Target: 2025!

[tandraschko]

As currently discussed in https://github.com/primefaces/primefaces/issues/11877: When should we remove Javax support? for 16.0? 17.0?

It seems that the usage is quite low actually: https://groups.google.com/g/jakartaee-ambassadors/c/-1cU-XkT810?pli=1

[melloware]

I think until that usage increases to a significant level we keep support of javax.

[tandraschko]

Currently only 7% of the users are on javax

can someone check which containers with javax are already EOL?

[melloware]

that seems REALLY low. I know NONE of my clients are on Jakarta yet. So those results may be skewed?

[tandraschko]

maybe i think that the most people that voted are propably the users, which are interessted in new jakarta stuff anyway

I think we should check the EOL of javax containers i dont think that we as OS community should support javax forever also it doesnt make sense to stay forever on old java/javaee but always upgrade to newest PF, we have PRO/Elite for this stuff

[christophs78]

"my" public sector organisation is currently moving to spring boot 3.x. (from 2.x) spring boot 2.x (=javax) is EOL since ~11/2023. (there is still some commerical support for 2.x) But i know the combination of JSF with spring (boot) is unusal.

Tomcat 9 will propably be there for a few more years. (2027? they just set 8.5 to EOL) Open Liberty maybe the same.

Some other public sector organisations i know already moved vom JavaEE to JakartaEE some time ago.

Some other maybe plan to use there legaxy JavaEE / JSF applications as long as viable and then do something new maybe based on JS-frontends. Maybe reuse some of the backend-stuff if refactoring this to REST / GraphQL / ... is viable.

Waiting until 2027+ to phase out javax on our side IMO is to far away.

"Things" like Hibernate 5 (last javax-verion) get very limited support. Similar things for other core parts of JavaEE. Don´t know how much effort Redhat, IBM & Co still invest to give JBoss, Open Liberty, still limited life-support. Or they hope there are not much CVE´s and Co they need to adress.

[tandraschko]

TomEE 8.0.x (javax) is or will be EOL soon, too

i think we could target 2025 to remove it

[melloware]

That sounds like a good plan to me.

[Bukama]

7% usage of javax seems very low to me, but I can't proof it either right or wrong. You could also argue that you drop it now as PF 14 needs Java 11 which does not brings javax build in.

[escay]

People on Weblogic 14.1.1.0.0 running Java 11 are stuck on using javax / JEE8. They need to migrate to an alternative application server. Weblogic 14.1.1.0.0 end of life is tier 1 support is March 2025 / tier 2 support is planned until 2028 (https://docs.oracle.com/cd/F52330_01/reference_guides/insbridge_hdwr_sfwr_guide/Content/Guides_Sales/Insbridge%20Hardware%20Software%20Specs/End%20of%20Support%20Schedule.htm). And no Weblogic 15 is planned, support issue https://support.oracle.com/knowledge/Middleware/2935815_1.html states: "There is no specific plan for JakartaEE9 or 10" or it is way too late and people migrated to alternatives.

Here a target of 2025 seems also reasonable.

TomEE people can upgrade to TomEE (Jakarta 10) or newer and do not need javax anymore. Payara people can upgrade to Payara 6 (Jakarta 10) or newer and do not need javax anymore. Glassfish people can upgrade to Glassfish 7 (Jakarta 10) or newer and do not need javax anymore. Same for Open Liberty and others: there are Jakarta 10 upgrade options available for over 2 years already.

I do not know what the overhead is for supporting javax. But thank you for letting me still using latest PF in Weblogic.

[tandraschko]

:/ i would really suggest to migrate to some other AS then