Lucee 6.2 is our new LTS release, with 5.4 LTS having reached the End of Life at the end of 2025.
LTS means no new features, only critical important bug fixes and of course regressions.
There are already major architectural changes in 7, especially for Java integration which will not be backported.
As the next LTS will be Lucee 7, personally I would recommend jumping to 7.0 for any new upgrades, rather than 6.2, as there is a lot of nice stuff in 7.0
Micha has completely overhauled the way configuration is handled internally, moving to programmatically defined configuration, with schema support, validation and incremental application of config changes (rather than requiring a full config reload for each change)
For those who missed the online meeting, the Lucee team visited the Mid-Michigan ColdFusion User’s Group and presented on these changes (plus a lot of other stuff!) : https://youtu.be/qN-BGkpcAaM?si=Kb1uFtIF7yieZFhe
Our team successfully deployed some of our larger CFML apps on Lucee 7 earlier this month. It’s been wildly successful. We have one app that is running about a 90 node EKS cluster used for for one of our classes at the university I work at to simulate some advanced networking designs, configurations and failures. The code base isn’t 100% the same as when we ran it on 5.4 last year but it’s pretty darn close – and with that we were able to save on CPU and memory usage across the board. We were able to reduce the ec2 instances in the cluster from an m3.medium to an m3.small thanks to the efficiencies in the engine, giving the same or better performance than we had before.
The new Java integrations are also a sleeper feature that won’t get a ton of headlines but has made my life a ton easier. We were able to rip out a bunch of custom java classes we had to write to bootstrap some AWS functions and use the native AWS provided libraries in our project. We could easily call everything from right within our CFML code and fully integrate the library in our projects in hours, not days. Because of the automatic imports in code via Maven, deploying java libraries is very straight forward and more importantly, easily repeatable when deploying projects.