Trellis Updates

[Aaron Coburn]

Hi All,

This list has been quiet for a while, so I wanted to send out a quick updates of some of the major developments within the TrellisLDP project.

First of all, most of the development conversations happen directly on GitHub, and you're all welcome to add issues or participate there: https://github.com/trellis-ldp/trellis

Second, we are getting close to being able to cut a 0.9 release, which will represent a significant milestone for the project. You can expect this sometime before the end of January 2020, though possibly sooner.

For users of Trellis, you probably won't notice too many changes in the 0.9 series, other than some minor adjustments to certain behaviors (e.g. access to the audit stream now requires acl:Control privileges). Otherwise, the biggest changes relate to:

a) refinement/maturation of the underlying Java API, making it more flexible and extensible. This is really of interest only to folks who might be writing custom back-end implementations, but it also means that custom back-end implementations will be even easier to write.

b) overhaul of the application wiring, moving to Microprofile and away from Dropwizard. You can think of Microprofile as being part of the next generation of JavaEE. Plus it's about 20% faster than using the current Dropwizard. The existing Dropwizard-based application will still be available, but we will be moving away from it as we look into the future -- and for those of you already using the existing application, *there will be no data migration required* when you move to the new system. We have started publishing new Docker images if you would like a preview of these systems (e.g. trellisldp/trellis-database and trellisldp/trellis-triplestore). The most significant change here will be related to how applications are configured: the new system uses environment variables rather than a YAML file. We are still working on documenting how configuration happens, so feel free to ask questions if something isn't clear.

c) integration of a Cassandra-based persistence backend. While the database persistence layer is fast, a Cassandra backend will be even faster and more scalable for systems that need that level of performance.

 

Let me know if you have any questions.

Cheers,

Aaron

[Joshua Allan Westgard]

Congrats on the progress!

[Ralf Claussnitzer]

Awesome! The move to Microprofile seems state of the art. I bet somebody is already thinking about Quarkus. ;-)

Thanks for the effort to everyone involved!
-Ralf

[Aaron Coburn]

Hi Ralf,

The core Trellis code is built and tested with both Openliberty and Quarkus (having two separate implementations is a nice bonus of microprofile), and all the new docker images are built with Quarkus.

So far, I have been very impressed with Quarkus, which will be releasing the 1.0 artifact any day now. For anyone who's unfamiliar with Quarkus, you can check out their website here: https://quarkus.io/

Cheers,

Aaron

--
You received this message because you are subscribed to the Google Groups "Trellis LDP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trellis-ldp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/trellis-ldp/2dfaf855-24e5-4bdc-bcea-f01c0d596b75%40googlegroups.com.