I work at Grafana, and Jsonnet powers our whole k8s infrastructure. It can get a little baroque sometimes but overall it’s tremendously powerful, and it’s fun to work with.
On that note, I'm already looking at migrating my codebase off of Spring. Just testing the waters with Quarkus, Helidon, Micronaut, Pekko, Vert.x, and plain Jakarta EE right now.
Red Hat effectively killed their JBoss/Middleware team and the rest of it moved to IBM https://www.redhat.com/en/blog/evolving-our-middleware-strat... Quarkus and other tools were pushed to CommonHaus/Apache. I believe Vert.X was also mostly developer by RH team, although moved to Eclispe Foundation a decade ago.
Oracle also ended up somehow sponsoring 2 frameworks: Helidon & Micronaut.
I'd bet Spring is still the safest choice next to Jakarta EE standards that all are built on top of nowadays.
Yeah my old colleagues who work on Kroxylicious are now IBM. I keep asking them if they're wearing a blue tie to the office yet, they still don't think it's funny.
I quite like Micronaut, especially the ability to use its compile time DI as a standalone library in a non-Micronaut app.
Quarkus is pretty similar, but is built on top of Vert.x so a lot of the fun of Vert.x (don't block the event loop!) is still present. It also does compile time DI.
And as popular and widely used as Spring is, that would 100% happen. To me at least, I wouldn't count this as a particularly huge risk. But in an enterprise setting, with mandatory auditing and stuff, I can understand why there would be a requirement to at least pre-identify alternative(s).
Probably a bit of overreaction given that Broadcom is now in charge of Spring. At the end of the day it’s a wildly popular open source project — it has a path forward if Broadcom pulls shenanigans.
That said, I have noticed that the free support window for any given version is super short these days. I.e. if you’re not on top of constantly upgrading you’re looking at paid support if you want security patches.
If there's no money in it for them - reduction of staff or funding leading to slower releases and bugfixes
Moving some features like Spring Cloud / Spring Integration, or new development behind a paywall (think RHEL)
Big users (like Netflix, Walmart, JPMorgan, LinkedIn/Microsoft, etc) would likely be able to pay for it (until they moved off), but smaller companies and individual developers not so much
I think it would be more of a Redis situation - steward changes the license, someone large enough to maintain a fork creates one, and everyone moves to the fork. In Redis's case, Amazon forked it into Valkey.
Spring is so widely used that there are multiple "large enough" companies who could do this
What advantage would this offer over a durable, transactional queue, like, say Pulsar? I don't see much of an advantage, and depending on how much you're pushing through your queue, I could see the outbox table causing issues with your production system if you're using that DB to hold the data.
Although the majority of what I do is in Python, I've taken to using Flyway for DB migrations.
A lot of what I do is using some of the more esoteric features of PG (think domains, custom aggregate functions, etc), so the ability to run migrations as direct SQL is great.
True, but built-in replication is still the whole installation or nothing. I believe Slony doesn't have this limitation; though, being trigger-based has some other quirks of it's own. Pgpool is another solution I've also used for a while. All have quirks which I don't recall running into with Mysql's replication.
