Lmst

hardwood - A minimal dependency implementation of Apache Parquet in Java

Looks nice, parsing only for now, intended as a replacement for parquet java, but heavily reduced dependencies (only compression libraries).

https://github.com/hardwood-hq/hardwood

atlcli - a CLI for Atlassian products.

Write documentation in markdown, sync bidirectionally with Confluence. Manage Jira issues from your terminal.

https://atlcli.sh/

psc - the ps utility, with an eBPF twist and container context

a fast process scanner that uses eBPF iterators and Google CEL, reduces piping and and scripting when trying to find the right processes. TIL about CEL, a common expression language, with bindings for Java, Go, C++.

https://github.com/loresuso/psc

mailpit - An email and SMTP testing tool with API for developers

Useful for testing via testcontainers, single binary, written in Go, web UI, link checking, html compatibility check, unsubscribe validation - nice set of features of testing emails on the receiver side.

https://github.com/axllent/mailpit

tuicr - A human-in-the-loop code review TUI for AI-generated changes

Fascinating to see, how development moves away from the IDEs and the web (GitHub, GitLab) suddenly.

https://github.com/agavra/tuicr

Your problem framing is sabotaging your strategy

Nice take that problems need to be solved from a user perspective and not what the current industry trends are. Also the way of pushing these things through within an org:

"As soon as any of the people responsible for solving the problem do not have a holistic understanding of that problem, you’re back on track towards a product users hate"

https://productpicnic.beehiiv.com/p/your-problem-framing-is-sabotaging-your-strategy

Elasticsearch Java SDK: No Magic, Just Solid Design Choices

Nice article providing more background about the current Elasticsearch client, which uses a specification to generate a fair share of code (thus making sure that all the languages are up-to-date), while staying typesafe, its HTTP client implementation being pluggable as well as serialization. Great work!

https://javapro.io/2026/01/13/elasticsearch-java-sdk-no-magic-just-solid-design-choices/

memlab - analyzes JavaScript heap and finds memory leaks in browser and node.js

Main use-case seems to be for browser leak checking, but looks simple to use and you can integrate into CI, plus a nice CLI.

https://facebook.github.io/memlab/

SymSpellComplete, a typo-tolerant autocomplete library in Rust

Like a combination of SymSpell and PruningRadixTrie, but better! There's a length blog post linked that contains a lot more information. Very interesting read.

https://github.com/wolfgarbe/symspell_complete_rs

Java: An ecosystem worth billions with IDEs in peril

Very valid point in here. Right now Java almost has a single IDE (plus a little vscode Java maybe). If something bad happens to IDEA, this might be a problem.

https://noprotocol.net/jan/posts/2026_01_02_java_ides.html

witr - Why is this running?

CLI tool trying to explain why a process is running, what service is bound to a port.

https://github.com/pranshuparmar/witr

Databases in 2025: A Year in Review

Andy Pavlo's yearly database review is always a great read. New file formats, acquisitions and deaths (i.e. PostgresML, Fauna and Hydra) were my personal highlights.

https://www.cs.cmu.edu/~pavlo/blog/2026/01/2025-databases-retrospective.html

How we made Python's packaging library 3x faster

Making code faster and more simple at the same time. Nice writeup!

https://iscinumpy.dev/post/packaging-faster/

Column Storage for the AI Era

Great blog post about parquet, it's history, way forward and that it's more than a file format being an Apache project that keeps evolving since more than a decade.

https://sympathetic.ink/2025/12/11/Column-Storage-for-the-AI-era.html

Wanted to test the Junie CLI at the beginning of the year during the holidays as coding assisstant. Even with a Junie subscription I was waitlisted. So I tested opencode with a commercial model instead. Will hopefully have time to write all of this up at some point.

Main question is: As anything AI moves faster than anything else, how do you pick? Or do you follow the k8s way and wait until a winner emerges?

Seems to be streaming service problem: Too many to pick from so you don't pick any

frameworks for understanding databases

Nice introduction into the understanding tradeoffs that the thousands of database had to decide on, in order to deliver well on certain use-cases, but fail on others.

https://www.bitsxpages.com/p/frameworks-for-understanding-databases

dotstate - Manage your dotfiles with ease

Not just a CLI tool to manage dotfiles, but a full TUI. I'm still good with lnk at the moment I guess.

https://dotstate.serkan.dev/

Alexander Reelsen boosted: