FP and House Music enthusiast. Lead developer on ZIO DynamoDB.
England, United Kingdom
Joined April 2016
- Tweets338
- Following199
- Followers207
- Likes1.2K
Avinder Bahra retweeted
ZIO Blocks now has its own home page:
zioblocks.com/
Next-gen building blocks for modern Scala apps, no effect dependencies, and usable from any ecosystem (or none at all).
Avinder Bahra retweeted
My company @zivergetech is launching a small business automation solution:
getverger.ai/
Most of my followers are in tech and won't need it, but you all know folks who could benefit.
Please share it within your network!
Avinder Bahra retweeted
As many of us, I was also thinking a lot about how code reviews fit into our new coding agent reality. Wrote down my current thoughts: blog.vigoo.dev/posts/code-re…
ZIO DynamoDB 3.x snapshots are out 🚀
vs Scanamo (props to @aplokhotnyuk):
• ~4× read throughput
• ~4× lower allocation
• ~20% higher write throughput
• ~10% lower allocation
Zero-dependency core. Effect-neutral execution.
github.com/zio/zio-dynamodb/…
#Scala #DynamoDB #AWS #ZIO
Avinder Bahra retweeted
Just merged Async[A], ZIO Blocks' new cross-platform, cross-version eager Async type.
Ready for kicking the tires, destined to bring Async streams to JS, and often hundreds to thousands of times faster than Cats Effect. 👀👇
Avinder Bahra retweeted
There was no e-mail, and I learned it by accident, but I feel all #Scala OSS devs should learn ASAP: Maven is introducing publishing limits. We most likely will have to apply for exemption:
- sonatype.com/blog/open-publi…
- central.sonatype.org/publish…
Avinder Bahra retweeted
Your billing query joins 9 tables. Postgres's planner stops trying at 8.
Famous last words: "It's just a 5-row lookup, the planner will figure it out."
0.3ms with 8 tables.
467ms with 9.
exobench.ai/blog/how-one-ext…
join_collapse_limit = 8. Default since 2005. Cross it → planner gives up → hash joins everything → 1.28s for 30 rows.
I would have shipped this. 0.5ms → 5ms looks like a fluke. ExoBench found a logarithmic bomb.
AI → ExoBench → real Postgres, real SQL Server, real MySQL at five scale points. No benchmark code.
- SQL Server: same 9-table query → nested loops at every scale. No cliff.
- MySQL: same query → no cliff.
- Only Postgres has this.
- The fix: SET join_collapse_limit = 12 → sub-millisecond restored.
Every prompt verbatim. Emotive query plans. The cross-engine chart that should annoy you.
Avinder Bahra retweeted
If you caught the article about Proteus but prefer a video format, check out @ghostdogpr's talk from Scalar 2026! 🍿
He shows how Proteus uses zio-blocks to bypass traditional Chimney/Protobuf boilerplate and speed up (slow down?) Scala compile times.
youtube.com/watch?v=zTq9gdIm…
Avinder Bahra retweeted
For all my Scala friends:
Golem 1.5 ships with support for Scala, so I hope you join the $2.5k, three-day hackathon and built something cool on Golem with Scala 3!👇
xitter.cf/jdegoes/status/2052778…
There comes a time in your life when your son looks you in eye and says "Dad, can we go for a drive" (him driving on his provisional licence)
When your Intellij IDE contains more bookmarks than actual code you know you navigating a complex domain
Just read this, looks like an awesome library we could use at work
Just gave my talk at @scalarconf about a new way to deal with Protobuf in Scala.
Couldn't be there? No worries, here's a blog post version of the talk! blog.pierre-ricadat.com/prot…
One of advantages of having a home gym is that you get to decide the music policy. Right now I'm listening to hse.fm/
I finally pushed to button on these TYR L2's, a mere snip at £210. Expect to add a plate to your lifts by just looking at 'em tyr.eu/products/tyr-mens-l-2…
