will be verrrryyy interesting if the sdf vision gets wrapped into dbt or it's just a case of catch and kill

I'm curious, is anyone here using EdgeDB? I was watching an interview with the founder yesterday, so wanted to understand how popular it is.

I just started using marimo for real today and it's beautiful https://marimo.io

Interesting PEP maybe the death of Jinja? https://peps.python.org/pep-0750/

speaking of PEPs: https://peps.python.org/pep-0771/ this would be very relevant for Kedro

A bit of shameless self-promotion 🙈 I was interviewed for the AI Insights Podcast and talked about my career so far, Kedro K, working on and maintaining an open-source project and the future of AI. 🎧 Give the episode a listen on Spotify: https://open.spotify.com/episode/17WHl9Yj6qgr44R4OskxUx?si=7xqpbvjeQW-yccjllk2OLA

I've taken notes on the tutorial series that was produced last year. The notes are on my blog. Feel free to point out any errors or typos. There undoubtably are some. Notes on Kedro Tutorial Videos

Life-changing dev tool I learned about recently: https://github.com/nektos/act Been using it on the https://github.com/dagster-io/community-integrations repo; works like a charm for almost everything there.

Hi @Juan Luis (and everyone) here you see two Kedro users (a beginner and an experienced one) meeting by chance at an Open Source Saturday meetup in Turin :). w @Giovanni Luddeni

I am conflicted on naming conventions for columns. Do I use spaces or underscores to separate words? • Underscores make it easier to select/edit the entire column name. • Spaces enable narrower columns due to word wrapping. I prefer longer column names (fewer abbreviations) to improve clarity. What are your thoughts?

I was suddenly reminded by the geospatial hackathon idea in #C03RKAQ0MGQ... If you do geospatial work and haven't tried DuckDB for geospatial data, it's apparently really good. I'm no geospatial expert, but last year we got a chance to see how one expert replaced complex Dask-based workflows with DuckDB and Ibis. https://docs.fused.io/blog/how-digitaltwinsim-models-wireless-networks-with-duckdb-ibis-and-fused/ showcases some of this publicly.

💡Cool insight on where Kedro is still introducing performance overhead from a use case I'm working on In almost all cases, either I/O or node execution time will outweigh this overhead by a lot, making it irrelevant but I'm sharing it anyways 😄 In my case, I am running ~800 dynamically created small nodes (the nodes are all interdependent in a very complex relationship graph so I rely on Kedro's orchestration without the need for me to codify the execution order 🙏) Initially my pipeline took 1min40, and the following two changes reduced it to 8sec: • Disabling pluggy tracing (see issue) • Using "list" inputs instead of "dict" inputs in

node

Post

vibe coding:

If anybody will be at PyCon US this week, come say hi! I'm giving a tutorial tomorrow morning and will be around through Sunday.

https://www.reddit.com/r/mlops/comments/19fadaz/comment/lngnp43/ sounds very cool, now I want to learn more about what exactly they've done!

with all the problems of the

.ipynb

format, it looks like it refuses to die. just got this on Databricks

Randomly noticed this, but Kedro seems to be a featured project on https://lfaidata.foundation/.

Hey I'm Richard, I was the original developer working on Kedro-Viz back in the day. When I built the project, I used a package called Dagre to generate the graph layout, but it had an issue with not allowing manual ranking for nodes. There was a PR that added this feature, but the maintainer had seemingly abandoned the project, so I forked the module from the PR. Eventually we moved on to our own graph layout algorithm that the very talented Liam Brummit wrote for us. Anyway I just received an email notification that after all these years, Dagre has a new maintainer who has finally gone back and merged that PR! So even though it's been 4 years since I left the Kedro team, I can finally have closure 😌

related to https://kedro-org.slack.com/archives/C03QPDQTLBG/p1750106258747589, with our number of peak concurrent users of the last 6 months, Campfire fits in a 2GB, 1CPU machine, which is smaller than the smallest machine on sale in Hetzner 😬 hence costing 48 € per year as opposed to... 2 200+ € per year https://once.com/campfire

Insane, I love it, I think https://www.typedef.ai

I know you've always wanted to query your Kedro config files with SQL, now you can https://github.com/AKSarav/YamlQL

Pandera now supports Ibis Could be relevant to the community, I know both pandera and ibis operate nicely with Kedro. I personally haven’t tried

ibis

but use

pandera

at every tabular data project. https://github.com/unionai-oss/pandera/releases/tag/v0.25.0

https://django-configurations.readthedocs.io/ interesting new pattern to configure Django applications given that Kedro settings are modeled after Django, would be interesting to evaluate the pros and cons of this approach

# mysite/settings.py

from configurations import Configuration

class Dev(Configuration):
    DEBUG = True

@Max Pardoe has left the channel

If anybody's interested in learning about data validation at scale, I'll be talking about the new Ibis integration for pandera with with Niels Bantilan, Chief ML Engineer at Union.ai and the creator of pandera next week! Check it out: https://www.linkedin.com/events/7357004656075579392/comments/ Hoping to also give the

kedro-pandera

integration some love later this year! 🤞

first stable release of

behave

in 7+ years! https://pypi.org/project/behave/1.3.0/