HNHacker News
TopNewBestAskShowJobs

kalendos

264 karma · joined January 21, 2016

Data U Engineering U Open

davidgasquez[at]gmail.com

submissionscomments
kalendos··on What I learned building an opinionated and minimal coding agent
You can use your ChatGPT subscription with Pi!
kalendos··on Show HN: OpenTimes – Free travel times between U.S. Census geographies
Totally! I've been using this pattern a lot and recently wrote about it.

https://davidgasquez.com/community-level-open-data-infrastru...

kalendos··on Show HN: OpenTimes – Free travel times between U.S. Census geographies
Nice! I know a couple of projects that have been using this pattern.

- https://bsky.app/profile/jakthom.bsky.social/post/3lbarcvzrc...

- https://bsky.app/profile/jakthom.bsky.social/post/3lb4y65z24...

- https://skyfirehose.com

Love this distribution pattern. Users can go to the Parquet files or attach to your "curated views" on a small DuckDB database file.

kalendos··on Ask HN: What's Your Cursor Setup?
Perhaps the most interesting thing I've done is to teach Cursor Agent some tricks like using `uvx` / `npx`, or searching the web with `ddgs`.

Wrote about it recently (https://davidgasquez.com/cursor-agent-tricks/) although is now outdated! You can find the latest setup on my Dotfiles repository (https://github.com/davidgasquez/dotfiles/tree/main/cursor).

kalendos··on Apache DataFusion
You might need to adjust filters to do an apple to apple comparison.

https://benchmark.clickhouse.com/#eyJzeXN0ZW0iOnsiQWxsb3lEQi...

kalendos··on MTA Open Data Challenge
I can only imagine. Many ETLs are already messy in companies with better tooling and processes.

Would love to read more about your experience with Open Data. Any place where I can reach out?

kalendos··on The new local data stack: Integrating Dagster, dbt and DuckDB
Sure! You can find my contact details on GitHub: https://github.com/davidgasquez.
kalendos··on The new local data stack: Integrating Dagster, dbt and DuckDB
I've been working with this stack (building Open Data Portals¹) for a few months and am super happy with how well everything plays together.

¹ https://github.com/davidgasquez/gitcoin-grants-data-portal

kalendos··on Ask HN: What companies have publicly available handbooks?
> Is is just a collection of random stuff?

Indeed. Is mostly a way for me to keep track of interesting things and have a quick way to re-learn something.

> I found this interesting: https://publish.obsidian.md/davidgasquez/Rationality . Somebody tried to explain what 'rationality' is, without tossing around the word loosely as most people do.

That note should be named "Thinking" to be honest. The goal is to collect a bunch of bullet points that I can read at any given time to remind me how to think better or common gotchas.

kalendos··on Ask HN: What companies have publicly available handbooks?
I keep a list of public company handbooks here: https://publish.obsidian.md/davidgasquez/Company+Handbooks.
kalendos··on Ask HN: What are some examples of websites where the author learns in the open?
They are called multiple ways. I like to call mine a Personal Handbook[1] but I've also seen this being referred as Digital Garden or Personal Knowledge Base.

I keep a list of open knowledge bases here: https://publish.obsidian.md/davidgasquez/Personal+Handbooks

[1] https://publish.obsidian.md/davidgasquez/

kalendos··on Ask HN: Show us your digital garden
This is mine: https://publish.obsidian.md/davidgasquez

Recently moved to Obsidian.md and couldn't be happier!

kalendos··on Most-mentioned links in Hacker News comments, 2006–2015
Nice catch! I used the first regex I could find on SO.
kalendos··on Most-mentioned links in Hacker News comments, 2006–2015
Very cool! Inspired me to get the most upvoted XKCD comics on Reddit for 2019.

https://gist.github.com/davidgasquez/3aeaac54c5a61216ffc8f7d...

kalendos··on Ask HN: How do you take notes when reading a book?
I have a similar workflow for adding content to my wiki (https://github.com/davidgasquez/handbook) from the computer.

On the go I'll add a task to taskwarrior and get reminded when I sync it from the PC.

>What would be even more amazing is when you can take ability to note under any line in a book (as Holloway already lets you), and see everyone's notes for any line/chapter of the book/paper you are reading.

I haven't used it but sounds similar to what https://web.hypothes.is/ is trying to do.

kalendos··on Our list of free software for remote work
Depending on the group taste, these suggestions might help!

- Drawful 2. Free on Steam. https://store.steampowered.com/app/442070/Drawful_2/

- Pictionary clones. I like https://sketchful.io/

- Local coop games via Steam Remote Play. https://store.steampowered.com/remoteplay

- Virtual scape rooms.

kalendos··on Rules of Machine Learning: Best Practices for ML Engineering [pdf]
These rules are also on Google's Machine Learning Guide[1]. I find the website easy to parse. Definitely an awesome resource to review before jumping into any ML project!

[1] https://developers.google.com/machine-learning/guides/rules-...

kalendos··on Timeline of the far future
Made me remember a Carl Sagan quote.

"We are like butterflies who flutter for a day and think it is forever"

kalendos··on Ask HN: Favorite note-taking software?
I have a similar setup. Markdown files published in the open¹ using Hugo. That said, I'd love to have an easy way to add tagging and searching without having to create my Hugo template.

¹ https://davidgasquez.github.io/handbook/

kalendos··on Ask HN: Good Python Machine Learning Codebases to read?
You can take a look at Open Source Kaggle solutions.

- https://github.com/davidgasquez/kaggle-airbnb (I'm the author)

- https://github.com/benanne/kaggle-galaxies

- https://github.com/rafacarrascosa/samr

kalendos··on How to make MongoDB not suck for analytics
Would you mind sharing the process? I'm curious which flow you use to materialize the oplog in Athena/S3.
kalendos··on Dtube – A decentralized video platform using STEEM and IPFS
> Having a platform that incentivizes popular content should in theory attract popular content to the platform. This should attract viewers to the platform, thus increasing the viewership available to see your less popular content.

Thanks for pointing that! I didn't though of it this way. To be honest, my fear is that this platform could turn possibly in some sort of niche. Right now, popular content is about X, so if I want to make quick profits, I'll create a video about X. More people will join and do the same.

I'm sure there are ways to fix that like giving people incentives for posting in unpopular topics. Potentially, each topic could be its own market!

> You're not being penalised, so why complain?

Not complaining at all for now. Just sharing my current thoughts! If I join a social media, I'd love to have diversity of content and opinion. I think in STEEM that could be hard, as it is in Reddit unless you subscribe to unpopular ones. Not a 100% bad thing though!

kalendos··on Dtube – A decentralized video platform using STEEM and IPFS
I like the idea and would love to try it! I have one small concern though. Since the incentive (money) is given to popular videos, why people should post unpopular opinions in Stemmit or videos in Dtube?

That is, if I upload a video of a cat doing some funny thing I'll get more tokens (potentially) that with a super technical video.

kalendos··on Atom 1.18 released
Also, Atom 1.19 Beta has native text buffer. This should reduce memory usage for large files. Will try it out!
kalendos··on Atom text editor 1.7.0 released
You can always try to make a Pull Request at see what it comes out!
kalendos··on I Open-Sourced All My Business Ideas
It will be interesting to sum up those ideas into an awesome list[1][2] at Github.

[1] https://github.com/sindresorhus/awesome

[2] https://github.com/tastejs/awesome-app-ideas