HNHacker News
TopNewBestAskShowJobs

stuartaxelowen

701 karma · joined March 26, 2015

stuart at sao dot dev
submissionscomments
stuartaxelowen··on Analyzing Articles on Hacker News Using NLP
> Firstly, any articles that received under 50 points were filtered out

Why? There's still a lot of information in the posts that didn't receive significant interest, and unexplained filtering here seems more suspicious than anything.

stuartaxelowen··on SEC Charges Failed Fyre Festival Founder and Others with $27.4M Offering Fraud
The tempting question: would you have shot Hitler at the height of WW2?
stuartaxelowen··on YC Startup School 2018: a free, 10-week, online course
Create systems by which people come to know about your site and want to visit it to solve a problem they have.

For example, SEO content marketing targets problems people search Google for and create content that helps solve that problem while advertising the product/service in some way. A good example of this is Digital Ocean's "how to set up software X on operating system Y" articles - good SEO, solves people's problems, and advertises directly to their customers/users.

There are other implementations, like more traditional display advertising and product reviews, but they all follow the same formula: create awareness and gives people solid reasons to come to your product/site/app.

stuartaxelowen··on Datasets for Machine Learning
Not at all - I released a customer support on Twitter dataset there specifically focused on unsupervised tasks! I think the focus on supervision in what people do with the data shows that there are still a lot of people poking around with the easier supervised tasks.

[0]: https://www.kaggle.com/thoughtvector/customer-support-on-twi...

stuartaxelowen··on John Resig: Introducing the GraphQL Guide
Totally agree. Sprinkle in some row level security and you've got an app server with data isolated at the DB level, and get the power of SQL for building out endpoints. Compared to this graphql feels a little like engineering theatre.
stuartaxelowen··on Command-line Tools can be 235x Faster than a Hadoop Cluster (2014)
Re: pandas, you could have done it streaming, a'la the `chunksize` parameter.
stuartaxelowen··on Are triggers really that slow in Postgres?
Interesting, I think that a good CI/CD story is probably the biggest thing holding broader use of stored procedures back.
stuartaxelowen··on Are triggers really that slow in Postgres?
Triggers in postgres are amazing - I've got a full automated phrase-extraction system built in postgres triggers, such that salient phrases are automatically detected in marked text fields.
stuartaxelowen··on A tool that lets you hear both Yanny and Laurel
Slider all the way to the right, only hear demonic "Laurie".
stuartaxelowen··on Cilium 1.0: Bringing the BPF Revolution to Kubernetes Networking and Security
Or Band Pass Filter :P
stuartaxelowen··on Why Is It Hard to Make Friends Over 30? (2012)
I see it happen first hand - the people that make friends are the ones that grab them, and don't let them go. If you push through the uncertainty and trepidation that happens early in every relationship, you'll find you've made friends thankful of your persistence.
stuartaxelowen··on The United States of Japan
The social safety net here in Japan is amazing, but the stigma of relying on it is so high that utilization by those most in need is very low, and that taking risks that could potentially lead to needing to utilize it are strongly avoided (among other social reasons for low risk taking).
stuartaxelowen··on Dapps
Page appears to be down from Tokyo.
stuartaxelowen··on California should emulate Tokyo, where housing stayed ahead of population growth
I'm no where near an expert here, but at that expense, why not just give them the money?
stuartaxelowen··on VisiData Lightning Demo at PyCascades 2018 [video]
VisiData is indispensable for data engineering tasks. I've been `head -n 1000 mybigfile.csv | grep somerandomthing | etc | etc`ing for so long, it's ability to give you fast summaries and exploration is such a breath of fresh air!
stuartaxelowen··on Scant Evidence of Power Laws Found in Real-World Networks
Might want to try that again...
stuartaxelowen··on State of Startups 2017 Survey Results
Wait, why are autonomous vehicles not AI anymore?
stuartaxelowen··on The Western Elite from a Chinese Perspective
All patterns are post rationalizations.
stuartaxelowen··on How Bubble bootstrapped a visual programming tool for non-coders to $100k/mo
"Josh and I both had enough savings from previous jobs to be able to live in NYC for more than two years (with a simple lifestyle…), and we didn't start paying ourselves a salary until 2.5 years in."

Critical factor here.

stuartaxelowen··on The GraphQL stack: How everything fits together
Modern SQL databases like PostgreSQL offer full JSON capabilities to serve this need.
stuartaxelowen··on [dead]
Previous discussion: https://news.ycombinator.com/item?id=15619541

Again, simple summaries of complex things removes nuance.

stuartaxelowen··on The GraphQL stack: How everything fits together
Why graphql instead of SQL with proper access control?
stuartaxelowen··on Strategy: How to Develop, Structure and Shape a Winning System
I listened for a couple episodes, but when they started talking about machine learning it became obvious how little they understood it. With how willing they were to confidently opine on how it would change industries based on an incorrect understanding, I lost confidence in the other things they talked about.
stuartaxelowen··on [dead]
"Simple summaries of complex things remove nuance."

GASP! More clickbait "content marketing".

stuartaxelowen··on Why would anyone choose Docker over fat binaries?
It seems you've misunderstood the problem.
stuartaxelowen··on Why we switched from Python to Go
For context, Instagram's primary web servers are Django.
stuartaxelowen··on Clojure/Conj 2017 – Opening Keynote by Rich Hickey [video]
Dynamism is certainly available to statically typed languages, but I believe the problem comes at how difficult it is to write dynamic code in static-typed languages, and to use the common building blocks that help you solve real problems.

Writing dynamic code in static languages is like running through mud, just like writing static checking in dynamic languages is.

I think both paradigms are important and valuable for different applications, but I have never seen a language that does both in any truly useful way.

stuartaxelowen··on How I Reduced My DB Server Load by 80%
The namedtuple is just a datastructure interface with psycopg2. For instance, you could write an arbitrary query "SELECT left.x AS cats, right.created_at FROM left JOIN right ON etc" and you'd get namedtuples with attrs `cats` and `created_at`. Similarly, you can insert namedtuples just like rows, since they are iterables, or as namedparameters with the nt._asdict() method. It's great!
stuartaxelowen··on How I Reduced My DB Server Load by 80%
If you're using python, you can use named tuples for everything. It's really nice! Check out the psycopg2 namedtuple connection.
stuartaxelowen··on Is AI Riding a One-Trick Pony?
Are you maintaining the machine code produced by your code when compiled? You don't care if it's ugly if it works in quantifiable and predictable ways.
← PreviousPage 3 of 7Next →