Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
pantsforbirds
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
91.
▲
by
pantsforbirds
3y ago
Can someone Steelman the Jones Act for me? I think it was a poor choice historically, but I truly don't understand it's need now.
92.
▲
by
pantsforbirds
3y ago
I lost access to all of our documentation at work for a week because my (small local) ISP banned all content from Somalia as being suspicious. I had to call their office and explain to them that Notion is a rather large and popular applicat
93.
▲
by
pantsforbirds
3y ago
Usually, when you buy things that are drop shipped they are also marked up a ridiculous amount.
94.
▲
by
pantsforbirds
3y ago
It can be difficult to tell what is drop shipped and what is still being handmade. I'm much more hesitant to use it than I was years ago.
95.
▲
by
pantsforbirds
3y ago
Are there any examples of PyApp using libraries that wrap other languages (Numpy, SpaCy, etc.)? In my experience, it has been extremely difficult to ship data/ml heavy applications built with Python as a binary.
96.
▲
by
pantsforbirds
3y ago
If you don't want the AI features it would make sense to be unhappy with the pricing changes.
97.
▲
by
pantsforbirds
3y ago
I've had really good results with it so far. I'm using it in the Huggingface Transformers library, and it's been great for my workflow.
98.
▲
by
pantsforbirds
3y ago
I've honestly had consistently better results with standard regression models. I really love the idea of it, and maybe I need to be tuning it better somehow, but overall I haven't had a great experience.
99.
▲
by
pantsforbirds
3y ago
Has anyone else struggled with Prophet? I've experimented with it on a few real world datasets and I've had very inconsistent results.
100.
▲
by
pantsforbirds
3y ago
I've had pretty good experiences with it. Most of my jobs have come over LinkedIn from recruiters reaching out and my current job (which I really love) came from applying to jobs on LinkedIn.
101.
▲
by
pantsforbirds
3y ago
ChatGPT has massively improved my ability to work with jq. I use it just infrequently enough that I constantly have to read the docs for anything non trivial, but ChatGPT lets me pump out scripts quickly. It's been super nice for my wo
102.
▲
by
pantsforbirds
3y ago
In my experience using tournament selection with a small elitism count works best. You keep the best `n` members of the population and for all other members you group them into random groups of size `m` and select the best value from that g
103.
▲
by
pantsforbirds
3y ago
Good blog post. I do wonder how it would do compare to an implementation of pycuda.
104.
▲
by
pantsforbirds
3y ago
Norvig has always been a huge inspiration for me. I read his sudoku solving blog ( https://norvig.com/sudoku.html ) when I was first teaching myself to code and it really helped me nurture a love for solving fun problems.
105.
▲
by
pantsforbirds
3y ago
Raysan5 (the author), has built a really neat ecosystem from tools he's also made. I really enjoy the way he has gone about it. It honestly reminds me of a minecraft world or something. Check out the full repo and see the projects!
106.
▲
by
pantsforbirds
3y ago
Lee Valley has excellent quality tools for a fair price. They are interesting in that they don't just make updated versions of Stanley tools, but have some pretty unique designs as well. All of my joinery planes are Lee Valley's V
107.
▲
by
pantsforbirds
4y ago
I find myself primarily working in the apache arrow domain these days, so I've done more work with the Apache Flight RPC than anything else ( https://arrow.apache.org/docs/format/Flight.html ). I'll make s
108.
▲
by
pantsforbirds
4y ago
polars is a really great library. Cool to see it expanding into so many languages too.
109.
▲
by
pantsforbirds
4y ago
Nikita is a wild builder. It was insane to watch this play out through his twitter. Now I'm really curious what Discord has planned for Gas.
110.
▲
by
pantsforbirds
4y ago
I've played with using it to interopt some python data science code with some C++ code efficiently without writing the project in Cython or using a tool like Pybind11. It worked pretty well in my test scenario, but I'm not sure ho
111.
▲
by
pantsforbirds
4y ago
Similarly does anyone use https://capnproto.org/ ? It's a project I was really interested in a few years back, but I haven't heard much in the way of it lately.
112.
▲
by
pantsforbirds
4y ago
They are a joy to use for ontologies or certain types of metadata. At my last job we had a bunch of entity categories, and each of those had a huge number of individual entity types. When an entity was picked up through the data pipeline we
113.
▲
by
pantsforbirds
4y ago
Would you bet that we'll see a paradigm shift of a similar nature to the classical physics -> quantum mechanics / relativity paradigm change, but for Cosmology in our lifetimes?
114.
▲
by
pantsforbirds
4y ago
Hard agree Cypher is the easiest way to read complex graph operations.
115.
▲
by
pantsforbirds
4y ago
I mean they asked for a client list and she paid a data scientist $18k to fake one. I guess at some point the question is "do i believe this person is committing fraud on the due diligence requests we already made"
116.
▲
by
pantsforbirds
4y ago
1. Woodworking. I'd like to timber frame a cabin or workshop and fill it out with my own furniture. 2. open source programming. I wish i had more time to give back to some of the ecosystems i enjoy using and there are many problems i&#
117.
▲
by
pantsforbirds
4y ago
A package manager has been the #1 thing keeping me from using Zig more. The build system is already really nice, the language is mostly really nice to use (a few oddities from my persepective, but I'm not a C programmer either), and th
118.
▲
by
pantsforbirds
4y ago
> Funny you said that since a microservice API is always dynamically typed and its usage cannot be checked by compiler. And the more microservices you use the more dynamically typed the whole project gets overall. I mean this is just not
119.
▲
by
pantsforbirds
4y ago
I tend to split things up based on the resources required. We had a decent sized data pipeline that was entirely microservice and serverless and it was a real joy to work with. - Our NLP code lived in a service that ran on GPUs - our inges
120.
▲
by
pantsforbirds
4y ago
It looks like they are still expanding some software roles. Even during layoffs companies usually have specific roles they still need to fill.
More ›