Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
paddy_m
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
17 ms
·
211.
▲
by
paddy_m
3y ago
The profit calculator from NYMag was a great series of articles about the economics of business from investment banks, to drug dealers, to yoga studios. https://nymag.com/news/features/2007/profit/
212.
▲
by
paddy_m
3y ago
Thanks for putting this together, it's very creative. Do you mind if I post videos of solving this to my youtube channel (that has absolutely no viewers)
213.
▲
by
paddy_m
3y ago
Check out generating ibis, which can output SQL and many dataframe formats (pandas, polars, modin,...)
214.
▲
Build better UIs faster on top of Pydantic
(github.com)
2 points
by
paddy_m
3y ago
|
0 comments
215.
▲
by
paddy_m
3y ago
Location: Boston Remote: Yes Willing to relocate: Yes Technologies: talking to users, python, pandas/numpy, jupyter, js/ts Résumé/CV: https://www.linkedin.com/in/paddymullen/ Email: paddy@paddymulle
216.
▲
by
paddy_m
3y ago
The Jones act.
217.
▲
by
paddy_m
3y ago
I would love to read a comparison explaining the relative advantages of each framework from an experienced practitioner who has actually built apps with each. Add in plotly dash, bokeh/panel, and voila too. Off the top of my head, bok
218.
▲
by
paddy_m
3y ago
Shiny has a DAG that enables intelligent caching. Here is a good talk explaining it https://www.youtube.com/watch?v=YNCPc9aWm_8
219.
▲
by
paddy_m
3y ago
By Ton-Miles the US leads the world in freight hauled. But for a variety of reasons, railroads are more interested in extracting profits from existing large customers than they are in increasing mode share. A couple of reasons for this, ra
220.
▲
by
paddy_m
3y ago
It took a lot of reading of that blogpost before I realized that it was about an operating system. I seriously thought it was about some parallel threading standard. Especially for a broad announcement like that, they should give readers
221.
▲
by
paddy_m
3y ago
I read that post, and I have been working slowly on a set of jupyter notebooks, with ipyleaflet (mapping) integration to make building train maps (with max speeds) easier. Looking for colaborators https://github.com/paddymu
222.
▲
by
paddy_m
3y ago
Acela isn't HSR, and Metroliner which came before had travel times from NYC to DC as low as 2.5 hours. Acela does the same trip in 3.5 hours. Metroliner lowered trip times primarily through faster acceleration.
223.
▲
by
paddy_m
3y ago
I built Buckaroo because I wanted a better experience for pandas/polars data analysis in Jupyter. Buckaroo is a modern data table for Jupyter that expedites the most common exploratory data analysis tasks. The most basic data analysis
224.
▲
Show HN: The Buckaroo Data Table for Jupyter
(github.com)
2 points
by
paddy_m
3y ago
|
1 comments
225.
▲
by
paddy_m
3y ago
The app is interesting. Who here rides? Dirtbikes in particular? I have a trials bike (Gas Gas TXT 250), and an enduro (husqvarna TE 150). I couldn't find a trials bike in the app, but it doesn't matter since trials bikes don&#
226.
▲
by
paddy_m
3y ago
What programming languages and libraries are well explained in this aspect?
227.
▲
by
paddy_m
3y ago
I am surprised that there's less mention of Brook's proposed team organization, a team of 10 led by a surgeon. The team is supposed to work on a single project. Every org I have seen has much smaller teams with much more individ
228.
▲
by
paddy_m
3y ago
You seem to know a lot about battery tech. What do you read to keep up to date?
229.
▲
by
paddy_m
3y ago
Not only that, our built environment is uninviting to walk in because it's built for cars. When have you ever walked by a parking lot and said "Oh my god, that parking lot was amazing, I want to spend time around it"? Yet w
230.
▲
by
paddy_m
3y ago
Oulu Finland would challenge your perception about good weather riding. https://www.youtube.com/watch?v=Uhx-26GfCBU
231.
▲
by
paddy_m
3y ago
That is rude behavior from those two wheeled users. If you honestly look at the dangers and attitudes involved I think you will find that drivers of cars are a much bigger problem. In a large portion of cases where a bike is using pedestria
232.
▲
by
paddy_m
3y ago
I'm glad to see ambitious development being done on the core emacs platform. Could you include some images of what is possible with webrender in the README and the introduction to your docs page. I think it will help others get excit
233.
▲
by
paddy_m
3y ago
I'm a software engineer, and the most concerned about marketing in my circle... and I still suck at it and promoting my stuff. How do I get better? I would love to work on actual marketing projects with experienced marketers as a sid
234.
▲
by
paddy_m
3y ago
I feel the exact same way. I miss F1 races showing at 730AM EST, it was very easy to get up and have a full day. I barely bother watching the 1PM races.
235.
▲
by
paddy_m
3y ago
I really liked seeing this page on tanstack's website explaining their collaboration with ag-grid. Better wayfinding to the proper project is a huge blindspot for most open source projects. I built a page specifically listing related
236.
▲
by
paddy_m
3y ago
I haven't looked extensively at react-datasheet. It looks like it is trying to build more of a full product than the other data tables. I have used ag-grid extensively, its an impressive product. Some pieces are a little awkward to u
237.
▲
by
paddy_m
3y ago
I have been hacking on a better calculation environment for train speed zones. Alon Levy wrote some python code that calculates how long it takes a train to accelerate and decelerate between different speed zones. I have been cleaning up t
238.
▲
by
paddy_m
3y ago
Exactly. A more charitable way of putting it is, you can make a pretty presentation when you know the data that you are trying to present and the story you are trying to tell. I can't imagine a 3 column pricing table that wasn't
239.
▲
by
paddy_m
3y ago
It's a term I made up for myself. I should write an article about it. Look at this screenshot of a Bloomberg terminal https://www.investopedia.com/thmb/trBFMDD8xeZkhLiLscngql3FV_...
240.
▲
by
paddy_m
3y ago
I'm comfortable installing jupyter and databases on my machine. What's the advantage of this approach? Is it the one step docker deploy? Could you talk about the pain of the alternative solution a bit more?
More ›