Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
mcdonje
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
33 ms
·
481.
▲
by
mcdonje
4y ago
This is a conservative think-tank article about how the people who are actively passing laws against teaching black history and sex education shouldn't have their hate speech infringed. Conservatives have been attacking institutions of
482.
▲
by
mcdonje
4y ago
>The treat’s name is a portmanteau of the surname of the bar’s inventor, Theodor Tobler, and torrone, a toasted-almond nougat confection typical of Christmas traditions in south-western Europe. I take exception at the word "inventor
483.
▲
by
mcdonje
4y ago
If it's tabular, self-describing formats have way too much overhead. I ran a query with a tabular result in the neighborhood of 100 columns by 215k rows, and exported it in multiple formats: - CSV: 166mb - JSON: 795mb That sai
484.
▲
by
mcdonje
4y ago
Apache Spark / Databricks is an example of this. Parquet files are stored in folders. A folder is assumed to hold one dataset split into multiple files based on specified partition criteria. The VMs read the necessary files into memory
485.
▲
by
mcdonje
4y ago
"Others have it worse" isn't a compelling argument for maintaining a status quo that isn't working for most people.
486.
▲
by
mcdonje
4y ago
I already said there is no cabal...
487.
▲
by
mcdonje
4y ago
I'm talking about institutions and incentives, and you're accusing me of inventing a boogyman...
488.
▲
by
mcdonje
4y ago
It doesn't suggest conspiracy. It suggests organizational structure. Reporters don't get to cover anything they want. They get paid to cover certain topics and are supposed to stay in their lane. Editors are gatekeepers of what ge
489.
▲
by
mcdonje
4y ago
You are assuming the person you replied to thinks there's a cabal or conspiracy. Alternatively, the 1% could be using mass media to project their viewpoint simply because most of the mass media is owned by the 1%.
490.
▲
by
mcdonje
4y ago
>Windows was a poor substitute for the feel of the Star interface. I'm intrigued. Any specifics about design decisions come to mind?
491.
▲
by
mcdonje
4y ago
That's a bad assumption.
492.
▲
by
mcdonje
4y ago
>would expanding that space really be such a burden? It would require dropping UTF-16, which is the limiting factor. That, or defining planes that aren't available to all encodings. Don't hold your breath. I don't know who
493.
▲
by
mcdonje
4y ago
Most of those are suburbs. 85% of the US population lives in metropolitan statistical areas. https://en.m.wikipedia.org/wiki/Metropolitan_statistical_are...
494.
▲
by
mcdonje
4y ago
>I hate being around so many people packed in a train, airplane, bus I hate being in traffic. On a bus, people are around but you usually don't have to think about them. You can read a book. In traffic, people aren't next to yo
495.
▲
by
mcdonje
4y ago
>unless you live in a large city Most people live in large cities
496.
▲
by
mcdonje
4y ago
Not exactly the same because the phone camera can take several sensor scans, bracketed or otherwise, to create one image that the user sees. Not surprised ML is involved, but it's more akin to a photographer taking several images and c
497.
▲
by
mcdonje
4y ago
So now when they say "Enhance!" in crime shows, it's more plausible. Depending on the application, this is a good model. It makes the maps look sharper. There are photo editing plugins that use similar technology. I would hop
498.
▲
by
mcdonje
4y ago
Excel has an option for formula calculation, automatic vs manual. The default is automatic. There is also a 'calculate now' button to use for if you have manual calculations set. I would expect that if automatic calculations set,
499.
▲
by
mcdonje
4y ago
I haven't seen this before, but it rings true. Nice conceptual framework.
500.
▲
by
mcdonje
4y ago
Very cool. I've done plenty of work supporting non-technical users who primarily interact with spreadsheets. This would definitely make my life easier. Is it going to be cloud-only, or are you planning on making a desktop app? If it&#x
501.
▲
by
mcdonje
4y ago
My first reaction was that you need to send all potential search results across the network in the DB, which would be ok for a small app. So, I guess there's a goldilocks app size where this is a good solution.
502.
▲
by
mcdonje
4y ago
I agree about Rust's memory safety advantage over C++, but I disagree that it's disappointing from a project perspective. Some DB experts made a good DB using a performant language they're comfortable with. You can't mak
503.
▲
by
mcdonje
4y ago
Yeah, I see it as the sqlite of the reporting world. Especially with columnar stores and parquet support.
504.
▲
by
mcdonje
4y ago
According to the comment you replied to, you're statistically more annoying.
505.
▲
by
mcdonje
4y ago
>Fuller is an inspiration, not a terminus. We can build on and improve his work. He would agree. "Call me Trim Tab" - Fuller
506.
▲
by
mcdonje
4y ago
Slow is fast.
507.
▲
by
mcdonje
4y ago
sccache seems awesome. I wasn't aware of it. Thanks.
508.
▲
by
mcdonje
4y ago
I don't know much about how the compiler works, so the answer here is probably that I should read a book, but can external crates from crates.io be precompiled? Or maybe compile my reference to a part of an external crate once and then
509.
▲
by
mcdonje
4y ago
In a datalake, PySpark should be the default for large operations with multiple transformations. Since you're essentially working with jupyter notebooks, it allows you to lay out your transformation steps in a way that's more clea
510.
▲
by
mcdonje
4y ago
You could be onto something there, but it'd be complicated by displacement. If it pushes the water boundary down 20', how thick is the ice above sea level? If they have a pretty consistent way to estimate that at the relevant leve
More ›