HNHacker News
TopNewBestAskShowJobs

jmatthews

331 karma · joined September 16, 2010

software/data engineer, python, golang, c#, java, c++, in that order. Domain expert with data: hygiene, acquisition, warehousing. Generalist.
submissionscomments
jmatthews··on Class Action Against General Motors LLC, OnStar LLC, LexisNexis Risk Solutions [pdf]
You can't take this as authoritative but my business has a data relationship with Toyota and they have a ton of juicy telemetry data.

Their attorneys are mad protective of the PII they have. Our relationship serves the public interest. We use the data to find people with open recalls where Toyota doesn't know who the current owner is.

I say this to say that we have other OEM relationships that are far more liberal with their encumbered data. This far Toyota seems to be playing it very straight.

jmatthews··on The Pursuit of Efficiency and the Science of Steam
The creation of the steam turbine is also fascinating.

It's a testament to engineering that hot air engines are remotely competitive when so many of the physics line up for steam.(phase change, heat capacity, heat conductivity)

The limiting factor as always is heat transfer between the heat source and the medium. If we had better heat exchangers, the theoretical maximum of steam is still the best.

jmatthews··on Debunking the Myth of "Anonymous" Data
I like to make an appearance in threads in this domain and just say, yeah, it's not just possible but "fun" to put the puzzle pieces together.

It is a very tough problem to solve. Especially when you consider the richness of the datasets you use to put the pieces together.

In my experience the only effective means is to poison the data, in addition to the common sense steps mentioned here.

Poisoning a dataset means seeding a wide variety of the datasets you use to discover PII with fictional look alikes that resist debunking.

Additionally you can poison the core set if you are very clever about it.

jmatthews··on Detroit wants to be the first big American city to tax land value
Munis stink compared to even moderately well managed REITs, especially in inflationary times.
jmatthews··on Detroit wants to be the first big American city to tax land value
At risk of Godwin's law, Hitler was elected right?

Or Ben Franklin, "Democracy is two wolves and a lamb voting on what to have for lunch"

Mob rule doesn't imply ethical or moral.

jmatthews··on Detroit wants to be the first big American city to tax land value
Real estate is the most durable investment. A plurality of wealthy people became so via real estate.

It is also reasonably resilient to inflationary dilution.

I'm fortunate to be still earning. You could spend your life building a small nest egg and paying off a home to retire in and have it just litigated right out from under you.

Your argument is "time marches on" but then what value in saving and accruing capital? It's all short sighted transactional consumerism or bust. That's no way to live.

jmatthews··on Detroit wants to be the first big American city to tax land value
Virtually every effective hedge against inflation is taxed heavily. There is no respite from dilution.

When you look at it from that perspective you see a systematic attack against any safe harbor.

jmatthews··on Stringzilla: Fastest string sort, search, split, and shuffle using SIMD
I realize "speed is a feature", but it's rare when that is THE feature. :)
jmatthews··on Ask HN: What's the coolest physical thing you've made?
Fan of the channel.
jmatthews··on Ask HN: How can I get better at writing production-level Python?
Avoid writing java... (fixed it for you)
jmatthews··on Ask HN: How can I get better at writing production-level Python?
Second or third reco for Fluent Python. Also, to be a purist for a moment. Explicit is better than implicit. Use this as a razor to decide all ties in your design.
jmatthews··on [dead]
Without speaking to the content of the post, there should be a special hell for this type of hype followed by a paywall. This tactic is responsible for a great deal of the shittyness of the modern web IMO. I hope it is judged as the low value click bait it is.
jmatthews··on StableLM: A new open-source language model
The AI war has officially begun.
jmatthews··on Downtown San Francisco Whole Foods Closing a Year After Opening
out of curiosity, why is this comment being voted down? It seems reasonable, is there more to it?
jmatthews··on Mastodon hit 10M users
I know I will 100% regret posting anything remotely viewed as political here but this...

Both sides actually want free speech, the blue side just argues that speech is more free when it's not stifled by hate.

Is the most alarming doublethink I have seen in some time.

jmatthews··on Ask HN: What is the thing you've built that you regret the most?
I wrote a largely autonomous agent that seeks and builds data dossiers on everyone you point it at. Then you buy legally encumbered data, genuinely sensitive stuff for fractions of pennies, and you reverse identify the person, unencumbering the majority of the data, because writing privacy laws that don't cripple well intended business is very difficult.

It is not inherently bad. Some of the real life uses are practical and beneficial, but it can also be weaponized.

jmatthews··on Python Asyncio
Just to swim upstream. For http requests and bounded IO I've found asyncio to be straightforward and a game changer. In the context of call an endpoint with a data payload and have the endpoint process it with outbound http calls it is a 10x'er for very little complexity and no external libraries.
jmatthews··on Katana: A crawling and spidering framework
Was there any specific library this was inspired by, or a specific use case it was built for besides the obvious generic case?
jmatthews··on 'Soleus pushup' fuels metabolism for hours while sitting
If you do back and legs you should be done for the day. Split your posterior chain. :)
jmatthews··on AI is getting better at generating porn
That's either a healthy attitude that I lack or a form of defeatism that I lack, I am not sure which. I love the turn of phrase though.
jmatthews··on Ask HN: Are authors of “how to be great developer” blogs great developers?
https://xkcd.com/1053/

The truth is that "great" is subjective, and any given developer blogger could be a proper resource for any number of people depending on where they are in their journey.

If it were otherwise you could say to any aspiring developer, "Read The Art of Programming by Knuth and just derive everything else from first principles"

A truly mediocre programmer could teach a great deal to a beginner. A master could skirt over various complexities and disappear up their own butt in abstractions.

jmatthews··on Pen and paper exercises in machine learning (2021)
this is shockingly good. to add more, I believe I am the use it or lose it dude.
jmatthews··on We will never have enough software developers (2020)
I understand the concept here but there is also a level of right tool for the job.

Some guys see a screw and reach for their trusty hammer. Some guys know to grab a screwdriver.

I had a project the last two weeks where the code was just going to fail about as often as it was going to succeed. I had to write a resource manager and an Erlang style supervisor and use an embedded key value store.

A better dev may have intuited what took me basically a midstream rewrite to figure out, a worse developer may still be grinding on the problem.

I think my solve is "robust enough" but there was no real way to power through that. You either found the right abstractions or you didn't.

jmatthews··on Nuclear war probabilities are useless - counter to the nuclear gambler’s ruin
This runs counter to the black swan concept popularized by Taleb. In that they are by nature unpredictable, what one can predict is that there will be a next one.

I like this framework but I'd appreciate hearing more about it.

jmatthews··on Researchers Build AI That Builds AI
I hate to be the naysayer in the thread, the paper is well written but the headline is super misleading. AI that tunes AI. It's a search algorithm.
jmatthews··on Ask HN: Who is hiring? (January 2022)
it's curious, your preamble reads AI shop but your infrastucture is Node-centric?
jmatthews··on Self-Hosting Dozens of Web Applications and Services on a Single Server
Thanks for the write up. I use roughly the same set up minus docker. My deploy is usually log in to ssh, git pull and restart the systemctl service.
jmatthews··on Intellectual Loneliness
To me it revolves around the introvert vs extrovert stuff. With myself as the model, it's not that I don't enjoy speaking to people, it's that I don't enjoy small talk. Give me a heady subject that is interesting and I will jump in with both feet. Rehashing the weather drains me.
jmatthews··on Big data may not know your name, but it knows everything else
Without digging in to details, I just want to agree that reversing into a clients PII, while not trivial, is certainly possible, and depending on the class of data, one can achieve fidelity levels in the high 80 percentiles.
jmatthews··on Why Not Rust? (2020)
Is that narrative still true post-modern cpp? On some level it feels like the cpp narrative was written in the early 2000s and hasn't kept pace with reality. Too cluttered, too object-centric, bad compile times, sure. Is the cognitive load of choosing safe pointers that much more than navigating the borrow checker?
← PreviousPage 2 of 6Next →