SOLAR_FIELDS

I do love how the software world has the caricature of a comic book villain in Larry Ellison

If someone proposes the right thing for selfish reasons, do we call that bad? Or do we call it proper incentive alignment?

In the course of my work I deal with organizations who sell cars out of impound lots. There is a whole capitalistic recycling industry of buying junked cars and reassembling them into functioning vehicles again.

Interestingly, a fair portion of junked cars get shipped overseas where the labor is very, very low cost and high tech, cheap labor chop shops in places like India or in some countries in Africa will have highly skilled workers basically cut out crumple zones and restore completely totaled vehicles into new looking cars and turn around and sell them. This would never be viable here in the States because of the labor cost, but when you are paying someone $1 USD a day to do highly skilled and technical labor like this it becomes very viable.

Another popular tactic where I live in Texas is to steal a car and drive it south of the border and strip it. Then buy a junked car same make and model with a clean VIN. Then ship that clean VIN second car south of the border as well and rebuild it with the stolen parts and ship it back over for resale.

My favorite quote from the article:

> So he calls Amanda Seyfried, who sits and watches the scene we’ve just seen of Holmes sitting and watching Seyfried portray her.

This is so very Nathan Fielder. Perfect encapsulation of his unique brand of humor. Seyfried portrays Holmes, giving her interpretation of Holmes. Holmes watches the portrayal, giving her interpretation of Seyfried's interpretation. Seyfried watches Holmes' interpretation of Seyfried's interpretation of Holmes.

This vibe is basically the entire theme of The Rehearsal. It's an exercise in how many Matryoshka dolls we can un-nest in our meta exploration of the topic. Keep zooming out and re-view yourself in the scene you just played which is a super interesting way to discover things about yourself and others

$5 wrench already in play, US citizen was detained at the border recently for simply wiping their phone

The thing that people always miss about this takeaway is that while it is a truism, most data is actually inherently relational. Even in your example given, the individual components that are made to assemble that document are better represented as relational datastores

Nowadays all providers are, or can be somewhat easily, just use Upjet to convert a Terraform provider. They walked away from the custom provider maintenance model awhile back due to its lack of scaling potential and loss of network effects from Terraform being the de-jure framework for infrastructure API's

I will say that yes, quite a few of the providers are half assed and have bugs. Mostly around how reconciliation works, but also gaps in the provider in how it writes fields due to some Terraform idiosyncracy. I often end up having to ignore various fields in the yaml to prevent drift. But now you can wrap your own without too much effort.

HCL itself definitely footguns you into it, it's perverse, because it's intended to be a mere configuration DSL and yet it has since its inception been square pegged into a round hole to do imperative language construct things that it was never designed for. Which is why you get insane practices like "never use an array to iterate over infrastructure in a loop, because removing an element forces a recreation of every resource after that loop because the position of the element in the array shifted to the left". These sort things are completely logical things you would do in a not insane framework (who doesn't use an array to loop over things in a normal software language?), but are a well-known landmine with HCL and Terraform.

You could say it was a mistake to use a language like HCL for Terraform and I would agree. It's a half baked idea that became production way too fast. Reminds me of Javascript.

Statistically, if it was LLM generated, is it actually more bug dense than human generated code was before this stuff all came out?

This weird take of “an LLM made it, it must be slop and buggy” ignores the reality of how almost all software written by humans was also quite shit before LLM’s came along

Double wordplay, not bad. Gives “Descartes before the horse” triple entendre vibes

“Journey” is an interesting way to frame purposefully creating an incomprehensible maze to force end users to walk through the entire warehouse. One of the worse “clever” design philosophies I’ve seen, it’s actively user hostile

I think there are two separate concerns here. The idea that long lived secrets survive in plaintext on a disk somewhere is a bad idea that should go away, but also tangential to that people have to have good hygiene around not printing evil things to logs. Latter is a lot harder to catch imo, and is a separate problem

There’s some rather infamous examples I’ve seen recently of people unpacking grocery store bins at farmers markets. There is no provenance around the origin of the goods in most markets, more common of a problem than people might think.

In Central Texas when peaches are in season people set up roadside stalls, but there’s no guarantee you are getting an actual local grown peach instead of whatever was on the shelf at the grocery store that day

I agree, we have single stream where I live and just understanding the logistics of how single stream works it’s easy to see that single stream leads to a large portion of recycling material simply being discarded because of the reality of sorting

Europeans seemed to manage to figure the problem out. It’s a cultural problem, not a scientific one. Because it requires gen pop to accept some sort of civic responsibility to do it properly, something Europeans have much more inkling to do

I’ve been having Claude do my business books for the past year or so, operating on a beancount ledger and it’s worked quite well

a bold attempt, how does it compare with https://github.com/nektos/act which has been the de-jure project in this space? Act has never been in my opinion super great, but it's also been several years since I've checked in to revisit the project.

Is this similar to Netmaker's offering? Platform wrapper around wireguard protocol?

Lot of people are building agent runtime platforms right now. Look at LiteLLM's play into the space, for example. It's the new hot space for innovation. IMO the main play here is going to be on the self hosted/provenance aspect, because every major provider is now branching into services now to try and entrap you into their hosted platform ecosystem with all of these integrations

I tried this in my house that has ancient copper telephone lines. I was trying to get some level of walled ethernet. I wanted to use the existing wires to pull new ethernet as is discussed here, sadly the builders had stapled them to the interior studs when they ran the telephone lines back in the 70's.

Meanwhile I’m over here with my 100 line flake.nix that installs system dependencies with direnv across hundreds of dev users and dozens of repositories wondering why everyone is all up in arms about boring and stable technology that solves a specific problem

There’s just something so charming about this era of game design. I think SkiFree is a shit game by modern standards, but I absolutely love the design of some stick figure yeti inevitably coming and eating you no matter how good you think you are. It’s a philosophy thing at some point

Can’t defeat that, hard to observe data that you never receive.

Anything that happens on the client, you can capture. You can capture anything delivered to the client that isn’t exposed by downloading the Claude code binary, unpacking it, and patching the nodejs HTTP library fetch call to listen for all traffic, then analyze it and build your own tracing. Or you can MITM with a proxy for a bit more heavy handed approach. It sucks you don’t get it out of the box, but you could literally paste my comment into a Claude prompt and have the tracing you desire in probably less than 30 minutes

> if you have a 500k line codebase and even > 50% is written by agents, you are in a world of pain that won’t justify the costs longer term

A bold claim to make with little to no supporting evidence

I did not try for long, but I was also never able to actually provision, this was a couple years ago when I tried, the only people I saw report success were using the scripted methodology to periodically attempt I mentioned in grandparent comment.

I’ve been PAYG for several years now and have straight up idled the instances for months at a time and never been deprovisioned

For what it's worth, I am a user of this and I'd like to clear up one thing that is not obvious here:

There are actually two tiers of Free here that Oracle offers.

Always Free - No credit card required, sign up. This is what the new limit applies to

PAYG - You give a credit card, but like AWS and GCP there is a "free" tier for people who are on pay as you go plans. From what I can tell, the PAYG subscribers are not subject to this lower limit, only Always Free users.

In other words, if you want the old provisioning limits for free you just have to put your credit card info in.

There are additional reasons that you would want to switch to PAYG - for instance, oracle will literally tear down your always free resources if you are idling them, they are watching CPU and memory and if they see them idled they will take them away. Also, even provisioning always free instances is a challenge - most people have to write bots that attempt on a regular basis, since they appear to only allow you to register when existing resources are freed up, either by the reaper referenced earlier or by someone deprovisioning their always free resources. Neither of these limitations apply on the PAYG tier

It's worth clearing up, because this amount of free resources is worth a nonzero amount of money, if you provisioned comparable resources on any of the tier 1 or tier 2 cloud's you would be paying well over $1k/year for what they are handing over for free. AWS + GCP free compute offerings are anemic in comparison. It's a significant amount of compute.

As an end user, I don't mind if you're building a product with rapidly changing API's out in the open. I'll still use the product if it's useful and deal with the churn. What I DO care about is not having half of that project rug pulled 12 months down the line when your Series B folks decide that this particular money knob needs to be turned up.

So if you want to not piss off your users, that's the way to do it. Build a model where you can actually sustain the product, and no, giving it away on seed and series A round money then rug pulling that gift is not the way to achieve that.

Okay, that’s slightly more reasonable it means if I’m going to be rolling stuff back into OSM in a new country I should at least have a quick chat and ensure what I’m doing abides by that country’s regulations etc like South Korea has some notorious law around data residency

I get this perspective. But I also find the standards to be onerous. They were honestly fine until this one:

> Contact the relevant local communities affected by the contributions

Do I literally have to do new local community outreach every little town one of my bookcases shows up in? Ridiculous if so