godzillafarts

Tend to the other agents

Recently went through a similar kick in the pants with both Apple (photos, videos, etc.) and Google (Nest thermostats).

I have been piecemeal migrating various services that my home uses into a self-hosted services on my homelab, following this as a rough (hardware) guideline: https://mini-rack.jeffgeerling.com/

There are self-hosted versions of everything you're asking about. You could conceivably create a relatively robust homelab + NAS that will be able to host photos and videos (check out Immich), run a Plex server, and host other services probably using something like ProxMox.

It won't necessarily be easy, or cheap, and you'll be on the hook to keep everything running. But if you're into that kind of thing, it's super rewarding and fun.

So far, we have:

- Home Assistant managing our smart thermostats (Google Nest), smart plugs and light switches (Kasa), and other misc. appliances. Prior to this, I had like seven apps on my phone to manage all our devices and it was a PITA.

- Plex media server. We don't have a ton of digital media so this is very lightly used.

- Grafana + Prometheus monitoring everything.

- Remote access outside of our home network.

Next up on my list is pulling in our extensive photo/video library from various locations like Google Photos, iCloud, two primary iPhones into Immich. The wife and I are both running into storage issues on our phones because we have so many photos and videos of our kids. Don't want to delete them to free up space, and don't want to pay Apple every month to store them. Right now in the mean time, we're exporting photos from our iPhones to the Photos app in macOS, and then archiving them on a 1TB SSD. It works, but there's friction in actually viewing the photos after that.

+1 this isn’t something new, it’s been possible all along in native git if you’re willing to do branch management and rebasing yourself. Just without the fancy UI / stack map.

“You cannot merge a PR in the middle of the stack before the PRs below it are merged.”

Huh? Some stacks need to land all at once and need to be reviewed (and merged) from the top down. It’s not uncommon, in my org at least, to review an entire stack and merge 3 into 2 and then 2 into 1 and then 1 into main. If 2 merges before 3, you just rebase 3 onto 1.

People who love using AI to create software are loving it because they don’t value the act of creating & understanding the software.

That's a pretty sweeping generalization. Just because I don't value the act of typing into a keyboard doesn't mean that I don't value the craft of creating and understanding software. I am not outsourcing my understanding to the LLM, I am outsourcing the typing of the code.

What you are describing is not engineering, it's (pardon the phrase) vibe coding. Claude Code is just a tool, and everyone is going to use that tool differently. There is nothing inherent in the tool itself that requires you to surrender your agency and understanding. If you do, that's on you, not Claude.

When you start a feature branch on your own, do you create an entirely new development environment to do it?

… yes? We have a few wrapper scripts around worktree operations that copy some docker volumes (pg data, bundle cache, etc.) from the base and spins up an entirely new stack on different ports with a host alias. We don’t have to install any deps beyond that because we copied over the ruby gems bundle cache and we’re using Yarn PnP + “zero installs” for client-side deps.

Hillsborough, NC https://www.hillsboroughnc.gov/Home/Components/News/News/856...

My favorites this year, be genre:

Sci-fi: Exiles by Mason Coile.

Fiction (Fantasy): The Bright Sword by Lev Grossman.

Fiction (post-apocalyptic / zombie): One Yellow Eye by Leigh Radford.

Non fiction: The Wager by David Grann.

This is effectively what I'm doing, inspired by HumanLayer's Advanced Context Engineering guidelines: https://github.com/humanlayer/advanced-context-engineering-f...

We've taken those prompts, tweaked them to be more relevant to us and our stack, and have pulled them in as custom commands that can be executed in Claude Code, i.e. `/research_codebase`, `/create_plan`, and `/implement_plan`.

It's working exceptionally well for me, it helps that I'm very meticulous about reviewing the output and correcting it during the research and planning phase. Aside from a few use cases with mixed results, it hasn't really taken off throughout our team unfortunately.

My wife and I have been so fed up with the streaming landscape that we’ve been amassing a library of physical Blu-rays on the cheap from places like McKay’s, etc. It takes a little work, but it’s been really good for our family and encourages us to be thoughtful about what we consume.

Honestly I don’t see us going back to streaming. The content isn’t that compelling; most of what we watch is older (we have kids, so lots of Disney movies), and we’re not really interested in most of the newer shows that would warrant us paying a subscription in order to watch new episodes as they drop. Before we cancelled, I remember looking at the carousel on my Netflix Home Screen and being completely uninterested in any of the content they were pushing.

I’m also not worried about some licensing deal nuking one of my kids’ favorite movies from the catalog now. No ads is just the cherry on top.

SvelteKit. All data is just markdown files preprocessed with mdsvex.

When I'm on the go I don't really do any building, I'll add/edit a markdown file, push it to the remote in a branch. Cloudflare Pages will then give me a preview environment. If I like what I see, I merge the PR and it automatically deploys to "production."

VTI for the majority of post-tax investments, VUG for a smaller percent that I'm more aggressive with.

If you find your interest piqued by this article and want to dive a little bit deeper into specific examples, I highly recommend reading Factfulness by Hans Rosling.

Yes, there are problems with crime reporting. Yes, violent crimes have been on a downward trend since the 1990s. Things can be both bad and getting better at the same time.

My experience, and my team’s experience, with Graphite has been the complete opposite. Graphite slowed us down considerably, which is a shame because it seems like Graphite was purpose built for a development workflow like ours. We continually stack changes to make code review less burdensome. We had been doing it in just git for some time and decided to try out Graphite.

Graphite is quite hostile to collaborating on stacked branches. If one dev restacks and a collaborator is working in a dependent branch, they’re hosed. If a collaborator pushes a change to a dependent branch and they're not using Graphite, hosed. I think one of the main problems is that Graphite abstracts out potentially dangerous git commands into a fancy CLI such that the dev doesn’t really know what’s happening under the hood. It makes things even worse if you’re trying to collaborate on stacked PRs and not all devs use Graphite. Just forget about it.

We could have just been using this tool “wrong” but we had to stop using it within our team because it became so disruptive. At the time it seemed like you should only use Graphite if:

1. You are the only person working on a stack. You do not expect collaborators to be pushing changes to any of the branches within the stack, and

2. Everyone on the team is using it too.

Maybe things have changed in the intervening time, but I’m always dubious of tools like this.

It’s just git. It’s a wrapper CLI around git with a nice GitHub PR integration and a dashboard. It’s already easy to stack branches in git.

Clicked the link. Immediately greeted with a CTA banner, a Google ad, and a cookie banner. Dismissing the cookie banner immediately reveals another Google ad, but this one is sticky and follows me as I scroll. Reading through the article I came across at least three more advertisements.

Kind of difficult to take this seriously.

I've played around with several platforms in the last year or so. I've landed on the following setup that works very well for me and ticks all your boxes:

A SvelteKit[0] app hosted on Cloudflare pages. The repo is hosted on GitHub and hooked up to the Cloudflare Pages app [1]. On PRs, I get preview environments. On merge, the changes get deployed to my "production" website. I write blog posts and other content in markdown, which is then processed by mdsvex[2] with very minimal setup.

Mostly, my requirements were more focused around getting the actual framework, hosting, etc. out of my way so that I could focus on writing. Gatsby and Next.js were too configuration heavy and turned me off once I scratched beyond the surface.

[0] https://kit.svelte.dev/ [1] https://developers.cloudflare.com/pages/configuration/git-in... [2] https://github.com/pngwn/MDsveX

You might already be aware, but Morgan Housel has recently released a new book titled "Same as Ever: A Guide to What Never Changes."

- Ballou, Brendan. Plunder: Private Equity's Plan to Pillage America.

- Smil, Vaclav. How the World Really Works: The Science Behind How We Got Here and Where We're Going.

- Doucleff, Michaeleen. Hunt, Gather, Parent: What Ancient Cultures Can Teach Us About the Lost Art of Raising Happy, Helpful Little Humans.

I like the general idea and also try to conceal my identity as much as reasonably possible when out in the world. However, not having a smartphone just isn’t an option for a lot of parents. I’ve long wanted to ditch my iPhone and get a dumb flip phone, but my infant son’s daycare uses three (!) smartphone apps for critical services. Think checking in/out, door access, critical communications, payments. There are no alternatives; you can’t do anything of these things outside of their respective app.

In my opinion, SO doesn’t really have much of a choice here. Not only has the quality of questions and answers been on the decline, the entire SO experience is notoriously hostile for new comers. I find that ChatGPT/GitHub Copilot tend to give much better answers for the vast majority of programming related questions. It’s much faster, easier, and less intimidating to ask your tools questions knowing that you’ll most likely get a high quality answer and not get admonished for having asked the question in the first place.

It may very well be, but it’s also an incredibly reductive view point. I’ve heard not one person speak to any sort of transition period wrt. training and upskilling of existing workers. So what do we do here? Just eliminate all junior positions? All internships, too? What happens when businesses decide they can’t economically hire any more juniors or do internships because LLMs can do the job better, faster, and cheaper?

Historically we’ve set aside a certain capacity, in almost every profession, to bring up and train the next generation. Apprenticeships, internships, etc. By definition, these people are shitty at their jobs, and that’s kinda the point. Were you brilliant at your first position?

Yeah the reporting here is laughable. The article literally states

A similar vulnerability exists in Google’s Android mobile operating system.

The two iOS widgets that I find the most usable and also provide quite a lot of utility are the MyRadar (weather) and Kasa (smart home) widgets.

I can get most weather information I need from the MyRadar widget; rarely do I need to open the app. If I do, it’s mostly to check weather in other areas like my parent’s home during hurricane season.

Likewise with the Kasa widget. The only time I open the app these days is to setup a new device or an automation/scene.

Useful widgets exist, but I imagine their use cases are fairly limited to mostly trivial things. For more complex applications, it probably simply isn’t worth the dev time to create and maintain a widget.

The coolest website I know right now is https://solar.lowtechmagazine.com/.

It’s a solar powered, self hosted version of Low Tech Magazine. The articles are always interesting, but there’s also a ton of good information about efficient website design as well [0].

[0] https://solar.lowtechmagazine.com/about.html#how

Any idea if this tangentially impacted LFS bandwidth somehow? We're seeing a huge spike in LFS transfer starting yesterday afternoon. The spike doesn't seem to correspond to CI builds cloning the repo (because it was sustained over night after our devs had signed off). No notable changes to the repo, automated processes, etc.

We've had to purchase additional data packs to unblock CI builds and folks pulling from the remote and assume we will need to purchase more. 11 days left in our billing cycle...

Lots of hate in this thread around the author writing about "only" three days offline. I get the sentiment, it isn't a particularly long duration of time, but three days offline can feel like an eternity if you legitimately have an internet addiction. Some of us are luckier than other in that we're able to self moderate.

I know that I sometimes struggle with pulling my phone out for no reason and refreshing HN or Reddit. Same with my laptop during work hours. I don't like it and I'm trying to do better; I want a healthier relationship with my devices and the internet.

I definitely took something from this article, though. For a while I've been using Do Not Disturb to cut down on distracting notifications throughout the day, with a whitelist for family members in case of emergency. But from now on, my phone's WiFi will be turned off and I've also disabled cellular data for all other non-critical applications.

Now if I want to refresh HN or Reddit for no reason, I either have to connect to WiFi or explicitly grant Safari permission to use cellular data. Overall, I feel like I'm taking steps in the right direction...

I've noticed that if I skip breakfast, I intently focus on how hungry I am until lunch time. However, if my breakfast is too "heavy", it makes me feel extremely tired. My go-to these days is whole milk greek yogurt and homemade granola.

I really enjoyed Project Hail Mary, by Andy Weir

My wife and I have been dealing with this conundrum for the last few months. At the end of 2021, we felt like we were drowning in subscriptions. We were subscribed to the following services: Spotify, Netflix, HBO, Disney+, Paramount+, Headspace, Peloton, and Adobe Creative Cloud. We're in the process of purchasing a home, so it was sort of a gut punch to go through our expenses and realize we were easily spending $150+ / month on subscriptions.

We've drastically cut back; the only subscriptions we pay for at the moment are HBO (we're finishing up a show and will cancel that when we're done) and Peloton (keeping this). The challenge for us has really been finding suitable alternatives that are either free, or involve one-time purchase.

Spotify has been replaced by NPR's music discovery channel and purchasing some of our favorite albums on the iTunes Store. Headspace has been replaced by guided meditation videos on Youtube. Our (my) Adobe CC usage has been replaced by purchasing Final Cut Pro X (lifetime license) and using Darktable (FOSS).

Most services have reasonable alternatives that you can either purchase outright or use for free.

Very cool.

One thing I noticed on mobile - if you pan far enough in either direction such that the tiles end up wrapping / repeating, you lose all of the map markers.