I wrote a short story years ago about a tech CEO going to trial over a similar technology that turned out to mass-blind people. Just food for thought I guess.
1,733 karma · joined May 20, 2016
wobsite: https://gabethebando.cc
Maybe I'm being too much of a purist because late model smart telescopes now allow you to grab the raw FITS files so the relative value of a fully open ecosystem is lower, but I'm trying to build towards the motivated teenager segment: someone who maybe has a Seestar or has seen one, and wants to have that experience, but wants to own it themselves and actually understand what's happening. I have the mechanics mostly working in Alt-Az mode, EQ mode is currently accomplished using a hinged plywood board fashioned into the world's shittiest polar alignment wedge, and now I'm working on figuring out how/where to source good enough optics. I have already learned a whack-ton designing the thing, and I hope that in the future the build will be interesting and informative enough for others to get a bunch of experience in Astrophotography for very little money.
You can buy the sensors used in a lot of these telescopes for <50USD on Aliexpress, the drive mechanics and electronics are ~80, and then all you should need (hopefully) is a cheapo Raspberry Pi 3 or equivalent, putting the total BOM at under 200 bucks. I'm hoping that by early next year I'll have something presentable and we can start building an open source ecosystem to compete with these very impressive but very heavily integrated instruments.
Either way, while this is true in the absolute, this is the value of building good MCP servers: they should expose only exactly the surface you expect your agent to need, and adding functionality should be carefully considered. The best MCP servers I use day to day (Cloudflare sticks out) do a really good job of exposing only what an agent might actually want to do on my behalf, rather than just all and sundry. Unfortunately the Chrome Dev tools MCP is less discriminating and is only as secure as a browser sandbox with full JS access (not fatal but not as strong as a well scoped REST API).
All of this is sidestepped somewhat by using good isolation primitives - I'm running a Hermes agent as of recently on a DigitalOcean VM that only accepts connections from my devices over tailscale, and it has all its own credentials so I can revoke them easily should they be used maliciously. Giving an agent root on a box is not _necessarily_ a huge deal, you just have to make sure that box has nothing valuable on it.
I kinda feel like we're rediscovering "Cattle, not Pets" when it comes to the environments we run our agents: give it root, sure, but a root that is almost meaningless outside of the functionality you granted it.
I'm not a Luddite or a drooling evangelist; my take on AI coding is basically that it's a good accelerator if you already know the shape of what you want, and have the software engineering fundamentals to build it yourself. I find myself building a lot more stuff these days because I don't have to wrangle HTML and CSS; it's one of those things I know I can do myself but I absolutely hate, and translating designs from my brain to the browser is a horrible pain. But I know, for example, what flexbox is, so I can tell my coding agent that I want something in a certain shape, laid out with flexbox, and it handles the particulars.
I think we have to acknowledge that most software engineering is just putting mostly compatible fittings together to join mostly standard sized boards. That's okay, and unless you're doing something extremely specialized most of the time the standard solution is more than good enough. I've been loving Supabase recently as my backend for AI built things because it turns out that PostgREST is a great shape for >90% of the APIs I would normally build, and RLS handles auth modeling way more concisely than domain code, and it's very easily scriptable to do all the complicated business logic bits in a contained, testable way. It's not optimal, it's not elegant sometimes, but it's fine. Every other engineering discipline coalesced on standard operating procedures decades ago, maybe we software engineers have to swallow our pride a bit and start accepting them in our discipline too.
Remote: Yes, remote or Cincinnati area only
Willing to Relocate: No
Technologies: LLMs/MCP development (recently for game development and developer support automation), Go, Python, Java, Typescript, Distributed systems, game development in Unity and Godot
Other skills: Developer education, partner engineering/FDE, solutions engineering
Resume: https://gabethebando.cc/resume
Email: g.albacarys@gmail.com
Jack-of-all-trades software engineer with ~10 years experience, focusing on developer experience, working on game platforms such as Netflix's cloud games stack and Niantic's real world platform. I love game development, and am passionate about helping partners succeed by helping connect them with the right resources and blazing a path forward where one didn't exist before. I'm looking for either a developer tools-focused SWE role or a Partner Engineering/Forward Deployed Engineer role. I'm also interested in jumping into engineering management; I've had significant informal management experience and would like to translate that into a full-time move into EM work.
I'm unable to relocate due to family obligations but have no issues with travel. Thanks in advance!
This project started when my partner told me I couldn't buy a Seestar, so I decided "fuck it, I'll build my own!" It's still early days but I think there's something cool here. I'm designing the whole thing to be easy and cheap to print and build. I hope someday this will be the kind of project that a sufficiently motivated teenager will be able to build with minimal experience and an old smartphone; maybe it's healing my inner child, lol.
We aren't a fully screen free family. Our kiddo watches probably 1/2 hour to 45 minutes of TV a day and we aren't so naive as to think plane trips and long car rides will be screen free, so we bring an old iPad loaded up with shows and movies he likes. We review the list beforehand and make sure it has what he wants (subject to our approval). But the night and day difference between a moderated amount of screen time and his peers who are full on iPad kids is just astounding. I just hope we can keep up the low screen time for as long as possible.
This is possibly the most dangerous sentence on this entire page. The hubris and ignorance of history wrapped up in this one sentence is breathtaking.
We've been trying to do "plain English" programming for 50-odd years. COBOL was supposed to be a programming language for business critters. BASIC was going to eat the world. We were going to UML our way out of writing software. Over and over and over again, it's been proven that there is irreducible complexity in programming that can't just be converted to "plain English" in a way that's meaningful to actual laypeople. AI has, in some ways, reduced the barrier to getting software from zero to one, but it's not a replacement for knowing how to code beyond the very basics.
The more I read of this page, the less I understand what's happening here. None of the examples show anything resembling "reducing the difficulty" of programming; it's still abstract symbols mapping to operations on a silicon chip people fundamentally can't reason about without training and practice. There's no universe in which
a <- 5, 6, 7
a = append(a, 5, 6, 7)
I don't know why this makes me so angry. Maybe it's because the pedagogy of software engineering is important and interesting to me, and projects like this are the latest generation of a long line of people believing they have finally figured out the secret formula to make coding "easy". It's just...never going to be easy. That's the nature of complex systems.
- structurally edited, ensuring syntactic validity at all times
- annotated with metadata, so that agents can annotate the code as they go and refer back to accreted knoweledge (something Clojure can do structurally using nodepaths or annotations directly in code)
- put into any environment you might like, e.g. using ClojureScript
I haven't proven to myself this is more useful/results in better code than just writing code "the normal way" with an agent, but it sure seems interesting.
In all seriousness, I am about 10k words into a new draft I started a couple weeks ago. It's a far-future post-catastrophe sci fi type story, with some inspiration taken from Dune and a whole bunch of post-collapse fiction. Writing is tough between a full time job, a toddler, and a social life, but I've managed to squeeze in at least a few sessions a week. Hopefully this will be the one that actually lands!
In other news, my first astrophotography rig is _finally_ mostly fully put together, and I'm going to try to go out and do some captures tomorrow night!
I've been getting into astrophotography recently, so I went out to my local Astronomy club's dark site in Middle-Of-Nowhere, Ohio, star tracker, DSLR, lens and nearly brand new HP Gaming Laptop I bought specifically for this purpose in tow.
It was cold as shit outside - 25 with a wind chill of just under 15 degrees. But I came prepared, and the club has a small heated clubhouse on the grounds of the site, so I set up all my equipment, did my polar alignment, and left my laptop plugged into a power outlet and remoted into it on my iPad so I could monitor the data capture from inside where it was warm.
About 20 minutes later, I lost remote access to my laptop suddenly. No problem, I thought. I headed outside to go debug what was going on, to find that the laptop had shut down randomly. That's weird. I tried to turn the laptop on, and it spun on the windows logo for over 5 minutes. I got worried that somehow out of all this gear I brought out to the middle of nowhere in the freezing cold, somehow the laptop was what had died. I try force-resetting a few times, to the point where I get the windows recovery environment, and it boots _so slowly_ that I think something is seriously wrong. Then the CMOS battery reset screen comes up (what the fuck?) and I finally get it to boot after about 8 attempts. However, it's so slow it's completely unusable - the CPU is pegged at the lowest possible frequency and just opening up the controller software for my star tracker takes nearly 5 minutes. decide to pack it in for the night, assuming my laptop is dying.
I bring all my equipment inside to tear it down, and leave the laptop in the warmth for 15 or so minutes while I tear everything else down. Then I hear the familiar Windows 11 startup chime behind me. I turn around and the laptop happily boots up, running at full speed, as if nothing was wrong.
Friends, the laptop got _too cold_. I have never experienced this before in my life, and I have put laptops through similarly extreme conditions in the past for other projects, let alone all the Raspberry Pi's I've left to bake in the sun and freeze in the cold. I am so done with modern technology, I want to return to 2011 when Thinkpads were good, Macbooks were great, and phones couldn't break my brain's dopamine circuits. I'm so tired.
No, what we need is for people to feel safe in public again, for them to not feel like they're constantly one questionable picture away from their lives being ruined. Kill social media, kill gigantic public face tracking dragnets, kill privacy-invading capitalism.
...As opposed to the actual medieval period, which was famously chaste, calm, and sober?
We are using 4-6 times as much storage as we need to, and these are often not small files (on the order of 100 MB - 5 GB, several dozen times a day) but fixing this overuse is so far down the priority list that I don't think it survived the great Jira purge of mid-2024.
Because I can't fucking stop myself, I created a fantasy ISA and am working on an assembler and basic interpreter for said fantasy ISA.
These people are irreplaceable. And I know one of these forecasters very well, he's an old friend of mine. He is done with the federal government; even if they offer him his job back he's not going back, because his trust that his job as a meteorologist was safe is smashed to pieces. That's irreparable harm.