eulers_secret

We need smart pantries to fully realize the “auto shopping list”, though… so a smart fridge is a half solution at best.

I even store produce (ex onions) on the counter tops, so I guess smart counters as well

I use opencode or pi harness with the deepseek api for all my at home coding usages.

Deepseek is at least on par with Sonnet (ghcopilot at work)- I don’t use opus, too spendy and I don’t need that level of ability.

The cost is for me was $5/6 months of use. Not a big user I guess! It’s good though, fast enough and incredibly inexpensive.

Been testing Qwen 3.6 28B on a 5090, and it’s also quite good for “free”.

I mostly do small self serving embedded projects based on esp32, so not very complex.

You can have these things without owning a house.

Community is IMO easier in apartments, the higher density means there’s a bigger pool of folks you might click with.

Community gathering spots are more prevalent as well: pools, playgrounds, green spaces, etc are common amenities. Especially if you get a dog, you will meet lots of people.

Of course it depends on which complex you live in. If I move I will miss my apartment and its community. I won’t miss my window aircon though, central air would be awesome.

Depends on the machine you’re targeting.

I do embedded Linux and ram usage is a major concern, same for other embedded applications.

I’m partying like it’s the 90s, on a 32-bit processor and a couple hundred MB of ram.

Tig is a nice and long-maintained git tui you might enjoy, then!

If nothing else maybe for inspiration

It’s interesting how internet backlashes can be large enough to move the needle: ring breaking with flock is evidence of this.

Yet simultaneously the internet represents the opinions of a very small and vocal minority.

I’ve never seen an internet boycott have an impact.

Since we're talking about GitHub Copilot I'll lodge my biggest complaint about it here! The context window is stuck at 128k for all models (except maybe Codex): https://github.com/microsoft/vscode/issues/264153 and https://github.com/anomalyco/opencode/issues/5993

This absolutely sucks, especially since tool calling uses tokens really really fast sometimes. Feels like a not-so-gentle nudge to using their 'official' tooling (read: vscode); even though there was a recent announcement about how GHCP works with opencode: https://github.blog/changelog/2026-01-16-github-copilot-now-...

No mention of it being severely gimped by the context limit in that press release, of course (tbf, why would they lol).

However, if you go back to aider, 128K tokens is a lot, same with web chat... not a total killer, but I wouldn't spend my money on that particular service with there being better options!

I know this is a dead thread, but...

How do so many people justify buying the new redesign? I mean it came out after the CEO went in front of the world and gave two nazi-like salutes, then did DOGE!

Do they buy his 'autistic' defense? Do they just not care about what the CEO does and support him with their money anyways? Do they actively support his ideology?

I suspect it's likely a mix of these depending on the person, and probably more that I can't think of.

I mean they're good cars, no doubt, and it's a damn shame many decent engineers and workers put in so much effort to have it all tainted by such nasty politics.

But I cannot ignore those salutes, nor the myriad other slights starting with calling those Thai cave-diving heroes pedophiles. Tesla is dead to me, a victim of this insane time and its CEO.

The new model Y AWD non-performance does 0-60 in 4.6 seconds! The performance version is ~3.3; Ioniq 5 AWD is 4.6 and the N is 2.9 (!!!). For comparison the latest Corvette does 0-60 in 2.9 and the Z06 in 2.6.

These cars are insanely, incredibly fast. My G70 (gasoline, ~370 HP) does the jaunt in 4.5. That used to be considered a fast car, now it's just average (though the warranty is almost over, and I'll be modifying it to ~450HP).

TBH, electric cars 100% broke auto enthusiast circles. When a highly modified, very fast car just gets stomped by an electric car hauling a family of 4 it smashed that world to pieces. Especially in the early days, when EV enthusiasts were mostly Tesla techbro fanboys - who didn't really mix well with the oil, grease, gasoline, and DIY culture that was there before.

Interesting times for sure.

This is a part of why I (sometimes, depending) still use Aider. It’s a more manual AI coding process.

I also like how it uses git, and it’s good at using less context (tool calling eats context like crazy!)

This reminds me:

When I was at a newish job (like 2 months?) my manager said I "speak more in a Brittish manner" than others. At the time I had been binge watching Top Gear for a couple weeks, so I guess I picked it up enough to be noticeable.

Of course I told him I'd been binging TG and we discovered a mutual love of cars. I think the Britishisms left my speech eventually, but that's not something I can figure out for myself!

Many of us just don't use JSON in our day jobs, weird I know, but true.

The only thing I use JQ for at work is parsing the copilot API response so I remember what the model names are - that's it! TBH, I could just skip it and read the json

If you really think this, `baby` is an apt name! Internet, Smartphones, and social media will all be more impactful than LLMs could possibly be... but hey, if you're like 18 y/o then sure, maybe LLMs is the biggest.

Also disagree with missing the train, these tools are so easy to use a monkey (not even a smart one like an ape, more like a Howler) can effectively use them. Add in that the tooling landscape is changing rapidly; ex: everyone loved Cursor, but now it's fallen behind and everyone loves Claude Code. There's some sense in waiting for this to calm down and become more open. (Why are users so OK with vendor lock-in??? It's bothersome)

The hard parts are running LLMs locally (what quant do I use? K/V quant? Tradeoffs? Llama.cpp or ollama or vllm? What model? How much context can I cram in my vram? What if I do CPU inference? Fine tuning? etc..) and creating/training them.

The FCC doesn't enforce very strongly, mudduck and others like him have been killing CB channel 6 (and sometimes 19) for like 2 decades and nothing has been done nor will be done.

There's also opencode which is a fork(?) of Claude Code that runs on any model: https://github.com/sst/opencode

And of course, not the same, but Aider still exists and is still a great tool for AI dev.

It's interesting how everyone is suddenly OK with vendor lock-in, quite a change from years past!

I wonder if this can work with OpenCode (Claude Code fork which allows for other model providers: https://github.com/sst/opencode)?

I really don't like being tied to a particular provider or model, switching models has been really helpful to get past blocks and save money (especially with Deepseek!).

And, of course, I need to use Github Copilot's Open AI-compatile API at work...

Wild take lol, I was recently unemployed and once I started facing the very real possibility that I'd have to go work at the local sawmill (or UPS) for $20/hr I was willing to do almost any humiliation ritual these companies wanted... and be happy I at least had a shot instead of being ghosted again.

I support the fight against this kind of thing, but I also think it's entirely hopeless: They have all the power in this situation, and this is the future they're going to force on us.

Hyundai/Kia use a traditional 6-speed auto in their hybrids. I drive one, the engine stays at a few RPM stepping points during most driving (it likes ~1600RPM, ~2200RPM, and ~3600RPM). I had a Prius previously, and I like this different setup because it reduces "engine droning noise", which was terrible on the Prius.

Though H/K have recently introduced a new hybrid system with a CVT, so maybe 2026 or 27 model years will be different.

Since I'm only making one comment, I also want to say hybrid cars are better than ICE because there are fewer belt-driven accessories. Aircon in particular on an electric motor is a big improvement. Without the idling engine producing heat, hybrids are much nicer in hot stop-and-go conditions!

Also my Prius made it its whole life (200k miles and ~20 years) without ever changing the brake pads... amazing!

YUP! This is why I've settled on Aider and it's "IDE integration" (watches all files for comments that end in "AI!", which then invokes the AI). I can then use it with whatever editor I prefer. I view the vscode mono-culture as a bad thing. I also like I can use any AI backend I like, which is really how it should be: Vendor lock-in tools are bad, remember?

I guess you lose tab-completion suggestions, but I am not a fan of those compared to 'normal' tab-complete (if backed by an lang server). If I want AI, I'll write a short comment and invoke the tool explicitly.

EDIT: Of course, it really depends an your usecase. I maintain/upgrade C code libs and utils; I really cannot speak to what works best for your env! Webdev is truly a different world.

EDIT2: Can't leave this alone for some reason, the backend thing is a big deal. Switching between Claude/Gemini/Deekseek and even rando models like Qwen or Kimi is awesome, they can fill in each other's holes or unblock a model which is 'stuck'.

Folks don't want to hear it, but our careers are based more on luck than any of us would like.

I avoided the first big round of layoffs at my last company. How? By being so overwhelmingly awesome and valuable that they couldn't afford to lose me?? NO!

I agreed to a smaller pay increase that year so a team member could get a much-deserved larger raise. I accepted more stock grants to help make up the difference. The next round of layoffs was based on stock packages granted to employees. More stock granted == more better, right?

Sheer luck I didn't get the notice that time.

Wait, what’s wrong with just staying on emacs? I’ve been using it as my daily (systems&os dev) for over a decade.

I find when I use vscode that most features have an emacs equivalent.

Am I missing out on something in that set of “stuff I don’t know I don’t know”?

Huh, I guess my 401(k) has already been mistreated. I made most of the money I have for retirement when I worked at Intel, and seeing the lawsuit is really interesting/concerning.

Guess it's time to roll over the 401(k) to my new job and take a more active role in managing it - I want it almost all explicitly in the market (and not in private equity) at this point in my life.

Thanks for posting this, my retirement funds are a big deal to me and I hate that people less informed will probably be screwed by moves like this.

This reminds me of my first internship, I was running a test suite and filling out a web form every week. It was never mentioned in any meetings or other comms.

After about 3 months of doing this, I asked my manager why I was doing this if no-one cared or noticed. He told me to stop and see if anyone said something. I stopped, interned another 15 months and it wasn't ever mentioned.

Hell of a lesson to learn as a newbie

Another reason to at least learn the default tooling is that often I find myself SSHing to another machine which has only the barest of default packages installed (often busybox, sometimes just a stripped-down docker container).

If I didn't know how to use "find" and "grep" (though I prefer rg) then I'd be at a disadvantage in these situations. Also command-line git.

It's why I learned to use Vim well, though I daily Emacs.

IME it's not that bad. My entire network failed when I was looking for work: either everyone was still at my old employer whom I didn't want to return to or they were also out of work. I don't have much online presence, because that's my preference.

I did ~11 applications (on company websites, tailored resume), of which like 9 were moonshots (NVDA, Valve, etc). I heard back from everyone, and then interviewed and accepted an offer with a smaller international company located locally. This was during the 2023/4 downturn (Dec '23 to be exact).

Caveat: I have 15YoE and work in embedded (especially embedded Linux); it seems this specialization has suffered less than others. I also don't have a degree. I had to accept a slight paycut and hybrid - but I was in office before... and hardware generally just requires you to be present sometimes.

Don't be afraid if you don't have a network, the advice is good, but it doesn't apply to everyone.

I got hired by directly applying via a company's website; it was my 11th application and my first that wasn't a moonshot (nvda/valve/etc). It was in late 2023 (very bad market), and I didn't have any connections to the company at all. I also have no online presence. Online applications are not fully dead.

Of course this is just my experience as a senior engineer in embedded, so likely doesn't apply to others. But if you're looking, it's worth your time to apply...

This is me, I'm a renter and my complex will never install chargers. I've tried, it's a no-go.

That said, 50MPG vehicles are common these days. My '07 Prius gets 48, measured, a newer Camry/Accord/Sonata hybrid will get ~50MPG as well. Add an openpilot driving system and it's almost like a private train car.

I drive 80 miles round trip, 5x a week for work. That works out to ~$2000/yr for gasoline. That's really not that bad at all! Just don't drive a crossover or truck as your daily.

I'll probably never buy electric, because I don't want to buy a house (just not for me) and I don't think apartments will install a charger-per-spot (personal requirement). That's OK, hybrid is pretty great.

On the heat pump side, I only have to heat 700 sqft - it requires little energy and is so little cost-wise I don't even track it.

I'm a simple man, and did something similar but with bash+mosquitto+imagemagick running on a Kindle. That's boring/been done, but what I want to talk about is screen burn-in.

I've been displaying slight variations of the same image on an old Kidle paperwhite for a bit over 5 years! I've noticed NO burn-in at all. I'm doing partial updates every refresh (30 second refreshes), so the display will go 1+ years without a full refresh (~1 million partial refreshes a year). There's some retention when I first work with it, but clearing the screen and showing a full-screen pattern is enough to remove that retention entirely.

Pretty impressive, and I am also impressed the old firmware stays attached to wifi for the whole time as well! Sometimes it can have trouble reconnecting (power outage, dog unplugged wifi, whatever), that's when I'll see if there's burn-in, probably once a year(ish).

Getting flashbacks to Gentoo building when I only had a single core. Good learning experience…

I have some information on that I ran accross:

https://www.reddit.com/r/askcarsales/comments/15nkbh3/new_ca...

https://old.reddit.com/r/CX50/comments/126aepa/mazda_connect...

There’s a few threads linked in the top discussion (you may need to expand the massively downvoted responses, and some are deleted), but Mazda is known to use tracking data to deny warranty claims and share that data with insurers.

I was also stunned to learn salespeople’s commission is denied if they don’t get you on the app! Absolutely wild.