Hacker Newsnew | past | comments | ask | show | jobs | submit | Scrounger's commentslogin

> Changing email providers sounds daunting, but it was relatively painless

If you have your own custom domain, changing providers is pretty easy, IMO. In fact, the portability factor is probably the best benefit of having your own domain.


Some TLDs get rejected or autocorrected by broken validators. I've had foo.dev changed to foo.de.


I was on the application for the .name top level domain back in 2000 and then there was a reasonable excuse for that - we were in the first new batch of TLDs. We had to contact a lot of site owners to try to smooth the path. But given the number of new TLDs since, you'd think people would've learnt.


I have a .name account and still occasionally run into sites that don't accept it.


This drives me insane. Lazy input boxes that only accept email domains from the big names.


Tradeoff: Failing to accept a limited number of rarer domain. Vs. all the errors caused by idiots who can't spell mgail.com, or gmail.cmo, or gmali.com, gmail.kom, or whatever the heck that thing is.


Excluding people from services because you can't be bothered to handle their perfectly correct and validated email address is actually a form of discrimination.

My pet hate is my bank refusing to email me bank statement notifications to mail@mydomain.com ("not a personal address"), yet their password validation and marketing emails get through quite OK. Somehow they seem incapable of acknowledging that one department of the bank has made a technical decision in conflict with other departments...


Is it discrimination if you decide to switch to custom power outlets in your house and then you complain that local hardware stores don't provide compatible parts?

> Excluding people from services because you can't be bothered to ...

Yes. But like most forms of discrimination, it's probably 100% legal. Same as Chuck's Pizza Parlor refusing to deliver outside an X mile radius, and 1st National Bank refusing to use the full legal name of anyone who's name exceeds the size of their DB's 'CustomerName' field.


"Not illegal" doesn't mean we should passively accept it whether we encounter it professionally at work or personally as a customer.


True. Vote your wallet as you will. Beyond that - discretion is the better part of valor, the low-level employees you can easily complain to are very unlikely to have any say in the policy, and Chuck is free to ask you to drive 40-mile pizza deliveries for the same pay and mileage as he gives at his 10-mile limit.


> I'm running these things in headless VMs

What's your setup like and what do you use it for?

I have a M2 Max MBP with plenty of ram and I use VSCode + Zoo Code plugin with Qwen3-Coder-Next-GGUF:UD-Q4_K_XL to run local agentic coding sessions, but I'm intrigued by being able to run headless as I could probably run multiple instances in parallel to do stuff?

Like are you using UTM with some pre-built VM and a local LLM?

Curious.


Location: San Diego, CA

Remote: Yes

Willing to relocate: Maybe.

Technologies: I'm a full-stack growth marketer with enough coding chops to prototype MVPs. Worked in the blockchain sector for the past 7 years and would like to transition into an AI-enabled consumer product.

Some of my writing (before AI/LLMs) straight from the dome: https://medium.com/@dilanka

In other words, I'm a straight up marketing nerd.

Resume/Work: https://tinyurl.com/bdzez9cv

Email: hackernews@notify.dilanka.cc


Interesting site/challenge; however, I had trouble browsing and finding "what to do" in a reasonable time.

I recently spent like $170 giving a new lease on life to a 15-year-old Lenovo S10-3 Ideapad with a 1-core Intel Atom CPU, 2GB of RAM, a WiFi card, and a 250GB SSD running AntiX Linux in TTY/Command Line mode.

So far, I've turned it into a picture/frame + vision board running Tailscale so I could SSH in and/or rsync stuff.

I am also attempting to run a no-AI version of Pwnagotchi to pwn WiFi networks.

I am also using it as an always-on appliance that does stuff like rsync/backup my entire server, run lightweight Python scripts to check the uptime and days until domain expiration, etc., on a set of websites I own and would like to own, etc.

I have all of this stuff connected to a Telegram bot that reports to me.

It's an interesting set of constraints, and you can surprisingly do a lot of cool stuff.


I gave my wife a 2012 macbook pro running ubuntu. It was a huge upgrade over her ~2014 macbook air running stock, which couldn't even update itself anymore.

It never occured to me to consider it might qualify as a "challenge" since it is 14 years old. It just works fantanstically and was my daily driver until 3-4 years ago.

I got it off ebay for approximately $100, cleaned it, and put in a new battery.


The 2012 macbook pro (non retina, at least. I never owned a retina one) is the last truly great macbook apple ever made. God I loved that thing, I used it to death and was truly saddened by its loss. It lasted a good 12 years


I had an 11” 2011 air. An incredible machine, one of the best I’ve ever owned. I stupidly replaced it with the 2016 retina touchbar MacBook Pro - which is hands down the worst Mac I’ve ever owned bar none. My modern M1 is fine. But that little air was somehow more fun.


Here’s an idea that’s been following me for a while, if you like low-level stuff:

Make a toy OS that boots into a Lisp shell.

Another to appreciate how fast computers that we call old effectively are: write a game for the shell. Depending on your level of skill, you can try pong, snake, lunar lander, or a 3D software renderer.


OLPC (one laptop per child) had Open Firmware, a Forth bootloader/firmware

https://lwn.net/Articles/209301/


Open Firmware was originally a Sun thing, then Apple adopted it at some point as well


Make something that makes music.


> 15-year-old Lenovo S10-3 Ideapad with a 1-core Intel Atom CPU, 2GB of RAM, a WiFi card, and a 250GB SSD

What a weird coincidence, I've just found one of these while clearing out a box of equipment I'm getting rid of and thought "I should stick NetBSD on this!"


Trying to use a 15-year-old Atom netbook as a modern laptop is mostly pain. But treating it as a small always-on appliance is a much better fit


> I recommend Magic Earth.

LOL Bruh... this has a 1.7 rating on Android based on 42k reviews

https://play.google.com/store/apps/details?id=com.generalmag...


It used to be really good, and then it went to a subscription model, with a lot of back-and-forth uncertainty about the change. I suspect the rating reflects that.


> Like allowing apps to restrict screenshots. I own the phone. If I want a screenshot, it should screenshot. This is not something for apps or Google to determine, and if the OS listens to me (not the app) it should allow screenshotting the display 100% of the time regardless of what the app cries about.

PREACH!

I hate this.


> What about banking Apps? No problem there?

Most banking apps work, but Google Pay/NFC payments won't work.


If your bank doesn't have its own payment app, but PayPal Card is available in your country, I got this setup to work on Graphene by installing PayPal into the Secure Folder, install the sandboxed Play Services there (required) and setup the mobile payments in the app. It's not a great solution, but for I keep it around for emergencies.


I know a handful of german banks that have their own nfc payment apps that still work in Graphene


Google Pay may not work, but NFC payments through yiur bankapp probably do. They did for me.


> May God protect us.

Today, data systems and algorithms can be deployed at unprecedented scale and speed. Unintended consequences will affect people with that same scale and speed

Michael Chapman


> Also M2 batteries are about maybe halfway done already unless the refurb replaced the battery.

My mom still uses a 2019 Macbook Air with 8GB of RAM. The battery requires servicing, but she's unaware and still using it just fine. I asked her to go to the Apple Store and get the battery replaced along with her iPhone 12 Pro Max battery, and she'll easily get 10 years out of each device.


> why Apple has nearly wiped it off their site all together, seriously go check out apple.com, not a single mention of Apple Intelligence.

TBF, the reason Apple removed "Apple Intelligence" is that they failed to deliver on its promises.

So much so that they just settled their false advertising in a class action lawsuit for $250M:

https://archive.is/efWkb

Also, P.S: Not to say that clothing/shopping is the primary use case, but I know plenty of women who use AI for clothes/fashion/interior decoration etc related tasks.


What feels already like old history is that Apple made a generous deal to OpenAI based on the premise that their AI could do the claims.

Apple engineers spend months trying to prompt engineer their way, thinking the prompter is at fault if the soon to be AGI system diverged. Some of these instructions were trending out there, as reveals of how naive Apple was at the time. They could be traced from the device's logs so not so much of a leak: Don't hallucinate, strictly follow instructions, followed by all sort of refined predicates, appended as if an LLM had reason

Then Apple released a paper to warn everyone (well, a few, and to save face) that we are getting fooled.

https://ml-site.cdn-apple.com/papers/the-illusion-of-thinkin...

In case Apple is a biased anti AI propagandist, here is a similar, more recent research paper from MIT and co:

https://arxiv.org/html/2603.24755v1


its really hard to read that first apple paper in context when it doesnt have a date on it. I know research papers are meant to be timeless artifacts, but when it says things like "Recent generations of frontier language models... " "frontier LRMs" etc i'd like to know what they were testing on and what the zeitgiest was around that time.

Please put a date on your research papers! I could figure it out roughly by looking at the "last accessed" date on their citations - 2025-05-15.


The specific models are specified. E.g gpt version's, deepseek R1 etc. but I agree that's terrible practice not to timestamp a paper aside the authors.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: