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

Isn’t that less efficient in power and can have different color?

It has to be an SMPSU in constant current mode, not a linear power supply. I'm sure it does still affect the colour, though.

It’s because building 500 units would be a non starter for many of them

Some of devices automatically lock once used on a carrier network


I don’t believe that’s true. Locking is entirely something done by carriers for promotionally-priced phones attached to active phone plans.

Can you point me to anywhere documenting your claim? That would definitely be a new low if true, but I really don’t think that’s how it works.


Sadly it is true. It’s called Flex Lock or US Carrier Flex Policy: https://www.usmobile.com/blog/what-is-flex-lock/

The gist is that it allows a third-party seller to stock a bunch of identical, not-yet-locked phones and offer a choice of carrier plans. The phone binds to whichever carrier the user first activated on.

So if you’re buying a phone, verify it is not one of these units.


If you buy them unattached? First time I hear of that, do you have an example or source?


Thats some lawsuit level bullshit.


Not in EU


No they don't


>If the goal is consistency, then wall-clock isn't the simple way to do it.

It’s simpler than doing a limit on number of states, and for some applications consistency isn’t super important.

Doing a time limit also enforces bot moving in a reasonable time. It puts a nice limit to set up a compromise between speed and difficulty.

Doing state limit with a time limit might be better way to do it, but is harder.


> It’s simpler than doing a limit on number of states

According to who?

A counter that you ++ each move sounds a lot easier to me than throwing off a separate thread/callback to handle a timer.

> Doing a time limit also enforces bot moving in a reasonable time.

It's designed for specific hardware, and will never have to run on anything significantly slower, but might have to run on things significantly faster. It doesn't need a time cutoff that would only matter in weird circumstances and make it do a weirdly bad move. It needs to be ready for the future.

> It puts a nice limit to set up a compromise between speed and difficulty.

Both methods have that compromise, but using time is way more volatile.


Why?


Does X even support hdr yet? When I looked last time the answer was none


Yeah but you can get a n100 on sale for about the same price, and it comes with a case, nvme storage (way better then sd card), power supply, proper cooling solution, and less maintanance…


The Orange Pi 5 Plus on its own should be much cheaper than an N100 system. Only when you add in those extras does the price even out. I bought mine in an overpriced bundle for 182€ a few months ago.

It supports NVMe SSDs same as an N100.

Maintenance is exactly the same; they both run mainline Linux.

Where the N100 perhaps wins is in performance.

Where the Orange Pi 5 Plus (and other RK3588-based boards) wins is in power usage, especially for always-on, low-utilization applications.


You can get an n100 system for $110 on sale. Price went up but I still see $135 on eBay now. However YMMV because Europe prices are different

For power I don’t know about orange pi 5 but for many SBC power was a mixed bag. I had pretty bad luck with random SBC taking way more power for random reasons and not putting devices in idle mode. Even raspberry pi was pretty bad when it launched.

It’s frustrating because it’s hard to fix. With x64 you can often go into bios and enable power modes, but that’s not the case with arm. For example pcie4 can easily draw 2w+ when active. (The interface!)

See for example here:

https://github.com/Joshua-Riek/ubuntu-rockchip/issues/606

My n100 takes 6W and 8w (8 and 16gb). If pi5 takes 3w that’s not large enough to matter especially when it’s so inconsistent.

Now one place where I used to like rpi zero was gpio access. However I’m transitioning to rp2350 as it’s just better suited for that kind of work, easier to find and cheaper.


I have no idea what US prices are like but I put in a reasonable amount of effort and at least right now here in Europe, N100 and RK3588 prices are pretty similar for comparable packages (RAM, case, power etc.). One other thing to note is that the N100 is DDR4 while the RK3588 uses DDR5.

I never ran into that bug but I came to the Orange Pi 5 Plus in 2025, so there's a chance the issues were all worked out by the time I started using it.

Looking at a couple of reviews, the Orange Pi 5 Plus drew ~4W idle [0] while an N100 system drew ~10W [1].

1W over a year is 8.76kWh, which here costs ~$2. If those numbers hold (and I'm not saying they do necessarily but for the sake of argument) and with an estimated lifespan of 5 years, you might be looking at a TCO of $140 hardware + $40 power = $180 for an Orange Pi 5 vs. $140 hardware + $100 power = $240 for an N100. That would put an N100 at 33% more expensive. Even if it draws just 6W compared to 4W, that's $200 vs. $180, 11% more expensive.

I'm not saying the Orange Pi 5 Plus is clearly better but I don't think it's as simple as one might think.

[0]: https://magazinmehatronika.com/en/orange-pi-5-plus-review/

[1]: https://www.servethehome.com/fanless-intel-n100-firewall-and...


What why?


Very little progress made this year after high profile departures (Hector Martin, project lead, Asahi Lina and Alyssa Rosenzweig - GPU gurus). Alyssa's departure isn't reflected on Asahi's website yet, but it is in her blog. I believe she also left Valve, which I think was sponsoring some aspects of the Asahi project. So when people say "Asahi hasn't seen any setbacks" be sure to ask them who has stepped in to make up for these losses in both talent and sponsorship.

https://rosenzweig.io/blog/asahi-gpu-part-n.html



Marcan (Hector Martin) resigned from Asahi Linux early this year [0].

Asahi Lina, who also did tons of work on the Asahi Linux GPU development, also quit as she doesn't feel safe doing Linux GPU work anymore [1].

[0] https://marcan.st/2025/02/resigning-as-asahi-linux-project-l...

[1] https://asahilina.net/luna-abuse/


marcan and asahi lina is the same person


GP's LKML link is very recent unlike your two links, implying something could've changed.


I have no insight into the Asahi project, but the LKML link goes to an email from James Calligeros containing code written by Hector Martin and Sven Peter. The code may have been written a long time ago.


Someone posting Hectors code or a quote does not mean he didn't leave. I'm really not sure how that could leave that impression.


Because key developers have left the project, and developers who are capable of such work are few and far between.


>are few and far between

They are more common than you would think. There just is not many willing to work on a shoe string salary.


> There just is not many willing to work on a shoe string salary.

You explained it well by yourself.


It's really hard to do and nobody is paying for it?


One of my friends ate two pounds of tomatoes every evening and would not feel full. I couldn’t believe it until I saw it with my eyes….


I feel like they should try the potato diet. I recommend it, not because I believe it is effective (it wasn't for me), but because it is very good at removing the pleasure from eating. When you are eating to survive, you will come to understand how what being sated is. From there, I found I could lose weight because I had broken the habit of eating for any other reason than to fuel myself. Although, admittedly, I enjoyed losing weight this way better than the potato diet.

With some people, I think they have just lost what being sated feels like.


Something like less then 2% of people keep weight off for more then a year from dieting alone


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

Search: