Hacker News Reader

Comment by dtx1 | Hacker News Reader

I think the main issues with all of these rasbi alternatives is the software. As ... quirky ... as the rasbi is, the software support and optimisation is excellent. There are several speed comparisons on youtube where the rasbi beats a lot of the competition through sheer software optimisation. Honestly, without complete Upstream Kernel support and Open Sourcing of all necessary drivers, none of these Boards is really useful.

View on HN

Allwinner based boards have mainline kernel support for most features, they're one of a handful of vendors that has mainline support (primarily due to community efforts). I personally like the OrangePi boards due to price & reliability, easy to set and forget without worrying about mSD corruption.

http://linux-sunxi.org/Main_Page is pretty useful for a list of boards and info.

nice, thank you

Not for any of the GPUs. If all you want is a terminal, you can pretty much buy anything..

You can run Gnome just fine, hardware accelerated video decoding/encoding works, HDMI with a few features is mainline, the only thing left is the Mali GPU, which is a fairly simple GPU as far as that kind of hardware goes, as compared to AMD or Nvidia GPUs.

On that note, there has been a fair bit of activity on Lima as of late, which is a big change from a few years ago: https://github.com/yuq/linux-lima/tree/lima-4.16-rc5/drivers...

Isn't Lima for 200 and 400 series. Tinkerboard needs has a 700 which already has kernel support but will need Panfrost for open source userspace.

Lima is for 200 and 400 series, which were what we were discussing earlier in this thread.

how mature is Panfrost? I have a board with Mali T720 (Allwinner H6) and I cannot get any GPU support currently.

Does libhybris work with Allwinner SBCs?

i.e. using ARM's proprietary driver for Mali from Android.

(FOSS drivers might be in development but pragmatically I'd choose something that works)

Not useable yet. It has shaders and can draw triangles but no textures or FBOs.

What makes Allwinner based boards less likely to have mSD corruption over others?

The orangePi boards don't use microSD, they have built in eMMC storage. Far more reliable.

On my 5 years old cubieboard server I experienced NAND corruption more often then not when power outage occurs. Never had this kind of problems since I switched to SD cards.

In my experience SD cards are much slower but more reliable and easier to manage (backups and such).

There's no real inherent reliability difference between SD and eMMC, it's all down to the individual controller implementation. Though some SD slots are quite flaky, i.e. the original RasPi models.

The eMMC that orangePi uses seems to be more reliable than the average SD card. Of course there are good SD cards out there and bad eMMC.

That hasn't been my experience during the design our last product. Allwinner was the vendor we had to pass on because their then-best-offering (H5) had poor mainline support and fragile, crappy code in the kernel dump. We decided to go with the slightly more expensive RPi 3B precisely because it had both usable mainline support and the handful of patches in the RaspberryPi Foundation's kernel were much better written and maintained.

We use H3 based boards, but the H5 seems to be fairly mature. I know its been under heavy development wrt mainline kernel support for the last year, might want to take a peek at the status matrix: https://linux-sunxi.org/Linux_mainlining_effort#Status_Matri...

H3 was already quite old back when we looked. The support offered by Allwinner for old chips (like the H3) back then was basically "If you're buying a ton, sure we'll support!", which was very off-putting.

H5's mainline support is still lacking in key areas. I2C, I2S, and cpufreq are deal-breakers for our product. Of course, combined with the fact that it will soon be out of support for a not-huge-buyer like us, it is out-of-question.

The status of their latest offering, the H6, in that matrix shows how bleak it was when we were looking at H5 for our board design. It was when the latest release of Linux was 4.4. You can see how few H5 features were supported with 4.4, almost exactly like how few are supported on the H6 today.

Sure, none of these chips are new, and the older models (which are easily & cheaply accessible) have mainline support. I wouldn't expect jack from Allwinner besides their crummy kernel 3.4 branch, but that is the beauty in mainline support, you aren't limited to what Allwinner supports!

> Sure, none of these chips are new, and the older models (which are easily & cheaply accessible...

The then-older chips were not available with the future volume we needed, unless we increased our volumes a few times. We had little choice but to build on their newer chips for the volume bracket we fell into. Perhaps the situation has changed in two years.

> I wouldn't expect jack from Allwinner besides their crummy kernel 3.4 branch, but that is the beauty in mainline support, you aren't limited to what Allwinner supports!

You can see how that is off-putting, especially when other vendors (e.g., the Raspberry Pi Foundation) themselves support the latest LTS (and non-LTS) kernel with maintained patches, and upstream their patches. And their code is not utter garbage. New RPis have a much easier time getting mainlined precisely because of these reasons.

I'd have thought they'd still be better than the chip in the Pi, which last I checked wasn't available at all to anyone but the Pi Foundation (and the proprietary bootloader license didn't allow you to run it on anything non-Pi anyway).

The licensing and supply for the RasPi were pretty off-putting to us. Sure, we aren't making a custom board, but I don't want to be held captive to a single supplier that can stop producing boards at any time, whether that is due to a lack of chips or a lack of desire.

We were afraid of that too. But Allwinner wasn't willing to give us any guarantees they'd even continue to make the H5 two years later. Nor was the H5 backwards compatible with either of its predecessors H3 and A64, giving us more reason to worry about the future.

The Raspberry Pi Foundation (and Broadcom, in contract partnership), on the other hand, had a proven track record of both continuing to manufacture RPis 1 and 2 and to maintain backwards compatibility when putting out new RPi SoCs and boards.

Perhaps the situation with Allwinner has changed now. Perhaps they're willing to support their products for longer for smaller clients. We'd have to look into that again. But, of course, without mainline support anytime soon (and the quality of their past code doesn't help a bit) it's very hard for us to actually build a good product with it.

It might be better in principle — though shitty code is only so much different from unmodifiable code — but it turned out to be the best compromise available in practice. The boards put out by the Foundation were cheap and good enough for our purpose. Of course we'd have liked to go with our own boards. We actually designed and commissioned a small batch of builds using the H5. But then we saw the software provided by Allwinner. It was so fragile, it wouldn't even boot correctly a second time. Looking at the code itself revealed it to be a nightmare. And at that time (over a period of three months) there was no software freedom to be had since the only software that would even boot at all was the shitty dump. We'd have had to wait a year or two for the code to get usable or clean, and by that time both Allwinner would have moved on from the H5 (as we were told) and our product would have been too late to market.

Long story short most other boards are not really cheaper if take into assumption that your time is not free.

Unless you enjoy hours of wtfs googling and trying to figure out why does this thing does not do what you want it to, then you are spending a few hours to save a few bucks. You can judge how rational that is based on your hourly rate. Just because RPi is popular it's way more likely that somebody has already solved given problem for you.

So the only argument would be that other boards are more powerful. Does that really matter in 99% of use cases though?

All I really want is a pi with networking that isn't forced through a USB 2.0 channel. I'm genuinely surprised that this is still an issue, even with their most recent hardware.

There's the olimex fully open source computers that have everything documented and don't need binary blobs to run.

https://www.olimex.com/Products/OLinuXino/open-source-hardwa...

The Armbian community provides a distribution that is not only optimized for almost each and every ARM based SBC board out there, but they provide a build environment that allows you to individually tweak and tinker your own image.

http://armbian.com/

Software can be fixed by a large enough number of developers and users. Where the Raspberry Pi wins is community. I've got so many other boards but none of them work quite as well as the Pi.

Mainline kernel support is a big big issue, and it can't always be fixed by the community. There are a number of well supported Allwinner boards out there at very competitive prices though, we use them in production due to their price, lack of mSD corruption issues, and bandwidth (all the USB/Ethernet ports are directly wired on the OrangePi PC+, no bottlenecking like on the Raspi).

Where the Raspberry Pi wins is community.

From what I've observed, the community feels almost like a walled garden in terms of mentality: if you're doing what everyone else is doing, it's great. If you try to go beyond that, don't be surprised if you encounter hostility or discouragement.

By "go beyond that", I mean simple things like trying to find the electrical specifications of the GPIO pins. The comments here are worth reading, for example: http://tansi.info/rp/interfacing5v.html

What type of optimization you specifically refer to, and which benchmarks? Changes to the ARM CPUs across producers are minor, and the packages are also practically the same (the distributions are few).

> without complete Upstream Kernel support and Open Sourcing of all necessary drivers, none of these Boards is really useful.

it's not clear if you refer to SBCs in general, or Pi alternatives. nonetheless, essentially all of them have Linux support (at least, Armbian support); of course, having upstream support would make them more usable, but this is different from "useful".

I have to say, having tried both the RPi and Tinker Board, the Tinker Board's GPU driver is far better. Yes, it's proprietary, but unfortunately the RPi driver is simply not reliable enough to run something like WebGL in Chromium. It's slow and glitchy. The Tinker Board runs WebGL like a dream by comparison.

This is my take away also. I have all the rpi models, a few beaglebones, banana pi, and odroid.

If there is something specific you are trying to do, you'll want an SBC that is geared towards that purpose (sata, adc, etc). If you just want an SBC as a general computer, rpi is your best bet.

The Rockchip RK3288 SoC used in this particular board is used in some chromebooks and uses ARMv7 on a 32-bit processor which I think may indicate decent kernel support

Reply on news.ycombinator.com