Support of Raspberry Pi

[tkaiser]

4 minutes ago, zador.blood.stained said:

without even thinking about contributing a single line of code or a single dollar for the hosting, hardware and consumables

It's even worse. Certain 'users' here constantly sabotage the work at least me is interested in by shitting their 'valuable opinions' constantly in threads and asking moderators to do some housekeeping later if they're not happy with it. I prefer to focus on development and not waste my time on housekeeping to keep threads readable and asked Igor to remove my moderator status now and will also leave this forum entirely for some time since it got such a mess in the meantime (opening this up to the Rasperry userbase might greatly improve this mess). 

[hojnikb]

@tkaiser

There is no need to read the forums or bother what other people say. 

 

While opening up armbian might be great to get new users/devs interested in, it also makes it one in many distros, that are available for rpi. It might make more sense to focus on boards (like allwinner) that are not well supported by other guys.

 

 

[tkaiser]

On 3.1.2017 at 5:38 PM, zador.blood.stained said:

There is an alternative, though pretty limited at the moment: https://github.com/christinaa/rpi-open-firmware

 

And here the main contributor to this project explains in detail why she stopped working on Raspberries: 'since rpi is shit and same could be said of rpi foundation': https://irclog.whitequark.org/linux-sunxi/2017-04-02#19158422  (also a lot of other technical details are mentioned there that are worth a look for anyone wondering why developers don't want to touch the 'inner workings' of these Raspberries and only fiddle around one layer above in userland).

[Landev]

TNX to tkaiser for in depth explanation of the Raspberry PI of the proprietary bootloader. Personally i own a Raspberry 1, a Raspberry 3 an Orange PI PC and a Tinkerboard and my question is now, can i contribute somehow eg. to the open firmware project (maybe a question for hojnikb) and how the policies of your project are? You only work with uboot and there is no option for a debian like "non-free" section for raspberry pi eg with such a script. https://github.com/braindef/debian_base_on_raspberry3 (i took from someone else and modified it for debian stretch)?

[slave2conquest]

You know, for a group which complains about the RPi foundation being bullies, you sure are a bunch of bullies.

 

I build custom debian (not rasbian) images for the RPi for robotics projects, and I would be willing to help, but why would I bother if people like tkaiser choose to be combative?

I, personally, have incentive to move RPi to this project, but I need to know this isn't a lost cause, politically.

[tkaiser]

1 hour ago, slave2conquest said:

I build custom debian (not rasbian) images for the RPi for robotics projects, and I would be willing to help

 

Helping with what?

 

BTW: I build custom debian (not raspian) images for the RPi for NAS projects (even if Raspberries are such lousy devices for this use case) and I try to ensure that this never will destroy the Armbian project.

[Marc Draco]

Although I used DietPi (clue in the name) for my daily NAS needs, it's running on OrangePi PC under and Armbian kernel.

Personally, the moment I learned that Upton had sold out to Google (partnered my arse) I determined that I wouldn't be giving them another moment of my time NOR any money.

Google is taking over the Internet bit by bit and it's like a damn cancer. I'm becoming convinced it's part funded by the CIA in an effort to spy on us all more effectively.

[Igor]

Solve simple problems first Dietpi is also selling you bullshit and you don't even notice. There are absolutely no technical advantages/optimisations at least towards Armbian. In fact, Diet runs slower in just about every case and on any hardware. Also on Raspberry Pi plus it is not just about speed or other technical advantages but security, support and trust.

 

I knew that RPi is bad and has bad karma and ofc Google is evil too and they stick together very well. Power corrupts if nothing else.

[jsilence]

Hey, I am new to Armbian, but not new to Linux, kernel building or Raspberries. Running a Tinkerboard and an Odroid HC1 at home right now and also toying around with the Armbian build chain! Thanks to all the active developers and the community for providing Armbian! Much appreciated!

 

I was wondering why apparently there is no Raspberry Pi support. Am I missing something? Is it too obvious? Or is there really no image for the grandfather of SoCs?

 

I have a couple of RPi3s laying around which I would like to activate. I'd like to run all my small machines with the same OS to keep maintenance and learning curve low.

 

[martinayotte]

Earlier in the current thread :

 

[Xalius]

RPi is not 'The grandfather of SoCs', if you think that, you are about 10 years late, and it is a good example of how you can still make a good project of obsolete hardware by having enough people buy into the hype...

[TonyMac32]

 

2 hours ago, Xalius said:

RPi is not 'The grandfather of SoCs

Indeed.  I was playing with Linux on the HP Jornada 720 and the Mini 2440 from friendlyARM before the RPi was even a rumor.  Oh, I forgot, I also have an ARM7TDMI board here somewhere...

[Larry Bank]

RPI is married to Broadcom because that's where their engineers used to work. They cobbled together their SBCs from older, inexpensive parts to be able to hit their "amazing" price points. The Broadcom SoCs were designed for set top boxes and are not necessarily the best for Linux SBCs due to their long list of limitations (e.g. slow USB, slow+old ARM cores, slow+limited RAM). For me, the worst offender is the Raspberry Pi 0. It's 11 year old technology (armv6) released with a "teaser" price that was never meant to succeed as an actual product. The other end of the spectrum for offensive Linux SBC board vendors is Qualcomm. They tease out their overpriced chips in a watered down board with odd and half-working (e.g. GPS) parts to get their feet into "this IoT market thing". The Chinese chip vendors are the only ones who are willing to put their best chips (e.g. RK3399) onto publicly available boards running Linux. I have a feeling that this is more of a "try anything to conquer all markets" than a real understanding or effort to help the IoT community.

[chwe]

People misunderstood the use-case of a RPi...  Just have a lock on their main page:

Quote

A small and affordable computer that you can use to learn programming

 

IMHO during the time the RPi1 was developed there wasn't that much effort by others targeting this usergroup. And it's still hard to find a good competitor on this field. Could you get a more powerful SoC for ~40$? Sure... A better software stack for getting in touch with linux and "programming" probably not or it depends on usecase...  You get a more or less recent kernel with hardware accelerated video decoding, a lot of 'learning hats', a lot of opportunities to start programming immediately (python, scratch, Mathematica etc.). I don't think that things like Mathematica will be there if the SoC is too powerful...  

But people try to blame the RPi for a lot of things (e.g. USB sharing bandwidth), but IMHO it helped a lot 'opening' this market. There wouldn't be any orange- banana-  etc. Pi without the RPi. Is the RPi and their Raspbian my favorite SBC&Distro? Not really, but I've still ongoing projects where I never would go away from the RPi as long as others do not improve (more or less every project which doesn't run on CLI or where a camera is involved..). And if you consider about long term support, for the price of ~40$ you get really a long time support for your device (RPi1 is still supported, whenever this has some drawbacks for the more recent devices)..

Do I wait for the next RaspberryPi? Not really, for the use cases I see for an RPi, model 3 does what it is supposed to do. 

For most of my use cases the RPi isn't 'the best board'.  Most things works without issues on the OPi0 (the shitty rev. 1.4, mostly IoT stuff), HC1 (mostly file server stuff) and my small webserver runs on a OPi PC+ (GbE would be nice, but it works OK-ish without GbE).

 

Do I think the RPi should be supported by Armbian? No, not really, I'm often annoyed when working with Raspbian on CLI cause things are 'slightly' different compared to armbian and having one OS on all SBCs would make it easier.. But on the other hand, you spot new things when working with other projects (as example, I saw NodeRed first on a RPi project, in the mean time all my NodeRed projects are deployed on Armian SBCs).

Armbian shines most where the boardmaker doesn't do the job he's supposed to do - deliver a good OS with a recent kernel... RPi does this job quite good. The only feature (I see) armbian users would have a real benefit when providing Raspberry images would be camera or hardware accelerated video decoding but to achieve this we would deal with another 'BSP kernel' which also means that developers time  is shared with another SoC (I don't think that they're bored because of lack of work to do.. ). 

[tkaiser]

9 hours ago, chwe said:

IMHO it helped a lot 'opening' this market. There wouldn't be any orange- banana-  etc. Pi without the RPi

 

You might better try to educate yourself about what really happened. Bananas and Oranges are Cubieboard descendants which itself is in a line with Mele A1000 (or more in general: Allwinner A10/A20 based Android devices that were popular in China and were exported later by Tom Cubie which kickstarted more or less linux-sunxi community, Cubietech and sites like CNX). These origins on A10/A20 happened in 2012 when RPi software support was still very limited. Below one rackmounted Mele A2000 cluster (real Ethernet and real SATA combined with SSDs made the difference to toys):

 

 

Even the power plug used on Bananas and Oranges is inherited from those first Mele TV boxes!

 

At the time RPi was in early development Beagleboard was already there, ODROIDs were already there and in China they had something similar to RPi already years before (QQ2440 and Mini2440 are FriendlyARM products, yes the company now selling NanoPis since Westerners are that stupid that they only can accept a good SBC design if it has Pi in its name)

Quote

Mr. Yang's QQ2440 and Mini2440 from 2007 were the Raspberry Pi of China. Every student in the country wanted one and they went hundreds of miles by train to reach a shop with stock. College textbooks were published that were exclusively about ARM architecture and programming.

 

When speaking about RPi it's more or less about Western perception and of course marketing (that's where the RPi Foundation really excels)

[chwe]

5 hours ago, tkaiser said:

RPi was in early development Beagleboard was already there

I know... 

It's still laying around...  And somewhere there's a small metal box laying around with beagle bone (I think it was the first generation of the beaglebone). There was also a tablet before apple presented the iPad.

It's not about who was first, it's about who produces an attractive product for an attractive price (I think my bones where around 90$ each). The RPi1 was somehow similar to the RPi0 at the moment (you could only buy one, and the chance to get one was really low - when mine arrived resell prices on ebay  where around 90-100$ ).

There where better products but none of them fulfilled  to grow such a big community (in western countries).  There's a similarity to the arduino project. I had an AVR programmed with basecom on my breadboard with a crystal and was able to let a led blink.  There was the AVR Butterfly, probably the 'best project' to start with microcontrollers in those days... 

but there was 'no' community around those projects. There was 'nobody' who tried to explain us idiots how to use those MCUs what stupid led blink projects we could deploy on those boards, no open available i2c, spi libraries for all those nice little sensors to work. Now we have all those small MCUs (various Atmels, STMs ESPs etc.), someone started to port Python for the more powerful devices and Adafruit/sparkfun made cool little sensor boards inclusive their drivers only to wait a month until  the 'same' board pop up on aliexpress from a chinese manufacturer which only makes the board without driver ('you can use Adafruit driver it works good') for half of the price(different story )...  

Every boardmaker which tries to catch a part of the 'customer cake' has a 'RPi compatible' pinheader. It's annoying but they define 'standards'. Everyone builds microUSB powered boards, a lot of boards have the RPi size to press them in the smallest cases you can find which ends often in terrible overheating devices, 'nobody' places the SoC on the right side - it should be on the bottom to cool them with a case with an aluminum baseplate --> OK, the HC1 did it right, the EspressoBin too, but most boardmaker do it the same annoying way than the raspberry pi does it, everyone places as much USB connectors as possible (mostly 4) on the board, even if you don't have any chance to power them properly (4x 500mA only for USB). All beaglebones/boards had a nice barrelplug years ago, they knew that a microUSB might be a shitty decision for powering a SBC but we have 2018 and there are still boardmakers which deploy microUSB powered boards cause RPi 'defined' it as a 'standard' (I'm sure, if they would power the RPi with a 5V 4.1mm barrelplug since model one, we would see a lot of barrelplug powered boards...  ).

The RPi might not be the 'grandfather' of all SBCs but it was the growing incubator to got more attention to those little ARM boards (OK, an incubator in western countries... )...

 

 

 

[tkaiser]

15 hours ago, chwe said:

The RPi might not be the 'grandfather' of all SBCs but it was the growing incubator to got more attention to those little ARM boards

 

Exactly. The average RPi user often has really no clue at all why he bought an SBC (and you should please stop spreading this 'charity/education' BS since 'Pi for education' reality looks different). Users choosing any of the alternatives did at least some research and thought about what they want to achieve and how. And this (the user base) is the real reason why Armbian should never start to support Raspberries since once these people start to arrive here in the forum Armbian is finally dead (but maybe the current approach to semi-support as much SBC as possible will already kill the project).

[chwe]

Note: this post is filled with a lot of off-topic, personal opinions, sarcasm etc. In case you're not interested, skip it.

 

8 hours ago, tkaiser said:

(and you should please stop spreading this 'charity/education' BS since 'Pi for education' reality looks different

There's a difference between 'charity/education' and 

On 18.1.2018 at 12:27 AM, chwe said:

A small and affordable computer that you can use to learn programming

Just here in switzerland we have (or we had, I'm not sure anymore) more than 4 school systems ('primary school' between 5 and 7 years, different types and flavors of 'middle schools' etc.) spreading this to europe/world and you'll end in a bigger mess than supported kernels by armbian.  I can't talk for other countries, but it seems that there's no time/willingness to 'learn programming' at school. Languages seems to be a 'more important' topic - hallo Thomas, or in french 'bonjour mes amis qui parler une langue terrible' @martinayotte (sorry, you're the only one here I know who speaks french and I was really bad speaking it at school - so feel free to correct my grammar in french.  ) or italian 'Ciao a tutti!' (in fact, italian is mostly optional).  But to go some years back: 

From this lego technic set:

I learned how a gear transmission worked.  (good old times.. --> for a 'good working' gear you need: a stable frame and different gears)

The RPi (or arduino) could be a similar 'product' for a bit older kids. You need a RPi, a mouse, a screen (doesn't have to be 4k capable), a keyboard, SD-Card &reliable PSU.

Does the RPi foundation a good job? I've no idea, I don't follow their PR stuff. But if I had to decide which SBC I would buy for a kid to 'learn programming' it would be a RPi, simply cause there's a lot of stuff done. If something gets corrupted I can simply burn a new SD-Card and I don't have to set up a whole computer. From the start all those small tools are there and I don't waste time in setting up a system. 

 

9 hours ago, tkaiser said:

Users choosing any of the alternatives did at least some research and thought about what they want to achieve and how.

 Is there a cheaper board than the RPi --> Ok, I'll go for an OPi0 [/sarcasm] IMOH it's not RPis fault if users buy their product and do not know what they want to achieve with it. Otherwise it would be the butchers fault if he sells you a filet and you make mince for a burger out of it. People buy stuff they don't need, people expect things from a SBC which the SBC can't deliver and companies lie to their customers with false advertising. 

To quote Getrude Stein:

Quote

A rose is a rose is a rose

So  to bring it to the 21th century: 'A SBC is a SBC is a SBC'. It's not a TV-Box, it's not a desktop replacement, it's not a professional server it is a single board computer. To go a bit further:

Quote

In Stein's view, the sentence expresses the fact that simply using the name of a thing already invokes the imagery and emotions associated with it

• Octacore is better than quadcore
• 4 USB ports means that you can use all of them @500mA
• Class10 SD-Card means fast (for SBC purposes)
• microUSB powered means: 'I can use my old phone charger and save the money for buying a reliable PSU'

As armbian community we can complain about false advertising and false expectations from users, we can try to educate new users what they can expect from armbian and/or from one of the various supported SBCs, we can drop support for boards which make too much trouble but we'll never fully avoid that people buy the false board or join the community with false expectations and blame armbian for 'why are you not able to deliver a good OS for my needs and why you didn't solve all my problems first cause this is a major issue'. 

10 hours ago, tkaiser said:

And this (the user base) is the real reason why Armbian should never start to support Raspberries since once these people start to arrive here in the forum Armbian is finally dead

I don't think we should support RPis, mainly cause there's no interest from one of the maintainers in bringing the RPi up and running on on armbian and cause they still provide a more or less good OS for their board. I don't think that a lot of former Raspian users would arrive here, it would be more a 'Armbian user who runs Armbian on his *random supported SBC* will run Armbian on his RPi too'-group (to keep administration as easy as possible).  But that's a personal opinion, maybe Armbian would hostile takeover the RPi community muhahahaha [/world domination mode off].

10 hours ago, tkaiser said:

(but maybe the current approach to semi-support as much SBC as possible will already kill the project).

I suggest we should open a thread called 'Armbian support philosophy' where everyone can make a proposal what's good or wrong with the current approach in supporting new devices. It's not about specific boards like the 'board bring up proposals', it's a more general discussion about support a (new) board, what's needed etc. 

This discussion is spread in so many topics&threads that nobody can follow this discussion...  Will this improve things immediately? I don't think so, but it could (at least) help to keep other threads clean(er) from this discussion, which is IMHO still an improvement.  

 

 

 

 

[Tido]

13 hours ago, tkaiser said:

education' BS

What I like on TK's post was everything beside education. And if TK lives under a rock in that regard.. posting from 2014  gääääähn.  Ups already 2018 did I miss something ?

I am not going to write a novel like @chwe it is already there..

https://www.raspberrypi.org/education/

Code Club   https://www.raspberrypi.org/education/programmes/code-club/

PiCademy   https://www.raspberrypi.org/training/picademy/

 

[martinayotte]

3 hours ago, chwe said:

so feel free to correct my grammar in french

Ok !

3 hours ago, chwe said:

bonjour mes amis qui parler une langue terrible

should be :

3 hours ago, chwe said:

bonjour mes amis qui parlent une langue terrible

 

[tkaiser]

3 hours ago, chwe said:

I don't think we should support RPis, mainly cause there's no interest from one of the maintainers in bringing the RPi up and running on on armbian

 

Wrong. Armbian already runs on tens of thousands of those Raspberries but fortunately users don't know. The whole thing (Armbian in general today) is not about technical details but user expectations and developers' motivations (and some commercial background gaining that might start to destroy the project). Feel free to open up a new thread babbling about what Armbian should do or not (count of devices supported and quality of support -- @botfap outlined this project's problems recently perfectly). I already gave up.

[chwe]

Support != Run. In case they ask why *random feature* on their/your OMV builds do not run smoothly they should ask in the OMV forum. There's also no 'official' support for the BPi M2 Zero cause non of the maintainers wrote the patches for UBoot and kernel, people from Sinovoip didn't send a PR with their patches. Armbian runs on a bunch of TV-Boxes which aren't supported officially.

Is dropping support the way to go? I simply don't know cause I don't have enough experience in this field. That's where people like you, @TonyMac32 @zador.blood.stained or @Igor are much more experienced. I guess (and that's just a guess not even an opinion) dropping kernels would make things easier. 

 

e.g.:

• drop 3.4 BSP H3 kernel (EOL, too old for building stretch images, hardware accelerated decoding is somehow limited)
• drop support for A10 (it's only one board but we deliver 2 kernels)
• etc. 

But neither this thread here, the meltdown & spectre (where @botfap made good suggestions) nor the 'support over forum' thread is IMHO the right place to discuss this. 

 

[off topic]

3 hours ago, martinayotte said:

should be :

6 hours ago, chwe said:

bonjour mes amis qui parlent une langue terrible

you shouldn't improve my skills in making fun about french natives...  

[TonyMac32]

6 hours ago, chwe said:

I suggest we should open a thread called 'Armbian support philosophy'

 

I have a few things I'd like to "babble" about along those lines, however I think it would be best termed "philosophy of application".  In this case there are a few things that may help, such as board config templates (what kernel options/etc are common as possible across all Armbian kernels/etc). 

 

8 minutes ago, chwe said:

Is dropping support the way to go?

 

I would say "freezing" would be preferable to "dropping".  This project provides more than enough information to recompile kernels and add modules/make small changes/etc, there's no need for any of the devs to pull out a board with a cm of dust on it and risk starting an electrical fire when they plug it in and the capacitors have dried out [dramatization for effect] simply because one guy wants 'x' feature on 'y' board that hasn't been built in 'z' years.

 

Reading over @botfap's post I agree, other than the RPi bits, like with the Tinker Board we'd constantly get the "But the Vendor OS can *insert something a *team* of professionals implemented that you weren't able to do in the 3-4 hours you put into this on an average night when nothing interferes*. Raspbian does everything you should and shouldn't do with one, a similar reality is why I haven't pushed a "community support" package for the Khadas VIM.  All Meson GX/GXL (Amlogic S905(X)) boards are very nearly identical and can be supported very easily as a family, but Khadas already has a pile of fully capable OS images, I have no desire to "compete" with them by pushing a 66% or so capable Armbian branded image.  RPi is the same sort of situation.

 

We have a lack of metrics for what constitutes "ready" or "supported".  Now, some requirements are difficult to make "hard requirements"  But this goes back to the "matrix of what works" I proposed some time ago, if the hard requirements are met, and the board's aggregate support/capability is over let's say 75%, then it could be considered (arbitrary numbers).  I won't pretend to be an impressive developer, but I am in automotive engineering, and my life revolves around gate reviews and processes.

 

3 hours ago, tkaiser said:

Armbian already runs on tens of thousands of those Raspberries but fortunately users don't know.

 

Keep it secret, keep it safe. 

 

There are, in my mind, two questions that must be asked before supporting anything:

 

1. What does Armbian bring to the SBC and its users that the Vendor does/can not?
   • This needs to be in "dumb user" terms.  A board for watching cat videos is not an Armbian device.
2. What does the SBC, it's vendor, and its users bring to Armbian that would be a benefit to the community?
   • Will these users contribute?  Will they donate?  Will they spread the good word?
   • Will the vendor contribute resources and will those resources outweigh the support effort for the device?

In the case of the Pi, it is a cat video watching, Scratch programming, "ZOMG I AM 1337" type of board.  I think in terms of a beginner Linux doodling machine it's fine.  But it does not have any hardware capabilities that would benefit from Armbian in real terms.  I'm sure the OMV images perform much better than if they weren't Armbian, but how much better is that when the hardware is so bad?  I'm glad OMV and tkaiser were able to squeeze some water from the performance stone that is RPi, but to go that one extra step and have a general-purpose official Armbian-branded image would be stupid I think, the answer to the "Important Questions" would be as follows:

1.  We can make the Citroen 2CV go 5 km/hr faster, but it is still outrun in a 1978 Diesel Rabbit towing a trailer.
2. The users would bring us exposure (probably bad when we're not as "friendly" (OS or devs) as Raspbian) and they would undoubtedly require more resources in support than we would gain in donations/contributions to continue the work on other boards.  Botfap's situation is quite different, if funding were assured and the customer defined then support is quite different than with an "anybody can download it and complain" model of an open project like this.

[Igor]

6 hours ago, TonyMac32 said:

however I think it would be best termed "philosophy of application".  In this case there are a few things that may help, such as board config templates (what kernel options/etc are common as possible across all Armbian kernels/etc). 

3

 

7 hours ago, chwe said:

But neither this thread here, the meltdown & spectre (where @botfap made good suggestions) nor the 'support over forum' thread is IMHO the right place to discuss this. 

 

Perhaps two pinned topic, one for philosophy/rules and the other for devices which are about to lose a prime support place. In here: https://forum.armbian.com/forum/22-board-bring-up/ where support wannabe devices start and also stop its journey? Even we drop few boards, a bottleneck of manpower remain.

 

10 hours ago, tkaiser said:

and some commercial background gaining that might start to destroy the project

Cash, which comes in here and there, is a consequence of past work on the project. In the economy term, that would be fruits of past accumulation. Those are the rules of the (western) world we live in and are not related to specific activity nor wish. So far I said "no" to many advertising agencies aka easy money because I hate sharing spam and want to suppress the motive which you might be referring to. Not to mention saying few times "no" to providing commercial board support, not necessarily related to Armbian. I asked once around, while now I don't bother and deny if anything comes in - obviously, we all have little to no interest in such moonlighting.

But. If a board maker or 3rd party throw pennies without requesting specifics, I don't see any morally disputable questions.

 

Initially, they gave boards around and ask, currently hope for support. Giving away a free board sample is already in a grey area. Now, at the state of this project/market, samples are becoming less significant and a donation is as is. Perhaps a wish for some sense of appreciation, connection, control?

We need to define very well what support is and throw out more deprecated boards for our own good. I certainly agree on that and have some propositions.

We are not busy solo because of too many board support count - new boards actually added little problems - but overall project expansions, lack of resources, rules, organization troubles ... For resources that are n/a but needed cash is sometimes the only realistic option to make things happen.

Income is a must have to run such project, while profit is not and there is no danger of having one.

[botfap]

17 hours ago, TonyMac32 said:

There are, in my mind, two questions that must be asked before supporting anything:

1. What does Armbian bring to the SBC and its users that the Vendor does/can not?
   • This needs to be in "dumb user" terms.  A board for watching cat videos is not an Armbian device.
2. What does the SBC, it's vendor, and its users bring to Armbian that would be a benefit to the community?
   • Will these users contribute?  Will they donate?  Will they spread the good word?
   • Will the vendor contribute resources and will those resources outweigh the support effort for the device?

Not actual answers but:

 

70-80% of other Armbian users I have encountered (outside this forum) are only looking for 1 thing. Ubuntu desktop or server on their TV box or new SBC toy. Armbian just happens to be the easiest way for semi or non technical users. Usually they come looking for Armbian or something similar because of problems or limitations in the stock Ubuntu image offered by their ODM. I say Ubuntu because thats what most have them have heard of or run in a VM so thats what they ask for. Very few of them have even heard of Debian. These people are not interested in development or bug fixing or contributing. They will however spread the good (or bad) word depending on their experience, and that is very important for the long term viability of any project. While the individual opinion of anyone from this group is pretty much irrelevant to development process the collective opinion is valuable and should be listened too.

 

How do you keep these people out of developer discussions? New users dont get posting privileges to Developer support unless they contribute (code, docs, bug fix, cash donations, etc) and you expand peer to peer tech support.

 

1. What does Armbian bring to the SBC and its users that the Vendor does/can not?

2. What does the SBC, it's vendor, and its users bring to Armbian that would be a benefit to the community?

 

Im not sure what Armbian brings to a particular SBC is important. It may be, but from my commercial perspective SBC's come and go fast, SBC retailers come and go fast, users mostly have no loyalty to either, they just want an operating platform that consistent and usable across projects.

 

You mentioned the concept of an "Armbian Device" which I think is a great concept, hard to define right now, but the concept of an "Armbian Device" that just works with your project code, no changes, is a very powerful concept. In my experience thats what devs and aspiring devs are looking for. Thats why I suggested the standardised functionality in the meltdown/spectre post. That standardised functionality would also elevate Armbian's usefulness well above any stock distro or ODM image.

 

Users benefit a community. ALWAYS. Very few users follow specific SBC's, most of them are looking for an SBC to use with the operating platform they are familiar with. Only hardware tinkerers (which admittedly is the majority of the Armbian community) are interested in specific SBC's.

 

Also in general:

Most end users bring nothing of value to the table except testing, expecting end users to bring other value is unrealistic

Most board manufacturers bring very little value to the table, free boards are not a good reason to support a board

Users bring devs, devs dont bring users. This is critically important and I think overlooked by this community.

 

The bottom line: SBC manufacturers and boards come and go, the operating stack is way more important than the hardware. Armbian can be a much more important part of the ecosystem than any specific SBC stamper but in order to do that it has to define what it is. That may mean 2 different variants (its what we do in house), one to cater to the generic desktop / server type user and a variant specifically tailored to IoT single use devices.

 

This thread is not really the place but if @Igor, @tkaiser, @TonyMac32 wants to create a specific channel or thread to discuss this I would be happy to contribute in terms of ideas, infrastructure and code. Im also happy to contribute more servers for hosting and building and even funding if the development direction is clarified.

[TonyMac32]

56 minutes ago, botfap said:

How do you keep these people out of developer discussions? New users dont get posting privileges to Developer support unless they contribute (code, docs, bug fix, cash donations, etc) and you expand peer to peer tech support.

Somewhat in line with the "users bring devs" statement, this is almost always a hindrance to users and strikes a negative tone. I don't worry about "dumb repeated questions", honestly we don't truly suffer from that as badly as we may like to believe sometimes, but more of actual hardware issues they refuse to recognize.  Now, sometimes we've been known to take a tone that is extremely negative and discouraging to newer users. 

 

1 hour ago, botfap said:

Users bring devs, devs dont bring users.

I came here for just the reasons you listed, user of some H3 boards and, when I got the Tinker Board, decided to try my hand at contributing.  With some barriers to discussion/etc, I may not have felt it welcome or even appreciated to contribute, there is a prevailing attitude in the Linux world in general:  "If you aren't smart enough to already know, I'm not answering your questions."  If there's anything I think the RPi group has done, besides making cheap SBC's available in general terms, it's to reduce that ridiculous mindset somewhat.  If you put up an entry fee, for lack of a better word, I think the $8-30 SBC folks aren't going to be extremely interested, and they are the primary advertisement for the project.

 

1 hour ago, botfap said:

Also in general:

Most end users bring nothing of value to the table except testing, expecting end users to bring other value is unrealistic

Most board manufacturers bring very little value to the table, free boards are not a good reason to support a board

Users bring devs, devs dont bring users. This is critically important and I think overlooked by this community.

 

If users bring devs, (example as noted) then users can bring value, if you support boards that you know have a certain audience in mind.  A board with 4+ GbE ports, for example.  No beginner is going to grab that board to be a desktop, for example.  Targeting that board would typically mean getting the attention of more dev-biased users.  A knock-off RPi-0, on the other hand, will only bring people who were trying to get a zero but couldn't find one or thinks they need a quad core to measure the temperature of their basement over I2C.

 

I agree that most board mfg's see projects like this as either noise or free support.  On the other hand, others see it as a valuable way to get some feedback on hardware/software/etc, and it is work they don't necessarily need to do, so it is value-added for them to provide boards/donations/etc.  Some are far worse than others, and to be honest the products of the worst offenders rule them out of support long before their lack of contribution would.

 

1 hour ago, botfap said:

Im not sure what Armbian brings to a particular SBC is important.

I agree in abstract terms.  However, when going to the trouble of supporting something, it might be good to see if it's worthwhile.  The RPi, for instance, really doesn't need another OS except in very specific situations such as OMV.  Otherwise we're simply another image in a pile of images.  I honestly don't believe the demand for an Armbian RPi image is very high for that reason.  If a specific board's platform is rock solid and support is as simple as adding it's DT and kconfig to the pile, then certainly, add it. (The VIM will be in that category once the Amlogic gremlins get sorted, at present it seems the Amlogic documentation and drivers are simply a mess, so other than boards already supported I'd not recommend any new ones at present based on GX(L)(M) or the new AXG)

 

 

[chwe]

5 hours ago, botfap said:

They will however spread the good (or bad) word depending on their experience, and that is very important for the long term viability of any project. While the individual opinion of anyone from this group is pretty much irrelevant to development process the collective opinion is valuable and should be listened too.

Since kodi seems to be a major use case and hardware accelerated video decoding is often quite hard to accomplish, spreading 'the bad word' is something we have to deal with... 

Armbian is somehow a 'front-runner'/early adopters distribution. Pick up different boards & SoCs quite early so that others can benefit from. It's also a discussion forum where people talk about SoCs which aren't supported at all (e.g. I found nothing about H6 OPiOnePlus on Xunlongs [except product annotation] forum, the discussion how to get their stock android to run happens here...). 

 

For me armbian is somehow a base where others can build their nice (or shitty) house on it. Depending on the building materials armbian gets, this baseplate is excellent, good or ok-ish (depending on hardware, kernel and sometimes developers time/interest).  My favorite example: OPi Zero rev. 1.4

-Wifi is a nightmare

-Thermals is broken by design

-MicroUSB can end in a mess (when people don't follow recommendations)

So there are a lot reasons to drop support for this board, it's cheap (all the people buying by price, not by their needs will buy it). Do I think we should drop it? NO, it is still a good SBC if you know how to handle it but I think as soon as mainline reaches stable, we shouldn't deliver any legacy images for it. It has no HDMI so desktop isn't really a usecase, 256/512MB ram makes it anyway more a CLI than a Desktop SBC (to be clear, armbian doesn't provide Desktop images for it).

5 hours ago, botfap said:

How do you keep these people out of developer discussions? New users dont get posting privileges to Developer support unless they contribute (code, docs, bug fix, cash donations, etc) and you expand peer to peer tech support.

that was his first post here... I don't think keeping such people out from parts of the forum only cause 'he's new'. That was an amazing peace of work, nailing down a problem and prove his assumption.

Since mainline reaches 'testing phase' on H3/H5 it might be an idea to allow mainline questions about this boards in the 'technical support' subforum to keep the 'dev subforum' a bit cleaner could be an idea.

5 hours ago, botfap said:

Users bring devs, devs dont bring users.

You never read @tkaiser 'spreading the good word' outside armbians forum in various blogs etc. 

 

5 hours ago, botfap said:

This thread is not really the place but if @Igor, @tkaiser, @TonyMac32 wants to create a specific channel or thread to discuss this I would be happy to contribute in terms of ideas, infrastructure and code. Im also happy to contribute more servers for hosting and building and even funding if the development direction is clarified.

If there's no other opinon clearly against, I would split this thread on @tonymac32s post (I have a few things I'd like to "babble" about along those lines,...) in armbians dev subforum (I think that's the best place where it belongs to) calling it "let's babble about armbian" (can't be 100% serious the whole time... ) 

 

Maybe I find a better name for it tomorrow..  

 

[Aux]

Would the new RPi3 B + be more interesting for ARMBIAN now?

[chwe]

Why?

Same SoC as before, same limitations (more or less) & same problems.

They improved:

-Thermal behaviour

-Wifi

-Ethernet

-PMIC

 

There are other distributions which provide proper support for RPis.. 

[tkaiser]

11 minutes ago, chwe said:

same problems

 

In fact one more problem: Gigabit Ethernet, crappy network equipment and ignorance: https://forum.openmediavault.org/index.php/Thread/18991-New-approach-for-Raspberry-Pi-OMV-images/?postID=171017#post171017