Hi to the team of six heroes who work at khadas
first, It surely was a very engaging decision to start this new company with a product which tries to find its specific public in an over crowded market : the market of chinese arm based SOC doing TV box and eventually competing with raspberry pi as a linux prototyping board for DIYers. I rushed to order my vim pro because I am a DIYer and because of the promise of linux support on it, I did and still do not care at all about running android on this board because I’ve already got much more than enough android devices .
I still do not understand how just another android TV box w/ some extra 40pin GPIO could made khadas become profitable while they also want to cover the linux DIyer’s market.
I wish them good luck because the DIYers like me are not satisfied with the software side of the product. Khadas cannot afford to spend the needed time to refining the linux software and tools and docs, the main result of this is deceived customers who will probably will not buy the vim2.
Hi to the team of six heroes who work at khadas
I think that releasing Vim2 is a rushed decision, especially when you have so much to do with one product, and as you have pointed out, you really are a small team.
This forums could really help you, but you need to fix all this stuff that we users need, that would be a great marketing, i am sure many of us would then recommend this device to anyone, to friends and on other forums.
If someone makes a search about Vim and looks here, he won’t find a reason to buy the device, he will see deceived buyers, Vim2 will then have less buyers than Vim1.
I understand that you have a lot of work to do outside Vim, but we would all like to have a fully working device, not Android box, because there are hundreds of these boxes, and i wanted to have a Linux mini pc.
I was promised an sd card Ubuntu image, and you still didn’t do that, that is a failure, i don’t understand that it is so hard to do, this is just one of these things you failed to deliver.
Nougat has a message “system ui has stopped” showing up every few seconds, and you didn’t even try to fix it, i have seen on yt that i am not the only one with it.
You could and should have ota update available, we need to burn new image with every update, it was a must, at least for Android builds.
There are a lot of reasons for us (customers) not to believe you, and unfortunately most of them are valid, but i hope you will soon prove us wrong.
I would sincerely like you to succeed, it would be great for all of us.
Khadas team does not write linux nor X11 nor wayland nor FB, let alone Gnome or KDE on its own. These are all open source projects. Support for some chips may depend on the chip vendor willingness to support open source or dedicated developers to write open source equivalent.
I am sure VIM has more support and developers than most other equivalents, end even if not so (I guess Raspberry and some others have even stronger community), all vendors can (not (yet)) take advantage of what software is (not (yet)) available. But I suspect VIM to pick up improvements equally fast (if not faster) than others.
As a buyer of the Vim1, I do not regret that I bought it. Having said that, I share some of the concerns Tommy21 has expressed. We need to understand that hardware, rightly or wrongly, is becoming a commodity these days, and the success of your product depends mostly on the software ecosystem and customer support. People are usually willing to pay a bit more money for a well supported product. Also keeping development alive for older products while designing new ones is a sign of long term strategy and respect for customer investment. There are many companies out there that always solve the problem with the next product, and the only solution is to discard the old one. Customers do not like that. It allows to make a quick buck, yes, but such companies do not get loyal customers in the long run.
I think Khadas has done a good job with the VIM hardware, yes, but guys you need to understand that you will be judged by the quality of the software running in the device. So while you need to dedicate resources to complete the VIM2, you also need, IMHO, to make sure your new builds will continue fixing bugs and bringing new features to the VIM1 as well.
Just my $0.02.
PS Hope the new VIM2 does not cost $131…
I have bought several ARM boards and usually Android works very well but traditional Linux desktop sucks on them. I agree that the vendor should at least provide a solid Android implementation. Most of Linux desktop technologies are developed on x86 and solutions that work very well on a $100/- Intel CPU become sluggish on $10/- ARM CPU. Expecting a small hardware vendor form China to fix this is unreasonable.
Regarding the VIM2 link on Gearbest, it’s only for their inside preview, not the true retail price, and the final retail prices will be at:
- VIM2 Basic: $79.99
- VIM2 Pro: $99.99
- VIM2 Max: $119.99
And the promotion price when VIM2 get launched will be lower, the Max model price may be $99.99.
Yes, we did agree that the software ecosystem and rightly technical supporting is the key for SBCs.
As VIM2 also specs with Amlogic SoC(S912) which is same architecture with VIM1(S905X), it means that most of the software developement will be common, for example:
- Ubuntu: both kerner Image, uInitrd, rootfs are same for VIMs and only dts and U-Boot are independent, we are trying to build a ROM to compatible both VIM1 and VIM2.
- Android: same Android SDK
- U-Boot: it’s also possible to develop one build to suit both VIMs.
It means that VIM2 will improve VIM1’s software and also technical support(we did know that importance that sw ecosystem , that’s why we didn’t follow on our original planing to develop Khadas Cross(RK3399) first).
Again, thanks for all the feedback from the community!
Thanks @Gouwa for clearing the the price of the vim2, I will definitely be getting a vim2 max was worrieid it would cost $161.00 as stated on GB site. Now I want to say I’m a proud owner of a Vim1 pro and in my opinion I feel khadas has done a reasonably decent job at supporting the vim1. We have several images for the vim and I see the khadas team is constantly improving them I agree that it would be nice to have hardware acceleration but compared to other SBC vendors like the makers of the Orange Pi line that doesn’t support there boards very well or even big corporations like Asus with there tinker board and Debian image that still to this day is still very buggy, I say the khadas team of 3 people have done a great job so far. I think what khadas needs is a bigger community take for example Pine most of the images available are maintained by the community now I still agree that the vendors of these SBC’s sould keep supporting there products. Well just wanted to say thanks to the khadas team for all your hard work and hope khadas keeps growing and also thanks for engaging with the community.
just send enough final examples to kernel developer like @narmstrong and Martin Blumenstingl and others and image builder for libreelec @kszaq and @balbes150 and to some testers who proved in past that they do test and report problems, and we could hope they improve Linux support.
I really don’t care about Android nor its development.
just my 2c.
Thans @Gouwa for your feedback on the strategy.
In the ARM world, there is a lot of pressure on the graphical part but none of the IP vendors (ARM MALI, ImgTech for PowerVR, …) understood the open-source strategy at all, and the price of these devices are so low they cannot generate enough money to put enough people to develop open source drivers and frameworks.
The situation on Amlogic is half bad, for the headless, Mainline linux will work like a charm on the VIM and VIM2, but if you need Video HW Acceleration and very good OpenGL ES support, this has a price !
Amlogic took years to develop their custom solution, but they went in the wrong direction by ignoring the standard Linux interfaces, and they missed the V4L2/OpenMAX road…
Anyway, there is still plans to add these missing pieces, and Amlogic recently published a Wayland compatible libMali that will work on the VIM.
But yes, it’s very frustrating to wait, but time is money and money is missing here.
I am very thankful for all work you and others did for Amlogic SoC.
But headless Server for me is something which works 100% reliable over Ethernet. And this is not the case currently with mainline kernel. As reported here in mailinglist and also and here in forum. Hope soon to have solution to this weird probem.
Keep up good work.
These are the reasons why I bought VIM1 Pro:
- It can run Kodi @4K
- It got 802.11ac WiFi which is essential for 4K
- It looks nice
I think Khadas really need to have a focus, maybe go for LibreELEC is a good idea. Make it a very good LibreELEC box, get supports from the community (VIM already has at least two members providing builds), keep it regularly updated.
For me, I don’t really like the Android on the box because It’s not the TV version. Even if it’s a TV version, it won’t have Chromecast function, also Kodi refresh rate would be a problem.
Then I really see these possibilities:
- A best looking LibreELEC box
- A decent Linux desktop
- A over powerful Linux headless server
For LibreELEC, the community can help. Please make everyone’s life easier by allowing multi-boot without UART cable.
For Linux desktop, maybe you could still focus on Ubuntu.
For headless server, I personally will just go for Armbian.
The VIM2 should not have any net issues since it will use an external PHY already used by numerous boards, but for VIM1, I have honestly no idea why it suddenly fails.
Most of our S905X board works like a charm, and we cannot reproduce the issue in a constant manner. And looking at the amlogic kernel, they did not add anything specific out of the PHY for the S905X support.
Anyway, you can try to look at what changed amlogic did on their 4.9 kernel and try to forwardport them to mainline, one by one, it’s painful but it’s how we added mainline support for all the other stuff !
Just to make sure you understand I do not include Khadas in the list of companies trying to make a quick buck. I bought the device after looking into the forum history, seeing how you guys were answering to queries, fixing bugs etc. You have been doing a very good job, and my comments were about my wishes to see the VIM1 fully supported in future releases, and you have answered my concerns, thank you. I will probably get a VIM2 basic (I’m a cheap b****d!)
Even the prices you listed are crazy for a S912 device. You want to compete with devices like the R-TV Box S10 that has the same hardware @ just over 60$?
according to khadas products positioning, their offer includes higher end software and dev support. Is that RTv brand offering also support like you can find in this forum?
You are confusing an empty TV box (unknown what support) and multifunction device (VIM2) having a much broader set of features and capabilities.
You can check Odroid C2 prices, it will cost more than 100$ with all the stuff that is needed to use it.
So i can’t say the price is too high, S912 is a better Soc, maybe there will be special deals that will cut down the price.
Only thing you can’t compare to Odroid is support, they are miles ahead of Khadas, but maybe that will change one day.
Maybe they have bigger community but not better support. They still don’t have everything in mainline kernel, and everything done in mainline kernel is done by @narmstrong (BayLibre) and others devs not directly from hardkernel.
Biggest issue here from my point of view ist Amlogic not pushing their stuff directly to mainline kernel and then also Khadas team with their custom stuff like internal PHY not pushing it to mainline kernel (see that confirmed ethernet problem).
I don’t know they have something new based on C2 product so I can’t comment on price.
When will the Khadas Cross? Competitors are not asleep, rk3399 too