Posts by mmk

Quote from wrxtasy

Whilst I applaud someone taking the initiative here is the thing. Cheap AML S905 hardware is "good" enough for most.

Its the Firmware and Software development support that kills most S905 boxes. Simply put. Its virtually non existent for budget Android devices.

And then we come to the very vital aspect of GPL Open Source Software. That being the AMLogic Kernel and LibreELEC / Kodi Open Source Code.
Any exclusive code modifications to support commercial LibreELEC boxes, MUST be published for all to see and then everyone else gets to use and improve on your code.

There is not greater way to piss off a development community than infringing upon Open Source development code.

Where is your exclusivity then ?

You need something in a product that no one else has.
WeTek for example have Widevine Level 1 DRM and Netflix in an easy as, Dual boot box. ODROID release an Open multi purpose platform with high quality componentry with great developer support.

A good example of recent GPL infringement issues concern Xiaomi HERE

Display More

wrxtasy, thank you. Although we haven't started the project yet but we do feel good when someone does acknowledge our effort to make something better.

I am not familiar with GPL open source infringement but I will read it up. But if we do develop a hardware and 'if' we managed to work out a deal so that Libreelec supports our hardware, I don't see any reason why we won't want it to be open-source. Many minds makes great work with little effort.

After all, we don't plan to develop a hardware where we want to control the software behind it. We are a small team and we all have full time jobs, we want to make money but of course we want to do it with as little as effort as possible. And definitely not pissing off the development community along the way.

But it has been several days since this thread started, there hasn't been any strong indication that there are 'MAGIC' features or options that could make KODI even better. Sure, more ram, bigger NAND or faster Ethernet is great but that seems to be a trivial matter to most users.

Well, we could throw in an aesthetic value like full anodized aluminium body with bright color that will look great on the user's table top. That is easy to make but is that what user wants? Who knows. Honestly, that is why we need input from the forum user.

We want to make a great hardware where we try to cram in as much flexibility/features as possible. Even going beyond the media center boundary. Then grows the software support and update it over the web.

Guys, what about if we add a RF transceiver into the unit? Those RFM69HW transceiver are used commonly pairing with Arduino for wireless hobbyist. If we can get OpenHab automation software (Smart Home) to run as well, you guys can use it to control the smart appliances in your home with it. Will this make you guys open your wallet to buy it?

openHAB
[hr]

Quote from Pelican

Please give a link where I can buy that! I haven't find ANY S905X box with gigabit LAN...

Well, I think we will be changing the SOC to 905x for the sake of 10bit 4K encoding features. So, yes, Gigabit Ethernet feature as well.

Quote from xsi

It's not so much that the response is negative, it's more that you dont have any differentiators for the product your pitching. You wont be able to compete on price with the vast array of Chineese made android devices which already offer very good hardware specs for very little money.

- Amlogic S905/S905X
- 2gb RAM
- 16GB Flash
- 1 Gig Ethernet
- Dual band wifi

The above specs can frequently be found in budget China devices for well under 50 USD, the majority of these devices work very well with Kszaqs LibreElec builds once the driver and device trees have been sorted out. The majority of DIY tinkerers will take a punt on a budget device because the prices are so low.

Then above this you have the likes of Odroid C2 - development boards and Wetek, so where exactly are you pithching yuor product to sit ? Inbetween budget china boxes and the higher priced devices ? Maybe if you can get Widevine Level 1 certification and DRM approvals from Google/netflix etc in a low priced device people will be interested i.e: in budget device terrotiry.

Xiaomi do have a android box on the market at the minute that offers DRM capibility and Widevine level1 at a decent price, but anyne that follows Xiaomi will know that they make good hardware but theire software support is useless, thats why it's never recommended to buy a Xiaomi android phone....great hardware...rubiish software, very slow reaction to software updates, bug fixes just dont happen a lot of the time...

Well that is what is troubles me, I can't seems to find a combination of features that will entice people into buying our products. Good manufacturing quality with moderate support doesn't entice people.

All the products out there seems to be a subsets from the big pool features. We were hoping we could concatenate a list of features and designs from all the best mind here and develop it. There doesn't seems to be a clear winner for a product with 'XYZ' features.

There is no way we can be profitable if we try to 'manufacture' the old fashion way. That is the regular development, manufacturing, marketing and then to the mass customer.
Well of course, forking out a huge amount to get it started won't be very profitable, so we thought we could crowd fund it. Unlike some dishonest "inventors" in Kickstarter, we are not scammers. We 100% legit and we do have the mind and resources to build it. We want to build a product that people will use.

Right at this moment, we are trying to convince our self that it is a feasible business. If we spend USD10k and if we target USD20 profit a per unit, we would have to sell 500 units!
It seems a product in 'middle' territory between those cheap China boxes and high prices doesn't catch on well.
[hr]

Quote from vitalero

I like the idea but not sure how profitable it is
The only option is to hire kszaq & his team full time and to kill the open source development, to allow it only on YOUR device that way his builds will support only your devices.
The reason is because once you do finish your device, it will just be between the known Chinese devices, since the hardware is not that different, and the software support [LibreELEC] is also free, there wont be much of a difference, the disadvantage is that others like Beelink can buy the chips much cheaper due to their order being huge.

I think you need to aim to the crowd that is waiting for the ODROID-C3 or the RPI4[since the latest versions were released 10 months ago, people are definitely holding themselves back from buying them since they expect something new to pop out soon], something like the ODROID-XU4 with LibreELEC support at a cheaper price [Amlogic instead of Exynos5422]

OR
Just a personal fun project, to make just 1 device, and enjoy the free already developed support, with the right quality and price you might even sell few pieces like 100 tops

Well, I don't have an ODROID-C2 but I do have a RPI2 and RPI3 at home. I like it the RPI series but it always lacking some features. For example RPI3 lacks NAND storage and no bluetooth. RPI2 running Raspbian and trying to browse the web nearly made me throw my keyboard at the TV! ORDOID-C2 is great but it too doesn't have a WIFI. Buying an option for WiFi is pricey.

What I have in mind is a product that has all these features combined. I feel that a media/server hub that serves as a media center and acts as a future bridge to IoT is the way of the future.
[hr]

Quote from 2ami

does this site has any public proposal for hiring its developer, such as how much money one need to spend for supporting a single product for one year, lets say something like this
kszaq only > $1000 /year
escalade only > $2000 / year

team work > $5000 /year
not so much but its better thank coffee donation, isn't it
if it does this way, maybe all company or start up would consider ask them for help

If kszaq is up for hire for USD1000 a year, I would throw him another free flight ticket to Penang and fly him here for holiday, :).

Ned Scott, Very true, a very saturated market. Tough.

I thought I would give it a try to sell the product here in this forum and see the responds before we actually do it. The feedback has been very negative so far.
ODROID-C2 is a very nice piece of hardware.

koenkooi, Yes, it is not easy to build a product and I have dealt with RMA all my working years. It is crap load of pain. Which comes back to the million dollar question, is it worth it? We could be talking of an investment over USD10k or more.

It would make more sense to buy semi-working android box, load libreelec and fix the overheating issue and resell them. I think ROI portion makes more sense. But then I like challenges and I hate getting ripped off by China seller again and again. Why can't we progressively move forward and work towards a honest business model and sell products at a reasonable price that WORKS!

Roti canai is awesome, there are others as well, roti issue (roti canai turned into a spiral tower coated by condense milk) and roti murtabak (roti canai with beef!). Lots of food here.
[hr]
Another thing that I wanted to point out is that because there are SO MANY China factories out there. So many variances of S905 hardware available for example different WiFi chip that requires different drivers to get it working.

Then users will post request to kszaq and ask for fix and upgrades. This is so unproductive. If our project do kicks off, all the hardware component will listed out. No surprises, one build to rule them all.

Raybuntu, I admit that I use kszaq's release on my unit at home and I didn't contribute a cent to it and yet I paid for my Windows 10 Operating System license. I also paid my steam games too. Occasionally, I watch movies at the cinema as well.

That is not right. I know that. I am pretty sure kszaq has a family to feed too.
I have never developed a commercial software before let alone an Operating System but I am programmer. It is a lot of work.

If this project do kicks off, we have never said we won't contribute to kszaq's work. No matter how compatible our hardware is going to be, it is going to require some tweaking from kszaq. I could do like what others do and start a thread and beg kszaq to do it but we didn't. When we said we were starting on this project, we said we were going commercial. If we are paying our packaging vendor to put a logo on the box, we are paying for kszaq's work for every copy we sell.

You may not believe it but this was one of the intention of this thread. We need support. Well, if we can't even manage to convince you guys or kszaq, we have no business into manufacturing a Kodi box.

Well, I beg to differ on a quote you mentioned. While it is true that China's labor cost is generally lower than anywhere in the world. Penang is a very competitive place to build PCBAs. Some of the top contract manufacturer like Jabil, Plexus and Flextronics are based in Penang. We are not saying we can engage them for this project because this is a low volume manufacturing. But what we are trying to say is that, the cost of material procurement, SMT, case production, assembly and testing are within China's higher range.

So your statement that China can do it for half the price, I don't think so. I am from the manufacturing industry, fake parts, low quality solder paste, poor solder connection, reworks are the common activities on how China manufacturing plants cut cost. Which is why so many of us gets conned once or twice buying China made Android box. I am sure there are many.

Maybe kszaq has some ideas that he wants to see in a Kodi box as well. After all, this project won't even give birth if I haven't found Libreelec OS that supports my S905. We feel that, is project like ours that may help to keep kszaq going, getting paid for his hardwork. Wetek is based in Europe, so a lot of the profits are lost in currency exchange. People in Asia can't afford paying USD120.

Quality, reliability and building something that works well is what going to set us apart from those China boxes.

Well, Irusak. You have a good point. Wetek Hub is indeed a well established product with almost similar spec that I am building. And it has a very fancy remote! It looks great too!

But the price is the primary factor. Wetek Hub is selling for almost USD120.

We believe it is possible to provide an affordable solution that is half of that price and looks and works as great as it does. But we do it because it is exciting and it is an opportunity to add / remove features of existing products. I don't mean by reinventing the wheel but if we have a choice to redesign cars, I am sure everyone has some ideas to chip in.

Well, this thread is to explore that. Is the current product that is available in the market fulfills all the general demands or is there opportunities to go beyond that? What if we can add more features / processing capabilities or even GPIO like those that is available in Pi? Not only this product can serve as a media hub but also a platform for IoT products that we have been hearing.

The possibilities are endless.
[hr]

Quote from Jaaxx

Unless you plan on completely rewriting audio and video drivers for Amlogic I would run as far away from that platform as you can.

Jaxxx, well it doesn't have to be a S905 platform in the first place.

I have not look up the selling price of this S905 chip yet. But I think it is a low cost SOC. It serves the general requirement well. It does 4K easily. If those China factories can manufacture and sell them for around USD35 a set and still make profits. We think it is a pretty decent selection.

Also, we choose S905 because Libreelec supports it. We don't the capacity or capabilities to write drivers support for it. Only kszaq can do it!

mmpp, you are right, S905 doesn't support USB3.0. I think so is S905x. I have amended it.

That brings the question, how many USB ports does a regular Kodi user needs? One for keyboard/mouse. One for a USB drive. One for an external HDD. Is that enough?

Hi all,

[Anger]
I am pissed. I love Kodi, I think it is an amazing piece of software. In the past, I use a HTPC to run it but recently I brought an Android box and it drove me nuts.
All those China made boxes are either missing Bluetooth module, fake parts, overheating, too little RAM/SPACE, flimsy WIFI antenna and a shitty Android OS with Chinese text. I am a Chinese but I can't read much Mandarin.

[Introduction]
I am from Penang, the Silicon Valley of the East. Not many people know where that is but Intel and AMD has a gigantic manufacturing plant in our island. I am a Electronic Engineer (majoring in Computer). I worked in an electronic factory.

[Design in Texas | Built in Penang, Malaysia]
I lacked the brain power to design a PCBA but I where I can sub-con it to local Penang based electronic manufacturing factories to built it. From SMT to plastic injection moulding. My friend who is a super talented hardware/software Engineer in the States will design it. No crappy parts, if we say we use SANDISK NAND, we will tell you which Part Number. We will buy only from approved distribution. I hate it when people advertise S905 but gives you a S802.

[Why]
Why did I start this thread?
Well I need some inputs from you guys as to what makes the perfect Kodi box. I have some ideas but a balance is needed to keep cost down and eventually make it into a profitable project. Yes, I am thinking off starting a KickStarter project.

[Hardware]
I have not done enough research yet to confirm which component part number I will be using for these components but they have to be compatible with Linux/Libreelec OS.

• Amlogic S905 S905X (Support for 4k-10bit encoding)
• 1x USB 3.0 Port (S905 or S905X doesn't support USB3.0)
• 3x USB 2.0 Ports
• 1x HDMI 2.0 (4K @ 60p)
• 2GB DDR3 RAM
• 16GB NAND Flash
• Realtek Chipset 802.11ac WiFi Dual Band
• External WiFi Antenna
• Bluetooth 4.0 Module
• MicroSD Memory Card Reader
• 10/100/1000 LAN
• 5V DC Power Input Connector
• -
• No SPDIF, No SD Card Reader
• No AV Port
• -
• Anodized Aluminium Case (Imagine Bright Red/Blue/or even Pink!)
• Case to SOC For Heat Dissipation (Passive Cooling)
• On Off Switch!
• IR Bluetooth Remote Control (Off the Shelf Part, building remote is expensive)
• 5V DC Power Adaptor

[OS]
LibreElec.

I don't think we will add support for satellite TV as this is meant for Internet based home theater system.
If you know where I can find materials for S905 reference board design, please do share. It is easier than starting from ground zero.

I love to hear from you guys. Anything you guys can think off? Please do share your thoughts.
mmk