Olimex Support Forum

OLinuXino Android / Linux boards and System On Modules => A20 => Topic started by: NK on September 04, 2014, 11:55:58 PM

Title: General Discussion: Olimex Vs Others (Raspberry Pi, Banana Pi, Hardkernel etc.)
Post by: NK on September 04, 2014, 11:55:58 PM
Don't get me wrong - I love Olimex products, they are cheap, feature-rich(HW wise) and tons of add-on modules at affordable cost for anyone to tinker with the ideas. But where Olimex as company, fall short is they are always onto something new and latest at the expense of their existing products. I would love to see them making the software stack improvements - not abandon them prematurely, or at the least make an honest effort helping the community improve the software base. I can imagine their proprieties always been developing the next thing, which is a never ending cycle.
To be fair, I do not like Raspberry Pi Foundation's priorities either, they are stuck with something that worked 3 years ago. But they have a strong software stack with excellent community support for their hardware. I would love to see a balance between Olimex mindset (fast-track HW development) with better support to the software stack. If that happens, in my opinion, Olimex products will be a big hit with small time hardware+software developers - not just for prototyping the ideas and than move-on with something else or build your own hardware. I would love to utilize Olimex SOMs or LIME products in my end-user products, if the software stack is a bit (a lot?) robust.
My main frustration is from the fact that there has been no improvements to Android sources for A20 since the very first release in early in the year (7 months ago). There is no proper build instructions for Android for A20-LIME products. I am mainly a software developer tinkering with some ideas. I finding it frustrating to discover things that are so obvious and should have been documented somewhere (product wiki page?)

I would love to hear what other community members think on this topic
Title: Re: General Discussion: Olimex Vs Others (Raspberry Pi, Banana Pi, Hardkernel etc.)
Post by: JohnS on September 05, 2014, 12:31:58 AM
To me Android itself seems a mess and too much work for each board maker, not helped by Allwinner just doing code dumps (if/when they bother).

So, it is what it is.  I can't see any way Olimex could do much unless they hire a load of software guys.  Is Android worth that?  I doubt it, but maybe it would be for you.  I suppose Olimex would have to increase the board prices to pay for the software people.

7 months is a short time, as I see it.  Nothing stops you getting your hands dirty and make a newer version that can be put on the wiki.

John
Title: Re: General Discussion: Olimex Vs Others (Raspberry Pi, Banana Pi, Hardkernel etc.)
Post by: dave-at-axon on September 05, 2014, 03:29:13 PM
Have to agree, Android can be messy if there are a number of boards to support. The same happened with the FriendlyArm boards I originally used. The Android source and information to build was really bad.

The Olimex Android wasn't any better but with perseverance I have managed to get it working and in at the same time learned a lot about how the Android Linux kernel and the Android Dalvik interact. :)

Now that I have a stable Android 4.2.2 running with the A20 and my custom LCD I have to say that it has been very stable for the last 6 months or so and developing on it has been a pleasure. I have 2 commercial designs based on it and just got my first order for 3 systems. Not a big order by any means but these are high end, high priced industrial type systems and Android allows me to develop different apps for them and update as easily as a download or inserting a flash disk.

I am keen though to move to Linux but lack of real online debugging has put that one hold for now.
Title: Re: General Discussion: Olimex Vs Others (Raspberry Pi, Banana Pi, Hardkernel etc.)
Post by: NK on September 06, 2014, 05:59:29 AM
Quote from: dave-at-axon on September 05, 2014, 03:29:13 PM
Have to agree, Android can be messy if there are a number of boards to support. The same happened with the FriendlyArm boards I originally used. The Android source and information to build was really bad.

The Olimex Android wasn't any better but with perseverance I have managed to get it working and in at the same time learned a lot about how the Android Linux kernel and the Android Dalvik interact. :)

Now that I have a stable Android 4.2.2 running with the A20 and my custom LCD I have to say that it has been very stable for the last 6 months or so and developing on it has been a pleasure. I have 2 commercial designs based on it and just got my first order for 3 systems. Not a big order by any means but these are high end, high priced industrial type systems and Android allows me to develop different apps for them and update as easily as a download or inserting a flash disk.

I am keen though to move to Linux but lack of real online debugging has put that one hold for now.

Dave, I have gone through all of your posts, blogs - they were really helpful in getting me started on the task. I really appreciate your efforts in proactively helping other members and sharing your knowledge. Like most people, my low level system (firmware, kernel, driver etc.) knowledge is limited and smallest things also sometimes becomes a major roadblocks. My intention is not to blame Olimex for that.

Congratulations on your commercial design win. It is a step in the right direction. I am sure with your perseverance and dedication you will sell many many more of them. Wish you all the very best.

Cheers,
:)