Google released some weeks ago Android Auto.
http://www.android.com/auto/
An app on your Lollipop device that can connect to cars head units.
So using your phone with the biger display in your car, with apps specially designed for that.
The problem is that you need a new car or a after market head unit that supports Android Auto.
That sounds really expensive!!
BUT what if we could replace that expensive head unit with a raspberry pi?!
Have no idea yet if that would work, that's why I started this topic, for brain storming about this. [emoji1]
Would be nice to have a multimedia central with Waze running... Hope it gets popular fast
that would be awesome and i am totally into that.
as far as i know, some guy made an lolipop port for the pi2. maybe we can use that as a base.
jackcaos said:
that would be awesome and i am totally into that.
as far as i know, some guy made an lolipop port for the pi2. maybe we can use that as a base.
Click to expand...
Click to collapse
Try my Headunit app of course; see sig...
But it requires Android 4.1+, a decently fast H.264 decoder, at least at 800x480 resolution, and USB Host Mode support, or in future WiFi Direct.
mikereidis said:
Try my Headunit app of course; see sig...
But it requires Android 4.1+, a decently fast H.264 decoder, at least at 800x480 resolution, and USB Host Mode support, or in future WiFi Direct.
Click to expand...
Click to collapse
Hi Mike, my apologies for cold outreach. Interested in your technology, please PM me so we can talk.
Thank you
thundpa said:
Hi Mike, my apologies for cold outreach. Interested in your technology, please PM me so we can talk.
Thank you
Click to expand...
Click to collapse
Anyone can email me anytime at [email protected] .
Is this already built? Or are you just suggesting an idea? I'd love to have it in my car. Do post here if when it is ready. (I mean the PI version)
mathewparet said:
Is this already built? Or are you just suggesting an idea? I'd love to have it in my car. Do post here if when it is ready. (I mean the PI version)
Click to expand...
Click to collapse
Quote from OP:
Have no idea yet if that would work, that's why I started this topic, for brain storming about this.
Click to expand...
Click to collapse
I will be/have released my Headunit app as open source, so someone with the skills could do this.
But I likely won't be doing it myself.
If there is someone with skills to compile your head unit app would be great.
Where can we find the code for your app?
Sent from my Nexus 7 using Tapatalk
luci84tm said:
If there is someone with skills to compile your head unit app would be great.
Where can we find the code for your app?
Sent from my Nexus 7 using Tapatalk
Click to expand...
Click to collapse
https://github.com/mikereidis/headunit
I might re-consider and do it myself, but my time is limited. I could also just to try help someone to port it.
I would rather have it working well on Android alone than having it half working on multiple platforms.
Hi,
I started to work on it. but I started from scratch for 2 reasons :
1/ I want to understand how the protocol works, it's important because there is no documentation available.
2/ The dev approach is very different : it's mainly based on libusb only and I try to implement the AA protocol without dependency with the underlying transport.
I plan to open code but now it's just too early. Sharing knowledge is a good start.
The Android headunit source code is very helpful to understand the protocol. Thanks for sharing the code.
So far what is done :
-100% asynchronous processing of usb transport so the thing is fully event based. It seems that the code becomes less messy this way and asynchronous programming model fits well with libusb.
-switch to accessory mode works
-ssl handshake is ok so the init phase of the protocol is working
Now I just got my first encrypted message from the phone. Encryption is not a big deal. I think I will be able to capture the first h264 buffers end of this week.
I imagine RaspberryAuto (codename for now) could run on a very lightweight JEOS style Linux with readonly flash filesystem. I will take look at the Openelec project which leverages this kind of system. I never developed UI for RPI, that's the main pain for me.
As it is a 100% C program it won't need a lot of RAM so the cheapest Raspberry model could be enough. H264 is not a problem as RPI supports hardware decoding. I'm a little bit more worried about the cost of the SSL processing (I don't know if the PI processor is good enough for that.
So why a PI ? Simply because it's open and documented, it's fast for h264 decoding, it supports composite output so it can be connected to the factory head unit of a car. Another interesting point is the my code is working on Ubuntu Intel too so there will be probably more choices for the target device at the end of the story. But PI is just enough for this project.
marcjero said:
Hi,
I started to work on it. but I started from scratch for 2 reasons :
1/ I want to understand how the protocol works, it's important because there is no documentation available.
2/ The dev approach is very different : it's mainly based on libusb only and I try to implement the AA protocol without dependency with the underlying transport.
I plan to open code but now it's just too early. Sharing knowledge is a good start.
The Android headunit source code is very helpful to understand the protocol. Thanks for sharing the code.
So far what is done :
-100% asynchronous processing of usb transport so the thing is fully event based. It seems that the code becomes less messy this way and asynchronous programming model fits well with libusb.
-switch to accessory mode works
-ssl handshake is ok so the init phase of the protocol is working
Now I just got my first encrypted message from the phone. Encryption is not a big deal. I think I will be able to capture the first h264 buffers end of this week.
I imagine RaspberryAuto (codename for now) could run on a very lightweight JEOS style Linux with readonly flash filesystem. I will take look at the Openelec project which leverages this kind of system. I never developed UI for RPI, that's the main pain for me.
As it is a 100% C program it won't need a lot of RAM so the cheapest Raspberry model could be enough. H264 is not a problem as RPI supports hardware decoding. I'm a little bit more worried about the cost of the SSL processing (I don't know if the PI processor is good enough for that.
So why a PI ? Simply because it's open and documented, it's fast for h264 decoding, it supports composite output so it can be connected to the factory head unit of a car. Another interesting point is the my code is working on Ubuntu Intel too so there will be probably more choices for the target device at the end of the story. But PI is just enough for this project.
Click to expand...
Click to collapse
Great work, keep us updated I've just ordered a car with android auto but it would be great to have something affordable for my wife's car.
Juicie said:
Great work, keep us updated I've just ordered a car with android auto but it would be great to have something affordable for my wife's car.
Click to expand...
Click to collapse
Hello Juicie,
Yes my new car will come with AA support as well so I have lost some motivation to complete this work these last few days. I can share my code with anyone that would like to continue.
I will do my best to release an alpha version that will work on Ubuntu but there will be more work to do to use the h264 hardware decoding of PI.
wow! that sounds great!
This could get really interesting now that RPi Foundation have announced a touchscreen....
Hi,
I have good news. I have a prototype fully working on Ubuntu (screenshots attached). The rendering is very good and smooth. Nav is fully working.
Now we just need to port the code on the PI itself. I only used libraries available on the PI so it should be straightforward except for the rendering. For rendering I use gstreamer because gstreamer is portable and supports omx (hardware accelerated h264 rendering on the PI). I think I just need to set the PI omx decoder and sink into the gstreamer code to make it work. The tricky part could be to capture the touch events if the sink doesn't provide them itself.
Problem is I don't have a spare PI to complete the job.
hi,
That's really great news didn't expect such good progress!!
I do have a rpi and can test if you give me the app.
The problem at the moment is that I don't have an android device running lollipop..but I can try to find one
Sent from my GT-I9300 using Tapatalk
I'm going to follow this !
If you need someone to test feel free to ask !
Hi
I will try to build an image pour the PI asap then.But I will have to disable the touch events at beginning because the purpose is to port the code and use the omx decoder engine.. That's a first step.
marcjero said:
Hi
I will try to build an image pour the PI asap then.But I will have to disable the touch events at beginning because the purpose is to port the code and use the omx decoder engine.. That's a first step.
Click to expand...
Click to collapse
Where are you currently hosting your code?
Related
hello, this is my first post, and i wish that somebody know if this adapter graphics card:
http ://www .dealextreme. com/p/usb-2-0-to-vga-display-adapter-cable-for-extra-monitor-screen-34756
http ://www .pandawill. com/usb-to-vga-usbvga-display-adapter-for-monitor-screen_p38476.html
or this
http ://www .amazon. com/Plugable-UGA-125-Multiple-Monitors-1280x1024/dp/B003IE49T8/ref=sr_1_3?ie=UTF8&qid=1304438526&sr=8-3
works in a device with android and usb host.
will be nice if i could connect any cellphone to a lcd tv or connect a tablet to two monitors.
leomg said:
hello, this is my first post, and i wish that somebody know if this adapter graphics card:
http ://www .dealextreme. com/p/usb-2-0-to-vga-display-adapter-cable-for-extra-monitor-screen-34756
http ://www .pandawill. com/usb-to-vga-usbvga-display-adapter-for-monitor-screen_p38476.html
or this
http ://www .amazon. com/Plugable-UGA-125-Multiple-Monitors-1280x1024/dp/B003IE49T8/ref=sr_1_3?ie=UTF8&qid=1304438526&sr=8-3
works in a device with android and usb host.
will be nice if i could connect any cellphone to a lcd tv or connect a tablet to two monitors.
Click to expand...
Click to collapse
This depends solely on which phone you have.
Also try going to forum.xda-developers.com and scroll down to your particular phone and search thier general forum for it, mainly because chances are someone already has asked or talked about this
I don't think any Android phone will support USB to Video adapters. Some android phones have hdmi or video out capabilities. But those types of USB to Video adapters need special drivers which I don't believe are available for Android OS.
zarathustrax said:
I don't think any Android phone will support USB to Video adapters. Some android phones have hdmi or video out capabilities. But those types of USB to Video adapters need special drivers which I don't believe are available for Android OS.
Click to expand...
Click to collapse
I think that too, but there are drivers for linux and then android should make it work. I was just wondering if anyone tried.
here i found a list of the devices that are compatible with displaylink:
w w w.displaylink.com/shop/index.php?product=5
and in "download driver" are drivers for linux.
someone know if the driver for linux works for android ??
Pd: If a linux based router can: w w w.engadget.com/2009/05/18/displaylink-for-linux-turns-a-humble-wireless-router-into-a-beau/
why android cellphone or tablet cant.
also i found an app ($1,99) that allow to convert ipad screen into a new display for a computer through wireless, is there some app like that for android??
leomg said:
here i found a list of the devices that are compatible with displaylink:
w w w.displaylink.com/shop/index.php?product=5
and in "download driver" are drivers for linux.
someone know if the driver for linux works for android ??
Pd: If a linux based router can: w w w.engadget.com/2009/05/18/displaylink-for-linux-turns-a-humble-wireless-router-into-a-beau/
why android cellphone or tablet cant.
also i found an app ($1,99) that allow to convert ipad screen into a new display for a computer through wireless, is there some app like that for android??
Click to expand...
Click to collapse
Android runs on top of a modified linux kernel, but it's not a true linux distro, so it can't run linux apps or drivers. Something would have to be written special. Maybe if you can run ubuntu off your device, you may be able to do something. What device do you have that has USB host?
Also, if its not USB 2, you may not have enough speeds to transfer. I think sending video output through usb would take a lot of processing power, which it may not be worthwhile with a phone.
What device do you have and exactly what are you trying to do... If I knew what you're trying to do, I might be able to help. I'm good with stuff like that.
You could maybe just get a DLNA adapter and wireless send video to a monitor or TV, instead. I'm not sure what you are exactly trying to get accompliched. give more details.
I bought a SmartQ t10, that still doesnt arrive, I bought it for the screen and battery, then i saw this adapter, so im seeing if can get video out.
I found a mailing list where a guy with a motorola droid (who managed to switch to USB host mode) trying to do this.
w w w.mailinglistarchive.com/html/[email protected]/2010-07/msg00073.html
according to what I understood you should enable the auxiliary output video and change the driver to operate with mobile device processor.
here i found how to do it:
sites.google.com/site/voyageofbeagleboard/Home/displaylink-for-android
Code:
DisplayLink for android (latested version)
How to build displaylink driver for android
Displaylink is asic vender for USB-DISPLAY converter. It's is little slow, but expand your gadet example for android based equipment. So I modified the displaylink driver called udlfb for android.
How to build
- prepare android source code. I checked on beagleboard and android2.0
- extract and replace files with framework-core.tgz under framegework/base
- extract and replace files with kernel.tgz under kernel
- rebuild kernel and android
Description of tar archives
- framework-core.tgz
Import mouse and touch features support from android-x86 project . This is depend on android2.0 eclair
Touch features is following as:
Touch the right corner on and off touch feature.
[HOME] Touch the status bar
[MENU] Touch the status bar from left to right
[BACK] Touch status bat from right to left
- kernel.tgz
udlfb driver for android. Original udlfb driver source code is in latest linux driver/stagine/udlfb
Add double size frame buffer for pan_display
Add pan_display function
Make maximu resolution 1024x768. Original code support much more high resolution, but It cause crash under memory allocation.
Contents of tar archives
framework-core.tgz
==============================
Import usefule function from android-x86
core/java/android/view/RawInputEvent.java
include/ui/EventHub.h
libs/ui/EventHub.cpp
services/java/com/android/server/InputDevice.java
services/java/com/android/server/KeyInputQueue.java
services/java/com/android/server/WindowManagerService.java
services/java/com/android/server/status/StatusBarView.java
services/java/com/android/server/status/TouchFeature.java
kernel.tgz
==============================
touchscreeen driver depend my equipment.
drivers/input/touchscreen/Kconfig
drivers/input/touchscreen/usbtouchscreen.c
DisplayLink driver for android
drivers/video/Kconfig ->add configuration
drivers/video/Makefile ->add udlfb/Makefile
dotconfig -> .config example
drivers/video/udlfb/ -> driver
this will work??
If your phone does not have USB host mode you can modify your kernel to support it. Then it's a case of getting drivers, which may or may not be difficult.
Yes, i found that too, usb host for nexus one: sven.killig.de/android/N1/2.2/usb_host/
i guess that work for any device.
but I'm interested in that first run the driver already having a device with USB host, because is there more research about getting usb host or usb otg
Here is a android 1.6 driver from the opensource .
http://libdlo.freedesktop.org/wiki/
or
https://github.com/magoroku15/DisplayLink-driver-for-android
Any one know how to use it or update it to work on newer android ?
Gingerbread has very good compatibility mode so it "could" run there.
Sent from my Huge Device 2 using XDA App
zarathustrax said:
Android runs on top of a modified linux kernel, but it's not a true linux distro, so it can't run linux apps or drivers. Something would have to be written special. Maybe if you can run ubuntu off your device, you may be able to do something. What device do you have that has USB host?
Also, if its not USB 2, you may not have enough speeds to transfer. I think sending video output through usb would take a lot of processing power, which it may not be worthwhile with a phone.
What device do you have and exactly what are you trying to do... If I knew what you're trying to do, I might be able to help. I'm good with stuff like that.
You could maybe just get a DLNA adapter and wireless send video to a monitor or TV, instead. I'm not sure what you are exactly trying to get accompliched. give more details.
Click to expand...
Click to collapse
he simply wants to project his phone display on a lcd tv or a pc monitor.
many ppl with no hdmi output would love to stream by usb, it also allows to connect other usb devices which not supported with MHL.
usb 2.0 is fast enough to stream.
i have same problem
dsa
vga adapter
i was thinking about getting 7 of these to run an office, but they are hdmi.
http://www.ebay.com/itm/TBS2910-Mat...1085685686?pt=Desktop_PCs&hash=item4d163ea3b6
and i stumbled apon this anyone tried this yet? idk how it interacts with android to get the display, but it must work somehow lol
http://www.ebay.com/itm/MHL-Micro-U...Phone_PDA_Cables_Adapters&hash=item4ac7878c5f
<img src="http://i.ebayimg.com/00/s/NTAxWDQyMA==/z/vKYAAOxy4M5R9wYk/$(KGrHqEOKpwFHcVK2kc5BR9wYkDdGg~~60_3.JPG">
I've been researching the same kind of thing. I have an android head unit that I want to show everything I see on my roof screen but so far no luck. I have tried UBS to RCA and UBS to VGA adaptors but so far no luck. I even thought maybe something like android to android screen share like 'say' mobile-phone to android-tv-box and connect the tv-box in the car and share my Android head-unit to the tv-box but I can't find an app that works with doing that either.
Hello,
Currently i'm working on a personal project to integrate my GS3 into my car display and audio. Well this was easy with the help of a MHL or Allshare Cast Dongle.
My next goal was to install a 4 wire resistive touchscreen on my car display and use it as an external touchscreen control. I bought a 7" 4 wire resistive touchscreen with a USB module build around the CYPRESS CY7C63723C chip. I hooked up the touchscreen with OTG to test if there any initial reactions from my GS3, needless to say there wasn't any (no surprise there).
After a little digging around I came up with the conclusion that I need to create a custom "Input Device Configuration Files" for my touchscreen based on the following links
http://source.android.com/tech/input/touch-devices.html
http://source.android.com/tech/input/input-device-configuration-files.html
My initial question before I start plugging away with some codes is how would I name the files, the documents states "Input device configuration files are located by USB vendor, product (and optionally version) id or by input device name." How would I find the product ID or the input device name for the touchscreen? I'm sure this can be done thru terminal or something but I am new to android development so cut me a little slack here
Also if you know of any Input device configuration files dealing with touchscreen please post it.
Anyone?
hechen said:
Anyone?
Click to expand...
Click to collapse
You would probably be better served asking this on one of the more general hacking sub forums. Here, you're just going to get people who are familiar with the S3. There, you may find someone who has done something similar on another phone.
hechen said:
Anyone?
Click to expand...
Click to collapse
Any resolution with this?
I have been pondering the idea as well, but have no coding or hacking experience to go off of.
Would love to hear about the progress for this project and maybe a writeup of parts and procedures so far.
Hello XDA developers!
A few friends and I have been developing a totally new USB open source gamepad called XOPAD. As part of the open source community, we figured we might ask you what you think, possible improvements and suggestions that may arise, before launching it.
Here's what it looks like:
It has :
Stereo speakers embedded in the controller, one on each side
Dual clickable analog joysticks (not sliders, actual tilting joysticks)
An internal battery with a booster that can charge the phone
Software-programmable SENSE-pin resistance, to be able to switch from On-The-Go to AOA 2.0 modes (beta function as of now)
Upgradable firmware through the micro USB port
Uses HID protocol, therefore there's no need for apps, it's real plug&play. We might need to make an app to easily remap buttons in the future for different games.
Adjustable to fit different phone sizes up to the Samsung Galaxy Note 2.
UPDATE:
We have now launched our project! Show us some support and make XOPAD a reality!
XOPAD - an Open Source USB Gamepad for Android smartphones
Tell us what you think!
PS: You can follow the progress of the project in http://www.facebook.com/xopad, our blog http://xopad2012.wordpress.com or our twitter account https://twitter.com/xopad
UPDATE:
Here is the behind the scene video of the XOPAD:
this is very cool
Thank you enadzzz2010! We'll be uploading more videos as the days goes by.
This looks cool, thanks.
wow looks great..keep up the good work :good:
Thank you for your responses! Looking forwarded to get these in your hands soon! :fingers-crossed:
Looks interesting.
Updated main post with Making Of Video
I've updated the main post with a new Making Of video!
Looks very good. I've been waiting for something like this.
Sent from my HTC One S using xda app-developers app
sound awesome
Thank you guys for your responses. We are just a few days away from launching this on Kickstarter. I'll send out an update once this happens on the main post.
Update
We have launched out project on kickstarter! See first post for update.
I know this question will come with some confused comments and answers... so Ill ask the question and qualify my question with some examples why I am asking and of what I am not asking.
Question: What is the likelihood of getting/shoehorning Full x86 Windows 10 (or 8.1 until 10 releases) onto this phone?
Qualifying statements:
1. I am not asking about Windows Phone OS at all. Everyone knows Microsoft screwed the pooch during their shift from Windows Mobile 6.5 to Windows Phone 7 then again from Windows Phone 7 to Windows Phone 8. That is why so many of us jumped ship from Microsoft's Phone offerings over to Android in 2010. This is 5 years later and Microsoft might be able to get some market share back, but only if they pull their head out of their a$$....
2. (Example of irrelevant answer... Why do you want full Windows when there is Win RT or WinPho???lol!!!!LMFAO!!!! TrolllFACE!!!)
There are plenty of usage cases to justify full x86 Windows on a mobile device. Microsoft and Intel are pushing on the tablet market but for some reason they have not begun to crack the phone market with Full WinTel.... Simply put, If someone is willing to pick up a Windows 8.1 x86 tablet with 1GB or 2GB of RAM for $200 or $300 bucks then they should be ok picking up a similar device with 4GB of RAM and the ability to make calls.
3. (Another example of irrelevant answer...But Android!!!! It rulz!!!)
First off I am an avid android user. We started with My wife and I getting the EVO 4G in 2010, then EVO 3D in 2011 (I know...), then Note 2 in 2012, and My wife switching out the Note 2 for the Huawei Ascend Mate 2 last year. Im holding on to my Note 2 for the final stretch with its 9300 mAh Zero Lemon battery. All phones we've gotten have been rooted and customized by me. I use Bluestacks and/or Andy OS on all of my Windows PCs and have owned several Android Tablets. In short I prefer Android over Windows phone and iOS and whatever else is out there... Now in saying that, I feel a full Windows device in a phone with sufficient processor and RAM would be able to run Windows as its main OS and Android as an emulator to satisfy my Android needs.
Very well put, I'm also interested in this.
Sent from my MT2L03 using Tapatalk
Also interested in this!
I think this needs Microsoft's direct support. IMHO these are some blockades the community will be met with:
1. Since Android uses a boot.img as stage2, some work have to be done for the boot image to be able to chainload Windows. Vanilla installation goes out of the equation.
2. Figuring out how to chainload a proprietary OS properly is, well, hard.
3. It seems like the device uses some special Intel modem and wireless chipset. Porting won't be easy (Look at Intel PRO...)
4. Although 5.5in is GIGANTIC for a phone, maybe explorer or metro won't be able to fit in it?
5. Onboard storage is lackluster to host a full desktop Windows.
6. Windows doesn't even have a dialer.
But generally, this is a great idea! Being able to run x86 apps on a phone, oh the feels!
I'd be very interested if it would run full x86 or x64 Windows! However as stated, I doubt that will happen.
Even then it would be a bit limited and the main issue I would imagine to be space. The screen is quite small for a 1080p window to display on. I'd want to probably run a 720p res for larger buttons and such, might fit a bit better.
Perhaps if you could have it all run off a memory card, but then it would be rather slow to boot and cache stuff?
Here check out this small presentation. Something could be possible with the virtualization extensions that intel processor has.
This totally depends on :
- how well Asus releases the source code.
- Bootloader unlockable or not, i.e. ways to work around the secure boot.
I tried similar things on Lenovo K900 which is running Z2580. Lenovo's open source release was just horrible since even building the kernel was difficult due to lack of info.
I was able to eventually built the kernel with KVM enabled, but was having trouble signing the kernel for the bootloader.
Just force loading the kvm.ko was not successful either since the stock kernel had some feature missing required by KVM.
I would be interested to work on this phone again if we can form a group.
kazuken said:
Here check out this small presentation. Something could be possible with the virtualization extensions that intel processor has.
Click to expand...
Click to collapse
chinabull said:
This totally depends on :
- how well Asus releases the source code.
- Bootloader unlockable or not, i.e. ways to work around the secure boot.
I tried similar things on Lenovo K900 which is running Z2580. Lenovo's open source release was just horrible since even building the kernel was difficult due to lack of info.
I was able to eventually built the kernel with KVM enabled, but was having trouble signing the kernel for the bootloader.
Just force loading the kvm.ko was not successful either since the stock kernel had some feature missing required by KVM.
I would be interested to work on this phone again if we can form a group.
Click to expand...
Click to collapse
sure we can start a group on slack.com
some other things that also come to my mind:
wine for x86
docker for x86
debian chroot
wine already has some ARM support
This would likely be much easier... Notice the 'high end' system reqs?
http://www.ubuntu.com/tablet/operators-and-oems
I think the biggest problem for Windows would be wrestling with the PowerVR-based gpu.. Those stupid things are usually a roadblock in just about every interesting project..
kazuken said:
sure we can start a group on slack.com
some other things that also come to my mind:
wine for x86
docker for x86
debian chroot
wine already has some ARM support
Click to expand...
Click to collapse
Sorry for ressurecting this old thread but you would definitely be able to run Linux X86 on chroot. Then through wine you'd be able to run a lot of windows apps. Only issue is that performance would be low *unless* you'd output the GUI to android's framebuffer (FB0) which would require a kernel supporting this (outputting to framebuffer) which in turn would need Asus releasing the kernel sources so that to bake FB support.
So yeah it's all doable even with relatively good performance and by outputting the image (through MHL or chromecast) into the big screen would give us a X86 PC on the go. In fact I'd prefer it from running windows X86 natively because then you'd be losing calls and notifications... Imagine your *phone* running all your PC's software (well almost all as wine often has issues). You can buy one of those 128gb micro sds and your "phone" would have plenty of space for your (PC) data...
Stevethegreat said:
Sorry for ressurecting this old thread but you would definitely be able to run Linux X86 on chroot. Then through wine you'd be able to run a lot of windows apps. Only issue is that performance would be low *unless* you'd output the GUI to android's framebuffer (FB0) which would require a kernel supporting this (outputting to framebuffer) which in turn would need Asus releasing the kernel sources so that to bake FB support.
So yeah it's all doable even with relatively good performance and by outputting the image (through MHL or chromecast) into the big screen would give us a X86 PC on the go. In fact I'd prefer it from running windows X86 natively because then you'd be losing calls and notifications... Imagine your *phone* running all your PC's software (well almost all as wine often has issues). You can buy one of those 128gb micro sds and your "phone" would have plenty of space for your (PC) data...
Click to expand...
Click to collapse
I tried it out. you can get GNUroot and GNUroot wheezy x86 on play store. I was able to get fluxbox with tightvncserver running (though no apps, but was able to get an image in vnc) i am now going to try lxde and then see if nomachine 4.0 will work. wine should be able to run photoshop cs2.
kazuken said:
I tried it out. you can get GNUroot and GNUroot wheezy x86 on play store. I was able to get fluxbox with tightvncserver running (though no apps, but was able to get an image in vnc) i am now going to try lxde and then see if nomachine 4.0 will work. wine should be able to run photoshop cs2.
Click to expand...
Click to collapse
Problem with running your gui on a VNC server is that it is slow. It's (far) easier to setup though.
On my android machines I prefer to (basically) output an X Server window on the (machine's) frame buffer. You get real time performance (almost the same as if you had installed the OS natively), plus you get sound which is useful if one wants to run sound and video editing software (or plainly listen to music ). It's (quite) harder to setup but it has all been streamlined lately by a play store app named linuxdeploy (IIRC it has added X86 distros support lately).
Yeah... Don't use vnc, use xserver-xsdl . It's in the app store. Best Android X server. In your chroot, export DISPLAY=:0 after starting it up.
Sent from my ASUS_Z00AD using XDA Free mobile app
ycavan said:
Yeah... Don't use vnc, use xserver-xsdl . It's in the app store. Best Android X server. In your chroot, export DISPLAY=:0 after starting it up.
Sent from my ASUS_Z00AD using XDA Free mobile app
Click to expand...
Click to collapse
That's a great solution too! Hadn't thought to recommend it. It's easy to setup too.
Still outputting directly to framebuffer instead of an xserver app is quite faster (even than that!). But I'd expect the Xserver-XSDL performance to be quite good too.
OMG, this is SO interesting. I have been looking forward to put windows desktop in my phone since ages. Virtualization never let you go any further than Winxp. But now, this is another story. I am thinking of getting one of my own to help with the testing
Keep it up guys!
I ran photoshop cs2, via xserver xsdl, takes a while to load but eventually does, but its very hard to drag windows via xserver xsdl. i tried with vnc and was able to open a picture taken from the zenfone's camera and adjust levels. its alot easier to use a physical mouse and keyboard. but here are some screenshots of it running all on the android. it took brute force to create the x with the paintbrush and to drag a window. I ran it at 720p, also at 1080p. photoshop loads a lot quicker using xserver xsdl vs vnc.
You can change mouse settings when you start up xserver-xsdl. By default it's set up like the screen is a laptop touch pad.
The other thing you might want to try is a different Windows manager. I prefer fvwm2 since it's very light.
Sent from my ASUS_Z00AD using XDA Free mobile app
*irrelevant reply alert*
This takes me back to running Linux on the Windows-based XDA Exec. Those were the days.
Anyway, this is a great idea and you're finding some interesting workarounds, but I think you should be looking to get Windows to run natively. Sure, it doesn't have a dialler, but I'm sure someone can hook something up - especially if the interface is anything like the old Voice Modems from when we could only get our internets at 56kbps. (You kids don't know you're born! In my day, etc)
Meanwhile, in the Enterprise world where we try to reduce the costs of people having a whole processor each that they carry around with them, we're looking at using PCoIP to deliver a PC experience on a tablet. Sure, it's a little laggy (we're talking milliseconds on WiFi, though) but you get a lot of processing power, and if you're using Amazon you'll get NVidia rendering too. That's more for workstation graphics - CAD etc, rather than gaming. But then, if you're looking at installing Windows on a Phone, you're probably not going to be trying to play GTA5 on it.
Again, this reply is irrelevant because I realise you probably don't want to shell out $20-$40 per month on a virtual machine with a full Office suite. Plus, it's less fun to play with and not quite as much of an achievement to have set up something that works out of the box.
Native Linux 64 bit maybe, you get a much better OS, customizable, better resources management, open source, faster and waste less battery plus you can create your own mobile friendly interface just like Ubuntu Touch. Someone said it might be possible to port dialer, modem and other driver's concept since android is linux based. Microsoft is a handicapped development private code and as linux creator affirmed, its therefore a crappy OS lol There is steam on linux and it can run OpenGL games faster with the same hw due to uncluttered OS.
The hardest part will be GPU acceleration.
aziz07 said:
Native Linux 64 bit maybe, you get a much better OS, customizable, better resources management, open source, faster and waste less battery plus you can create your own mobile friendly interface just like Ubuntu Touch. Someone said it might be possible to port dialer, modem and other driver's concept since android is linux based. Microsoft is a handicapped development private code and as linux creator affirmed, its therefore a crappy OS lol There is steam on linux and it can run OpenGL games faster with the same hw due to uncluttered OS.
The hardest part will be GPU acceleration.
Click to expand...
Click to collapse
The .Net Framework is already Open Source. It's likely Windows 10 will go Open Source at some point. It's said to be the "last version of Windows" - probably similarly to the way MacOS X hasn't been replaced with MacOS XI. (There will still be a market for desktops when we have 128bit CPUs, and they won't just stick with the same 64bit kernel.)
Hi,
Sorry, I had already posted this on another section but I think this one is more appropriate and can't delete other post.
I have a Joying JY-NL124 android car radio, which does not have an accelerometer. I bought the Yocto 3d V2 http://www.yoctopuce.com/EN/products/usb-position-sensors/yocto-3d-v2 to see if I could get it to work with torque and other android apps for the track but although android sees the device the apps do not. I am not a programmer, but is there a way to get the apps to recognize an external accelerometer like some do for an external GPS?
Very interesting. I would like to know as well. It might also be usefull in navigation to get a faster direction orientation.
I was reading up on this again and this bit means someone with the know how would probably need to code something
The main advantage of this solution is that you don't need to install a driver to communicate with a Yoctopuce module, as the HID layer is always present. You only need to add our library, that we provide in source form, to your driving software for it to be able to directly talk to Yoctopuce modules.
Click to expand...
Click to collapse
Have you tried contacting them directly and asking if they can do this for you?
sinnedone said:
I was reading up on this again and this bit means someone with the know how would probably need to code something
Have you tried contacting them directly and asking if they can do this for you?
Click to expand...
Click to collapse
Uff, sorry for the late reply. Life has a way of getting in the way of interesting projects. I'm about to restart this project since I'm using the car a lot on the track. It's weird that the units don't come witha compass, gyroscope, accelerometer, etc, considering they are meant to be used in moving vehicles.
Definitely post up if you figure it out.
It would be nice to get it to work with Android as a whole. If that's the case you might need to talk to one of the ROM developers to see if it's something they can do ROM wise or even a custom kernel.(the kernel bit might be a little harder)
Did anybody get this right. It would be cool if these usb accelerometer could work in the apps
flash_xx said:
Hi,
Sorry, I had already posted this on another section but I think this one is more appropriate and can't delete other post.
I have a Joying JY-NL124 android car radio, which does not have an accelerometer. I bought the Yocto 3d V2 http://www.yoctopuce.com/EN/products/usb-position-sensors/yocto-3d-v2 to see if I could get it to work with torque and other android apps for the track but although android sees the device the apps do not. I am not a programmer, but is there a way to get the apps to recognize an external accelerometer like some do for an external GPS?
Click to expand...
Click to collapse
I'm very interested to understand whether this is possible as I am thinking about a similar setup. You tried the three Bluetooth tips in the roll up thread?
https://forum.xda-developers.com/an...roll-joying-2gb-sofia-mtcb-mtcd-tips-t3555249
"Bluetooth Tethering & BT Settings", "Difficult to pair BT devices" and "Modified stock bluetooth app to allow connection to all devices"
Have you tried external GPS for higher refresh rates or is the Head Unit gps refresh good enough for the track?
Bob
MX5DrIver said:
I'm very interested to understand whether this is possible as I am thinking about a similar setup. You tried the three Bluetooth tips in the roll up thread?
https://forum.xda-developers.com/an...roll-joying-2gb-sofia-mtcb-mtcd-tips-t3555249
"Bluetooth Tethering & BT Settings", "Difficult to pair BT devices" and "Modified stock bluetooth app to allow connection to all devices"
Have you tried external GPS for higher refresh rates or is the Head Unit gps refresh good enough for the track?
Bob
Click to expand...
Click to collapse
Sorry....USB not Bluetooth... Also discouraging. I was hoping USB would be an alternative to BT for track add-ons...
I too have an Android head unit and do track days, and have the exact same interest as you. I use Harry's Lap Timer. I have the Pumpkin AE0273B head unit. It too lacks the compass and accelerometer. I too found this to be a very unfortunate omission.
I have developed apps actually so I do know Android program to some degree (although I am sortof a hack, not a pro). I have apps on the Play Store (since I can't post links just search Play Store for developer JimRoal). I saw the Yocto stuff. When I get some time I will look into this some more.
Any update on this? Would also like to add a compass / accelerometer to my joying.
Also interested. But on a newer 10.0 unit. Something I am going to look into.
nFiniti said:
Also interested. But on a newer 10.0 unit. Something I am going to look into.
Click to expand...
Click to collapse
Sad that over 4 years after I first posted this we still don't have a unit with these features. I finally gave up and bought an AIM Solo 2 DL.
I'm looking for a newer radio though, since mine is older than this thread
Still waiting as well.
Looking forward for solution, we have also WIMOTION sensors in the market with sample android apk but I'm not a developser so, put things together is hard for me !
I think the hard job is try to mock the sensors in Android AS IF external accelerometer was factory embeeded (allowing apps to recognize it)
Nothing new about this project / idea ???
?? dead ???
mariodantas said:
?? dead ???
Click to expand...
Click to collapse
Once you've figured out how to modify the kernel, you should be good to go!
Another option could be sensor to serial over USB, via a compatible adapter, write an app and there you go. Of course it won't be standard Android though.
Drivers are here
https://elixir.bootlin.com/linux/v4.14.133/source/drivers/iio/imu/st_lsm6dsx
The problem is adding them to kernel !
You dont need to add them to the kernel. You will need to implement an Android HAL that can read the values of these sensors. Its not a trivial task but also not very hard if you have access to the source code