TP2 Robotic Controller/Maze Auto-Solver - In development... - Touch Pro2, Tilt 2 Windows Mobile General

I'm starting a project that will automatically solve a tilt-maze (Teeter, for instance) on the Touch Pro 2. Since WM programming is a bit new to me, I'm looking for suggestions and advice on how to deal with some of the development issues.
Here are some details of the intended design/process:
- Open the maze program and lay the phone down on a tilting table
- The program reports the ball position within the maze as X-Y coordinates over bluetooth serial port
- A microcontroller with a bluetooth modem receives the data and determines the desired ball direction using a state-machine/path-planning approach
- The microcontroller uses a gravity-rolling model of the ball dynamics to determine the required tilt angle to move the ball along the desired path
- The microcontroller commands two R/C servos to tilt the table in pitch and roll, tilting the phone and causing the ball to move
- The phone updates the controller with the new position and the loop repeats
Alternatively, the phone itself could compute the control law and simply send servo commands to the MCU. Either way, I love the idea (irony?) that the phone generates commands that move itself to solve a maze on itself.
Much of the hardware work is already done. I built an M&M dispenser maze last year that is a robotic implementation of Teeter/Labyrinth.
Search "Office Maze" on YouTube, video 'kUqRcLjFOUc'
We were/are working on getting computer vision to recognize the position of the M&M within the maze and have it automatically solve itself. For now, I like the idea of solving the maze autonomously on the phone.
I've made a basic app for the TP2 that moves a ball around the screen via button interface, but I am still struggling with access to the accelerometer data. I also have had no luck sending data over the bluetooth serial port.
I've read many useful articles and threads, but so far have not been able to implement these functions. Any suggestions on references or example code that describe how to do either?

Related

HTC Incredible doesn't have the same multitouch issues as Desire/N1 with video

Everything in the title;
http://www.engadget.com/2010/04/26/confirmed-droid-incredibles-multitouch-support-is-better-than/
Feeling like I should have waited .
What routes are there for us 24m contract holders to upgrade? Sell/rebuy or do HTC do some in house upgrade?
regards,
niklz
(not sure if this has been proven yet) but as far as I can make out I reckon it's only a software issue with multi-touch.
I downloaded the multi-touch test software from the android market and the screen was able to track two fingers independently until they crossed axis - this could be an issue with the actual software from the market.
The fact that it was able to track both fingers at the same time moving in different directions suggests to me that a fix could be possible.
We'll have to see what access root access brings to the desire.
any way the Desire looks way better than the Incredible, why change?
here's hoping froyo will fix this for n1/desire but it maybe a hardware fault
Personally I'm indifferent on the looks of both,
The Incredible uses a newer screen, which is what fixed this issue, not software atm.
The issue afaik is due to what information the screen records, I believe it records only x-y data for each input, but keeps no record of which input is which. So when you cross axis, there is some possibility of an incorrect allocation of the x-y data for each input as they move apart.
If HTC could optimize the data collection and analysis, I'm sure they could trace each input and tag it as such, or just reset the inputs after a cross of axes, such that it see's the inputs leave the screen for a short period of time and picks them back up again. All my poorly educated guesses though.
regards,
niklz
am i missing something here because i don't seem to have this problem on my desire using multi touch vis test app from the market. doing exactly what was shown on the video.
On android 2.2 spec list, it says something about fixing the multitouch issue???
Hopefully it will Be fixed!
It's a planned fix (For the Nexus one) in Android 2.2. Hopefully it will come down to the HTC Desire aswell as they use the same screen. All this will fix is the axis crossing. The screen will still only detect two fingers as that is the hardware maximum.

Windows Phone 7 owners, test an app?

Hey, i just got my dell venue pro today and was finally able to try out my app i created!
I added it to the market on the 15th of oct and it received poor reviews from the guys in the UK (not sure why).
Using it on the dell and it works great.
Anyway i was wondering if you guys could give it a try and give a review of what you thought.
My app is in the Health Fitness section and it is called My Workout.
Description:
It is a handy app that allows you to save your workout routines to an easy to access and easy to read schedule. My app also has a stopwatch and 1 rep max calculator.
It doesn't keep track of workouts, nor does it have any pre-loaded. This was created to allow gym savvy people (anyone who's worked out before) to digitally have their workout routine with them at the gym (i use my phone for music) to replace paper.
Video of app:
http://www.youtube.com/watch?v=sVoqy0xYmNg
* BUGS/ FEATURES NOTICED SINCE USING ON REAL DEVICE***
1. The lock screen wipes out any in progress routine that you were entering. Same with stopwatch.. If you are running the stopwatch and the screen locks, the stopwatch will reset. ( will fix with new lock screen guidelines that were just released)
2. I need to make the return key allow for the exercise to be added to your list instead of always hitting the add button.
Discovered that new users to windows phone 7 are oblivious to pivot controls and they love hitting return to enter data. WILL fix the return bug, the pivot control is essential to the metro ui. Cant change that.
* End bugs/features**
If you guys find anything let me know!
Thanks,
Blake
do not mean to hijack this thread but does the venue pro have a sd card slot? if it does it will be the phone i get.
I have nothing to test it on, but I do think the timer should have milliseconds as well.
the dell venue pro does NOT have a sd card slot.
Nice app, wish I had it when I was in the gym.
It's simple and straight to the point. Good to keep up with your routines.
Micro I just tested it in my HD7. Very nice app. I don't go to the Gym. (Too many posers there). Instead I ride my mountain bike and go swimming. Still your app will come in handy to many.
I've just written a brief review on Marketplace and given it 5 stars- user name jfsfoster.
Looks solid, I have two issues, but they're just aesthetic. The flexed arm doesn't really fit in too well with the default icons. I know, silly gripe, but it caught my eye. I'd try something more low key. Such as a round weight.
Remove the white border around Start/Stop and Reset and remove the speedometer graphic. It will fit better with the WP7 look.
These are just my opinions. The actual usefulness of it looks great and I will be downloading it once I find a Venue Pro.
After having friends use the app for the first time (first time wp7 users as well)
The most common problems were:
1. Unable to figure out that the navigation is pivot.
2. Why hitting return doesn't add exercises to the list.
I cant do anything about #1
However i have made some significant changes to how the textbox entry is maintained and have fixed some major bugs:
Disabled the lock screen from enabling via phone idle (microsoft wont like that i bet..) - Had to do it, if you are half way entering data and you set the phone down for 30 seconds it locks and then guess what, NO MORE DATA, you have to start over if you haven't finished the routine.
Allowed for the return (enter) key to add exercises to the list.
fixed misc bugs.
Version 1.4 look for it in about 2 days.
what do you mean navigation is pivot? as in, how they have to slide left/right? people just have to get use to that with the phone, it's not your fault. if they owned the phone, after a couple of days use they would work it out and it would be natural.
what do you mean no more data? i thought the new guidelines said the app could stay alive even if the screen goes to stand by? can you write to isolated storage upon screen lock?
Yes, you are right. People do need to get used to that but what's happening is that they download the app, attempt to use it for 2 min and then rate it 1 star and uninstall. I've noticed many apps getting poor ratings and they seem perfectly fine to me.
You are allowed to run in the background, however you cant have any timers running as per power saving guidelines. So that would mean that my stopwatch would have to be disabled.
I'd rather just disable the phone idle while using my app..
Also as for the data erasing, I write to isolated storage but not until the routine is at a specific point in the process. If you are say entering exercises and you have to go do something and set the phone down and it auto screen locks, the app re-sets and all that data you were entering gets cleared since it didnt get to the isolated storage save point yet.
that is fixed, it's just up to microsoft to see if it passes testing now.
Will download and test it, but can't report back for a few weeks. Can't go to the gym this week and next I'm out of the country.
App updated 11/10/2010
Added the ability for users to import exercises from previously entered routines. This saves time for those who do the same routine throughout the week.

[Q] ts_calibration HTC Touch pro Xandroid

Hi everyone.
First time on a forum for years. This one is pretty daunting though. Not even sure if Im in the correct thread
Just recently got myself an HTC Touch Pro. Yeah! it's old but I guess its better that the Ericsson w350 I had.
I was not happy with the 6.1 Win Mobile OS on the HTC and I heard that Android can be loaded up instead. So I started doing some research.
To cut a long story short, I managed to load XDAndroid onto the microSD card. all is working fine except I messed up on the touch screen calibration. or at least it was not made clear to me that what I was indeed doing was a screen calibration.
After some reading on the wiki files, I removed the ts_calibration as instructed and rebooted the system. The graphic for the calibration test does not seem to display correctly. Hence the readon for the initial stuff-up so I simply had to guess where the pointer were using the win6.1 OS calibration(Don't ask - it involved a non-permanent marker). Even that does not seem to work.
It would seem that the co-ords where transposed after some fiddling around. Left was right, up was down etc. So I got to thinking to open that ts_calibration file and started playing around with the numbers. Even the screen functionality was reversed when the I slid the screen open. Navigation via the Keyboard is no hassle though. The touch button navigation seems to work fine except the selection key does not actually select anything, I have to use the enter key on the keyboard. Not sure if this is normal. So if anything, navigation involves the keyboard to be open thus the screen orientation is always landscape as opposed to portrait view which sucks! esp if I'm texting or making a call. What the work around for this is general?
So I thought of opening the ts file in a txt editor and I transposed the numbers. I seem to be getting somewhere as the up and down seem to be working correctly now but the left and right are not, center screen is still out though. I'm not sure what sets of number represent which co-ord. I know that the calibration requires 5 co-ords, but I dont know in which order they should be. Is middle, top-right, top-left, bottom left, bottom right?
It would be easier to just have the correct co-ords for this phone. instead of playing around with some random numbers. IOW would the ts file co-ords be the same for all HTC Touch Pros? if so where can I get the correct co-ord numbers for that file? then I wont have to use the calibration function. easier still, is there a free app that will do this from within the XDAndroid os?
The start up txt I chose was Raph100 for the root folder. This is correct for my phone according to the wiki help files. I've not updated the files and used them as is. Would the incremental updates fix this?
I'm having the same issue for my TP. After some discussion, I downloaded the ts file from the FRX07 fixed discussion and it's pretty darned close to accurate. MAYBE just a little off center to the lower left of icons.
Here's a utube vid that shows clearly the touch pattern for calibration.
http://www.youtube.com/watch?v=FB6MVFH4Jv4
Hope it helps.

JY-UL128N2 now available

I emailed Joying again yesterday to ask about UL128N2, and they responded by adding the unit to their webpages and accepting orders.
Canada website (Canadian $):
http://www.joying.ca/best-joying-ne...t-pip-sleep-model-bluetooth-4-0-dab-obd2.html
Global website (US $):
https://www.carjoying.com/new-andro...h-4-0-gps-navigation-professional-head-u.html
and the website is down. As usual.
If the website isn't down, it's their forum. Sigh.
morgish said:
and the website is down. As usual.
If the website isn't down, it's their forum. Sigh.
Click to expand...
Click to collapse
Well its back up now. Or at least its working for me.
Sometimes international routing through the great firewall can be a bit.... spotty.
Has anyone bought already?
Is it better than the ones with bezel and volume knob?
KamaL said:
Has anyone bought already?
Is it better than the ones with bezel and volume knob?
Click to expand...
Click to collapse
My two units arrived. Played with one of them on the kitchen table late last night.
The row of capacitive buttons are thoroughly pointless -- touches to them don't register at all if your finger doesn't touch it absolutely perfectly.
Having said that, I like them better than the one with the knobs, for the reason I've already mentioned to you -- the lack of deep bezels.
Better? It is definitely a matter of preference. I don't need the capacitive button row, so that doesn't bother me at all.
Note: they have this weird dot thing stuck in the UI as configured by default. Touching the dot brings up a set of buttons -- back, home, menu, etc. So.... CHINA. Obviously, unless you have AT LEAST three copies (SystemUI header bar, row of buttons, the weird dot thing, steering wheel controls) of every button, somebody is going to miss it. Fortunately, there is a control application in the application menu that allows you to disable the stupid dot.
doitright said:
My two units arrived. Played with one of them on the kitchen table late last night.
Click to expand...
Click to collapse
I am sitting on the fence with this purchase, any-chance you could do a video review with these units?
doitright said:
My two units arrived. Played with one of them on the kitchen table late last night.
The row of capacitive buttons are thoroughly pointless -- touches to them don't register at all if your finger doesn't touch it absolutely perfectly.
Having said that, I like them better than the one with the knobs, for the reason I've already mentioned to you -- the lack of deep bezels.
Better? It is definitely a matter of preference. I don't need the capacitive button row, so that doesn't bother me at all.
Note: they have this weird dot thing stuck in the UI as configured by default. Touching the dot brings up a set of buttons -- back, home, menu, etc. So.... CHINA. Obviously, unless you have AT LEAST three copies (SystemUI header bar, row of buttons, the weird dot thing, steering wheel controls) of every button, somebody is going to miss it. Fortunately, there is a control application in the application menu that allows you to disable the stupid dot.
Click to expand...
Click to collapse
That dot was terrible. I also disabled it immediately, thankfully they gave an easy option to turn it off, unlike many of the other things that are obfuscated(such as disclosing to owners the hardware keycombo to boot into recovery).
Upon receiving my unit the capacitive buttons weren't always responding even when pressed perfectly, but after holding down RST till reboot they haven't had the issue again.
When inquiring about a larger touch area for the capacitive buttons I got "Hi friend, Nice day Yes, you can choose the bigger or small touch area in below settings:" with this picture attached.
Increasing the delta values in Settings of the Key Study app doesn't seem to do a whole lot but I think the trigger area is a bit bigger. Unfortunately the border values are for the entire touch screen not per button area.
tally3tally said:
I am sitting on the fence with this purchase, any-chance you could do a video review with these units?
Click to expand...
Click to collapse
No. I don't do video.
rcll said:
That dot was terrible. I also disabled it immediately, thankfully they gave an easy option to turn it off, unlike many of the other things that are obfuscated(such as disclosing to owners the hardware keycombo to boot into recovery).
Upon receiving my unit the capacitive buttons weren't always responding even when pressed perfectly, but after holding down RST till reboot they haven't had the issue again.
When inquiring about a larger touch area for the capacitive buttons I got "Hi friend, Nice day Yes, you can choose the bigger or small touch area in below settings:" with this picture attached.
Increasing the delta values in Settings of the Key Study app doesn't seem to do a whole lot but I think the trigger area is a bit bigger. Unfortunately the border values are for the entire touch screen not per button area.
Click to expand...
Click to collapse
That one looks like the buttons (and their corresponding areas) are considerably larger than with the 7". They are surprisingly small on this! I might see if I can disable them altogether, since I suspect that it would likely interfere with "slide finger in from left" gestures.
Ok, so got the thing installed last night.
Well, sortof.
Seems that China can't make up its mind regarding how far the front of the radio sticks forward relative to the mounting holes. So my first attempt to mount it resulted in the radio's face sticking so far forward that my DDIN face plate wouldn't have been able to make it down to the dashboard.... assuming that the holes fit. More on that later.
Second attempt, I moved the radio back to the next set of holes on the mounting bracket, which are a bit more than 1/2" behind the holes that I had the UL135 mounted with. Ok, well depth is *better*, but now a bit too far back. Middle of the night, so as long as I can use the thing to route my way to work in the morning, I'm ok with that. Maybe on the weekend, I'll drill another set of holes about half way between the two sets I tried so far. That should line it up OK.
So the DDIN sizing. Another thing that China can't make up its mind on. The UL135 fit in the Toyota Tacoma Metra dash kit... very snugly in horizontal direction, roughly 1mm clearance in the vertical direction. This UL128N2 is about 2-3 mm too big to fit in that dash kit in BOTH directions. They *claimed* the UL135 to be 100mm x 178mm. They *claim* the UL128N2 to be 102mm x 178mm. I will tell you that the UL128N2 is WIDER than the UL135, despite the claimed equal width. And given the 2mm (actually more) increase in height and that I had at least 1mm of available vertical space, I *should* have been able to force it in. NOPE. The difference is much greater than that, and in both directions. If the UL135 really is 100mm x 178mm, then the UL128N2 is probably 104mm x 180mm. WHY? Oh, the reason is absolutely HILARIOUS, it is because they have a thin plastic ring around the glass digitizer. They ordered the GLASS to have dimensions matching the DDIN target, but didn't compensate for the thin (~1mm) plastic ring around it! Very typical of the way that China manufacturing works. Ok, so I was able to grab a file and open up the dash kit a little bit to make it fit, and now its ok.
So next up... actual functionality.
First observation; the AM radio frequency standard is set to something that is most decidedly NOT North American standard. The closest frequency I could get to my preferred station was off by 2 kHz. While it was adequate to get the station and understand, without TYPICALLY an overabundance of noise or added sound from the next station over, it definitely sounds bad. Thanks so much China, for giving me a well working AM radio tuner! With any luck, I'll be able to hack the program to tune the proper frequencies.
On the physical utility of the unit, when you look past the "size" issue, its not bad. As expected, the screen is much more convenient without the deep bezel around it as the UL135 had. In order to achieve this, they replaced the physical control buttons with a row of capacitive buttons. Those buttons are pretty well useless, you can't reliably press them without focusing carefully on them. That is what is often called "distracted driving". I would go so far as to say that this row of buttons is DANGEROUS to try to use. Not only that, but the row of buttons is ***COMPLETELY REDUNDANT***. After all, they are just duplicates of the home/back/volume buttons that are already present on the SystemUI upper panel anyway, and, of course, the STEERING WHEEL BUTTONS (for those who have them hooked up). Why did they even bother to include that row of buttons? Because.... CHINA! Oh well.
Next up; there are some serious bugs when it comes to AUDIO ROUTING. If the radio is playing, and GPS voice comes on; radio mutes. Oh, it STAYS muted until you go BACK into the radio application, which seems to trigger a reset of the audio routing. This improved once I set google maps as the default nav application, now it just volumes down the radio to "pretend" to play the GPS sounds, and volumes it back up after.
Well, starting with the "included" ancient version of google maps... while it does make sound, it gets clipped by about 3 seconds at the start of the sound. So you miss about the first 3 seconds of the instruction. Update google maps to the latest version, and no sound comes out at all. Ok, what is happening with the newer one is that it is trying to use an audio pathway that is obviously BROKEN on these units. So that means that there isn't going to be any voice nav from google maps until either (a) THEY fix the audio routing problems, or (b) WE fix the audio routing problems. I'm not sure yet if there is an audio routing policy file that we can configure, or if this is something hard baked in somewhere. That's very annoying, but something I can live with for now. I prefer the newer google maps enough to live with the lack of voice guidance, rather than to try to use the ancient version.
DPI: Obviously, Android believes that the screen is as large as a TELEVISION. The fonts are all incredibly small, even if you set it to HUGE font. Ok, so right angle triangle has H = root(1024*1024+600*600)=1186.83 pseudopixels. Divide that by 7, makes 170 DPI. Its set to 160 DPI, so Android thinks that the screen is 7.4 inches. Ok, that's not really too far off from reality, but it IS in the wrong direction, to make everything look SMALLER than it should be. The problem is that the screen is roughly 3 feet away from your face. Take a tablet and how far do you hold it away? Roughly 1.5 foot. So lets figure out how big Android should *THINK* the screen is, in order to blow things up to the same scale as a 7" tablet at 1.5 foot.
We're dealing with right triangles again, 7:3 = x:1.5, x = 3.5. Android should think that the screen is 3.5 inches. That puts it up at 338 DPI. Ok, so these numbers are pretty rough, since I'm just guessing the viewing distances, but you get the idea. Crank that DPI WAAAAAAAY UP! Only problem I see, however, is regarding how they coded the SystemUI. If they followed the Android design guidelines, then they coded it with DP's (density independent pixels), which means that when you crank the DPI far enough to make applications look proper, it will make the CHINA look HUGE. Think about it -- they designed their SystemUI panels to show at a "normal" size when the DPI is set way too low. I can see bad things happening.
I wonder if it would be practical to just SWAP OUT the chinese SystemUI for one built from source? That might actually be a practical way of repairing the UI. Might also be nice to rebuild the System Settings application from source. I'll have to go over it again, but I don't recall there being anything special about System Settings -- they just colorized it, broke the bluetooth part, and added a password to developer options. WHY???!?!
Like the American proverb says... "if it ain't broke, don't fix it!" That ESPECIALLY goes for China, since if it ain't broke, they're GOING to break it.
And now, to BLUETOOTH.
Well, knew going into it that they futzed up the bluetooth settings as badly as they possibly could. Why? Because... CHINA. The good news is that they tried so hard to break it as badly as they could, that it was EASY to fix.
In fact, I had written up the most important fix WELL before I even had these head units. I will refer to my fix thread HERE: http://forum.xda-developers.com/and...elopment/bluetooth-settings-launcher-t3504526
There are two obvious problems with bluetooth on these units;
1) They deleted the System Settings button that brought up the Bluetooth settings. That is what I fixed in above thread.
2) They BROKE the bluetooth pairing pin prompt. On a normal Android device, when you try to pair with something from it, it will first try one or two "standard" pairing pins, and they don't work, then it will pop up a dialog requesting that you enter the device's pin. This one does NOT do that. It JUST tries to pair using whatever pin is hardcoded in the CHINESE bluetooth application. So while I haven't "fixed" this, there is at least a reasonably easy (if not particularly convenient) workaround. When it doesn't pair, you go into the chinese bluetooth application, change the pin there, and try again.
So the good news is this; the REAL MEAT of the bluetooth stack, is the product of INTEL. Not China. That means that once you work around the severely broken UI (again, WHY BREAK it when you can just leave it be and have it ACTUALLY WORK?), it actually works WONDERFULLY! That means you can connect to AND USE multiple devices at the same time. I.e., torque + **ANY** OBD device, while simultaneously a voice call AND bluetooth data tethering.
Hmm, what else....
Bootup.... this is a very dramatic change from the UL135.
A full bootup takes about 20 seconds.
So when you get into your car in the morning to go to work, after it has been sitting all day, you turn the key, and it is IMMEDIATELY UP. Wait, what? Its wired correctly -- if the +12 and IGN were reversed, then it would lose power with the ignition switch off, which means that it would have to do a full bootup. If the IGN wasn't connected at all, then the screen will never turn on -- when the IGN turns off, the screen goes off, radio stops playing, and another thing I observe, is that bluetooth connections are lost. So, either the thing is only doing a partial shutdown and proceeding to "drain the battery" mode, or they have actually made it sleep *correctly*, switch off all the parts that suck down the power, put the CPU into deep sleep (clock off), and sit there draining negligible power until it receives a wakeup interrupt (IGN line powers up). If that is so, then I am VERY pleased, at least with this element.
On performance and stability;
Its better than the UL135 across the board here. This thing is no rocket, the CPU is actually quite a TURD, though definitely better than RK3188 (despite lower benchmarks). The increased RAM helps with keeping things in memory and responsive. The stability is a night and day improvement over the UL135, which I believe to be attributable to the INTEL element. eMMC performance is abysmal.
Summary so far;
1) The front panel is bigger than DDIN.
2) The front panel sticks out a different amount than UL135.
3) The audio routing policy is BORKED.
4) The UI is ALL CHINA.
5) The Bluetooth stack is INTEL
6) Performance is "OK". Nothing special.
7) Wakeup after sitting all night is INSTANTANEOUS.
More on the GMaps audio routing.
While I haven't yet tried out the latest system update, I HAVE found a way to make GMaps actually speak.
Unfortunately, the "way" is by being on a PHONE CALL.
Yep. When you're on a PHONE call, google maps is HAPPY to speak to you. In fact, the damned thing will NEVER SHUT UP. Obviously. When better to harass you then when you're trying to listen to someone else? Or when you're trying to speak, since it shuts off the microphone in order not to echo back the nav prompts.
(fortunately, there is a setting to tell it not to issue nav sounds during call).
Oh yeah, gmaps also speaks LOUDER than the phone call.
EDIT: people on the joying forum are indicating that the newest update will fix the gmaps voice.
Ok, on to the December 8th UPDATE.
First off, I have NOT really run the update yet. Supposedly, it fixes the GMaps sound problem.
I can confirm that they have NOT (yet) caught on to Blueballs.apk that I wrote (and shared) to regain access to the System Settings --> Bluetooth activity, so bluetooth functionality has not been degraded.
First the install process;
Take their install archive and EXTRACT it to a uSD or USB disk. Either is fine. I used USB.
msdos type partition table, ONE partition, formatted as fat32.
Power on the vehicle, stick in the card or USB, and wait until it prompts you to install the update.
Now is where things get scary....
Starts by updating the MCU, fine, then proceeds to update the Android image. The progress bar moves about a 3rd of the way across, and STOPS. And it just sits there. For about 10 minutes. The only indication that the device isn't completely crashed out is that the little websphere thing is still spinning in the little robot's hands. Ok, eventually it goes, then it says to yank out the disk and it will reboot itself. OK, no problem.
You do it. And it reboots. It plays the boot animation, right up to the last frame... and stops. It just sits there saying "ANDROID" across the screen. Now you're thinking to yourself that THIS THING IS JAMMED UP AND YET WE HAVE NO WAY TO GET INTO RECOVERY!!!!! Crap. Ok, lets just let it be. It sits, and it sits, and about 15 minutes later, the panel lights go out. Ok, at least something changed. Wait another minute and it finally boots. So WHAT THE BLOODY HELL CHINA? Obviously, it must have been optimizing application number 27 of 258, so WHY DID YOU DELETE THE PROGRESS DIALOG, CHINA? WHY?????
So it boots back up, and you notice something funny right off the bat. It has returned to the china launcher. Ok, that's minor. Keep on moving, we can reset it to google launcher momentarily. So now you go into something like k9mail, and it asks you to sign in.
HUH?
But the program remained installed.
THE BLOODY MORONS..... THEY DELETED EVERYTHING FROM /data/data/. They kept /data/app/, but they deleted /data/data/. So everything you've set up now has to be reset. GEE, THANKS SO BLOODY MUCH, CHINA.
So my suggestion... if you're going to update to the Dec8 system image, the sooner you do it, the less of your settings you will lose, because the less of your settings you will have actually SET.
Quick observations;
1) They have NOT fixed the frequency selection on the AM radio. I'm still stuck 2 kHz away from my target frequency.
2) There were THREE phases to the system install; MCU, Android, and "Allapp.pkg".
3) Allapp.pkg contains Settings.apk. Yet there was a Settings.apk in the Android update. Hmmmmm..... wonder if we could just force back the REAL Settings.apk and have a proper functioning system?
4) Allapp.pkg *also* has a SystemUI.apk. Could it be? Could the one in the Android update actually be... no, it couldn't. There's no way that they would actually *include* an AOSP SystemUI, is there?
This is the list of all of the apk files in Allapp.pkg:
Code:
EasyConnectedFY07.apk
assets/CarMan.apk
assets/CarManNotChinese.apk
assets/EasyConnServer.apk
Gmail2.apk
JY-1-C9-BT-V1.0.apk
JY-1-C9-Guide-V1.0.apk
JY-1-C9-Launcher-V1.0.apk
JY-1-C9-MUSIC-V1.0.apk
JY-1-C9-Radio-V1.0.apk
JY-1-C9-Video-V1.0.apk
JY-2-C9-Launcher-V1.0.apk
Maps.apk
PlatformServerSofia.apk
Settings.apk
SFY-1-C9-Steer-V1.0.apk
Sofia-1-C9-Aux-V1.0.apk
sofia-1-C9-CarMark-V1.0.apk
Sofia-1-C9-FileManager-V1.0.apk
Sofia-1-C9-Gallery-V1.0.apk
Sofia-1-C9-RightCamera-V1.0.apk
Sofia-1-C9-Server-V1.0.apk
Sofia-1-C9-TV-V1.0.apk
SystemUI.apk
SYU-1-5009-FloatBtn-V1.0.apk
SYU-1-C9-Calculator-V1.0.apk
SYU-1-C9-Caliration-V1.0.apk
SYU-1-C9-Canbus-V1.0.apk
SYU-1-C9-CarRadio-V1.0.apk
SYU-1-C9-CarUi-V1.0.apk
SYU-1-C9-Graphics-V1.0.apk
SYU-1-C9-Navi-V1.0.apk
SYU-1-C9-SubServer-V1.0.apk
SYU-1-C9-UIServer-V1.0.apk
SYU-1-C9-Update-V1.0.apk
SYU-1-C9-Widget-V1.0.apk
SYU-2-C9-CMSettings-V1.0.apk
SYU-2-C9-EQ-V1.0.apk
Notice... CarMan.apk and CarManNotChinese.apk, LOL.
Oh, and "SYU-1-C9-Caliration-V1.0.apk" is also pretty cool.... ***CALIRATION***
doitright said:
My two units arrived. Played with one of them on the kitchen table late last night.
The row of capacitive buttons are thoroughly pointless -- touches to them don't register at all if your finger doesn't touch it absolutely perfectly.
Having said that, I like them better than the one with the knobs, for the reason I've already mentioned to you -- the lack of deep bezels.
Better? It is definitely a matter of preference. I don't need the capacitive button row, so that doesn't bother me at all.
Note: they have this weird dot thing stuck in the UI as configured by default. Touching the dot brings up a set of buttons -- back, home, menu, etc. So.... CHINA. Obviously, unless you have AT LEAST three copies (SystemUI header bar, row of buttons, the weird dot thing, steering wheel controls) of every button, somebody is going to miss it. Fortunately, there is a control application in the application menu that allows you to disable the stupid dot.
Click to expand...
Click to collapse
I have the 10" version and I got the buttons to work 100% after relearning the touch keys. Open "carmediasettings" then click on Panel keys assignment. Hope that helps =)
xlinkx said:
I have the 10" version and I got the buttons to work 100% after relearning the touch keys. Open "carmediasettings" then click on Panel keys assignment. Hope that helps =)
Click to expand...
Click to collapse
The issue on the 7" isn't the button calibration, just the physical size of the touch areas.
Doesn't matter anyway, since the buttons are all available in other places -- systemui bar and steering wheel.
So today installed the second UL128N2 into the wife's Subaru Forester.
Popped out a parrot asteroid smart (obsolete). The dimensions of the parrot are about a full quarter of an inch in both horizontal and vertical directions. Really had to do a job on the ddin dash adapter on that one, but fortunately was able to avoid having to cut up the actual dashboard.
doitright said:
Summary so far;
1) The front panel is bigger than DDIN.
2) The front panel sticks out a different amount than UL135.
3) The audio routing policy is BORKED.
4) The UI is ALL CHINA.
5) The Bluetooth stack is INTEL
6) Performance is "OK". Nothing special.
7) Wakeup after sitting all night is INSTANTANEOUS.
Click to expand...
Click to collapse
Any news on the instantaneous wake up after a full night? No car battery issues after several nights? I would be very interested in measurements of power consumption when the car is switched off.
Regarding the small font settings, would you suggest to buy the 8" unit if it fits in the car? I guess that the UL138N2 has the same hardware as the UL128N2 only with an 8" screen? And I hope that it has also the instantaneous wake up?
Sent from my SM-N910F using XDA-Developers mobile app
nomailing said:
Any news on the instantaneous wake up after a full night? No car battery issues after several nights? I would be very interested in measurements of power consumption when the car is switched off.
Regarding the small font settings, would you suggest to buy the 8" unit if it fits in the car? I guess that the UL138N2 has the same hardware as the UL128N2 only with an 8" screen? And I hope that it has also the instantaneous wake up?
Sent from my SM-N910F using XDA-Developers mobile app
Click to expand...
Click to collapse
No, I wouldn't suggest that. I would just suggest changing the DPI setting on the 7" to something more suitable.
The UL138N2 WILL NOT fit in a DDIN dash adapter. Its basically a DDIN *box*, with a TABLET bolted out in front of it. Unless your car is an ugly beater, you probably won't want to install something that ugly into it.
As far as overnight power consumption, I have not had any opportunity to test more than a 24 hour stretch. No issues, nor, frankly, do I expect any. They appear to be sleeping the unit correctly, so all of the wireless signals are turned off, and CPU is in clock stopped mode, so like a good quality tablet with the wifi turned off, it'll go for months like that. On a battery as big as a car has, possibly years (assuming that the car battery didn't self discharge in that time, which it will).
doitright,
With your help I was able to get full system menu. Thank you! I was wondering if you were able to get "Ok Google" to work on any screen. It only works on the home screen for my 10" head unit.
xlinkx said:
doitright,
With your help I was able to get full system menu. Thank you! I was wondering if you were able to get "Ok Google" to work on any screen. It only works on the home screen for my 10" head unit.
Click to expand...
Click to collapse
I have no interest in retard mode, so no. I have not.
doitright said:
DPI: Obviously, Android believes that the screen is as large as a TELEVISION. The fonts are all incredibly small, even if you set it to HUGE font. Ok, so right angle triangle has H = root(1024*1024+600*600)=1186.83 pseudopixels. Divide that by 7, makes 170 DPI. Its set to 160 DPI, so Android thinks that the screen is 7.4 inches. Ok, that's not really too far off from reality, but it IS in the wrong direction, to make everything look SMALLER than it should be. The problem is that the screen is roughly 3 feet away from your face. Take a tablet and how far do you hold it away? Roughly 1.5 foot. So lets figure out how big Android should *THINK* the screen is, in order to blow things up to the same scale as a 7" tablet at 1.5 foot.
We're dealing with right triangles again, 7:3 = x:1.5, x = 3.5. Android should think that the screen is 3.5 inches. That puts it up at 338 DPI. Ok, so these numbers are pretty rough, since I'm just guessing the viewing distances, but you get the idea. Crank that DPI WAAAAAAAY UP! Only problem I see, however, is regarding how they coded the SystemUI. If they followed the Android design guidelines, then they coded it with DP's (density independent pixels), which means that when you crank the DPI far enough to make applications look proper, it will make the CHINA look HUGE. Think about it -- they designed their SystemUI panels to show at a "normal" size when the DPI is set way too low. I can see bad things happening.
Click to expand...
Click to collapse
I tried some changes to the dpi in build.prop. I went up to 180 and the ui looks alright still. Then I decided what the heck and went to 300 since my galaxy s5 is at 480...bad idea. The screen gets stuck on the kill apps screen and you can't open anything. Luckily I was able to open es file explorer using adb and edit the build.prop. After rebooting, all os well again.
Moral of the story, don't try anything over...maybe 200?

Speed Sensitive Volume Control via Tasker + Root

I've been playing around with speed sensitive volume control via Tasker (and ROOT is necessary for this function) on my Dasaita PX6 Android 10 MTCE_HA_V3.56_1, and I think I've reached a point where I'm satisfied with the functionality. Since it took me a while to compile the information needed to achieve this, I thought I'd share it for others who may be interested.
Basically, you need to create a looping task that reads GPS and converts it to MPH. You then need multiple profiles that will monitor the GPS loop task and increase the volume when they become active at a certain speed (and decrease it when no longer active). The trick, however, is that in order to actually change the volume you will need to use a broadcast intent to the MCU (which will require ROOT to work properly).
I found a list of key codes that work for my particular unit posted here by Serge_SM: https://forum.xda-developers.com/showpost.php?p=67855471&postcount=17 I'm not sure that this list is exhaustive, and different MCU versions may use different codes for the various functions, so you may need to do more digging if the volume tasks do not work on your headunit.
I have attached the prj.xml file so that you can simply import it into Tasker as a project. From there you can take a look at it and see how it works (and also make any tweaks that are necessary for your own needs).
I found that the volume transition points were best when placed just under various cruising speeds to avoid hysteresis issues, but you can adjust the speed points as you see fit. A two-second delay on the GPS loop seemed to work well without too much of a delay and while not being a huge burden on the system
As always, with stuff like this, there are no guarantees that it will work for you and you try it at your own risk, but hopefully this will be helpful to others out there!

Categories

Resources