• After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.

Root [ROM][Beta][FINAL] CM7 Android For Triumph [Tickerguy Edition, Kernel: mine] 11/19

metatroid said:
When selecting an image taken with the camera under the gallery app, to share over messaging, I get a series of SQLiteExceptions, and the MMS hangs on "sending..." until I delete the message.

There are many of these exceptions that flash by on logcat.

Code: 
E/Database(  334): SELECT locale FROM android_metadata failed
E/Database(  334): Failed to setLocale() when constructing, closing the database

E/Database(  334): android.database.sqlite.SQLiteException: no such table: andro
id_metadata
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.native_setLoca
le(Native Method)
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.setLocale(SQLi
teDatabase.java:1987)
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.<init>(SQLiteD
atabase.java:1855)
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.openDatabase(S
QLiteDatabase.java:820)
E/Database(  334):      at android.database.sqlite.SQLiteOpenHelper.getReadableD
atabase(SQLiteOpenHelper.java:197)
E/Database(  334):      at com.android.providers.telephony.TelephonyProvider.que
ry(TelephonyProvider.java:382)
E/Database(  334):      at android.content.ContentProvider$Transport.bulkQuery(C
ontentProvider.java:174)
E/Database(  334):      at android.content.ContentProviderNative.onTransact(Cont
entProviderNative.java:111)
E/Database(  334):      at android.os.Binder.execTransact(Binder.java:320)
E/Database(  334):      at dalvik.system.NativeStart.run(Native Method)
I/PowerWidget(24561): Setting up button: toggleSound
E/DatabaseUtils(  334): Writing exception to parcel
E/DatabaseUtils(  334): android.database.sqlite.SQLiteException: no such table:
android_metadata
E/DatabaseUtils(  334):         at android.database.sqlite.SQLiteDatabase.native
_setLocale(Native Method)
E/DatabaseUtils(  334):         at android.database.sqlite.SQLiteDatabase.setLoc
ale(SQLiteDatabase.java:1987)
E/DatabaseUtils(  334):         at android.database.sqlite.SQLiteDatabase.<init>
(SQLiteDatabase.java:1855)
E/DatabaseUtils(  334):         at android.database.sqlite.SQLiteDatabase.openDa
tabase(SQLiteDatabase.java:820)
E/DatabaseUtils(  334):         at android.database.sqlite.SQLiteOpenHelper.getR
eadableDatabase(SQLiteOpenHelper.java:197)
E/DatabaseUtils(  334):         at com.android.providers.telephony.TelephonyProv
ider.query(TelephonyProvider.java:382)
E/DatabaseUtils(  334):         at android.content.ContentProvider$Transport.bul
kQuery(ContentProvider.java:174)
E/DatabaseUtils(  334):         at android.content.ContentProviderNative.onTrans
act(ContentProviderNative.java:111)
E/DatabaseUtils(  334):         at android.os.Binder.execTransact(Binder.java:32
0)
E/DatabaseUtils(  334):         at dalvik.system.NativeStart.run(Native Method)
D/dalvikvm(18172): GC_EXTERNAL_ALLOC freed 96K, 52% free 2782K/5703K, external 9
39K/1033K, paused 213ms
E/SqliteWrapper(25701): Catch a SQLiteException when query:
E/SqliteWrapper(25701): android.database.sqlite.SQLiteException: no such table:
android_metadata
E/SqliteWrapper(25701):         at android.database.DatabaseUtils.readExceptionF
romParcel(DatabaseUtils.java:158)
E/SqliteWrapper(25701):         at android.database.DatabaseUtils.readExceptionF
romParcel(DatabaseUtils.java:114)
E/SqliteWrapper(25701):         at android.content.ContentProviderProxy.bulkQuer
yInternal(ContentProviderNative.java:330)
E/SqliteWrapper(25701):         at android.content.ContentProviderProxy.query(Co
ntentProviderNative.java:366)
E/SqliteWrapper(25701):         at android.content.ContentResolver.query(Content
Resolver.java:271)
E/SqliteWrapper(25701):         at android.database.sqlite.SqliteWrapper.query(S
qliteWrapper.java:59)
E/SqliteWrapper(25701):         at com.android.mms.transaction.TransactionSettin
gs.<init>(TransactionSettings.java:66)
E/SqliteWrapper(25701):         at com.android.mms.transaction.TransactionServic
e$ServiceHandler.handleMessage(TransactionService.java:628)
E/SqliteWrapper(25701):         at android.os.Handler.dispatchMessage(Handler.ja
va:99)
E/SqliteWrapper(25701):         at android.os.Looper.loop(Looper.java:130)
E/SqliteWrapper(25701):         at android.os.HandlerThread.run(HandlerThread.ja
va:60)
W/TransactionService(25701): Exception occurred while handling message: { what=1
 when=-84ms arg1=1 obj=com.android.mms.transaction.TransactionBundle@2b062ae8 }
W/TransactionService(25701): android.database.sqlite.SQLiteException: no such ta
ble: android_metadata
W/TransactionService(25701):    at android.database.DatabaseUtils.readExceptionF
romParcel(DatabaseUtils.java:158)
W/TransactionService(25701):    at android.database.DatabaseUtils.readExceptionF
romParcel(DatabaseUtils.java:114)
W/TransactionService(25701):    at android.content.ContentProviderProxy.bulkQuer
yInternal(ContentProviderNative.java:330)
W/TransactionService(25701):    at android.content.ContentProviderProxy.query(Co
ntentProviderNative.java:366)
W/TransactionService(25701):    at android.content.ContentResolver.query(Content
Resolver.java:271)
W/TransactionService(25701):    at android.database.sqlite.SqliteWrapper.query(S
qliteWrapper.java:59)
W/TransactionService(25701):    at com.android.mms.transaction.TransactionSettin
gs.<init>(TransactionSettings.java:66)
W/TransactionService(25701):    at com.android.mms.transaction.TransactionServic
e$ServiceHandler.handleMessage(TransactionService.java:628)
W/TransactionService(25701):    at android.os.Handler.dispatchMessage(Handler.ja
va:99)
W/TransactionService(25701):    at android.os.Looper.loop(Looper.java:130)
W/TransactionService(25701):    at android.os.HandlerThread.run(HandlerThread.ja
va:60)
I/Database(  334): sqlite returned: error code = 8, msg = statement aborts at 1:
 [PRAGMA journal_mode = WAL;]
E/Database(  334): sqlite3_exec - Failed to set WAL mode for [/data/data/com.and
roid.providers.telephony/databases/telephony.db]
I/Database(  334): sqlite returned: error code = 8, msg = statement aborts at 37
: [CREATE TABLE IF NOT EXISTS android_metadata (locale TEXT)] attempt to write a
 readonly database
E/Database(  334): CREATE TABLE android_metadata failed
E/Database(  334): Failed to setLocale() when constructing, closing the database

E/Database(  334): android.database.sqlite.SQLiteException: attempt to write a r
eadonly database
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.native_setLoca
le(Native Method)
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.setLocale(SQLi
teDatabase.java:1987)
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.<init>(SQLiteD
atabase.java:1855)
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.openDatabase(S
QLiteDatabase.java:820)
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.openOrCreateDa
tabase(SQLiteDatabase.java:854)
E/Database(  334):      at android.database.sqlite.SQLiteDatabase.openOrCreateDa
tabase(SQLiteDatabase.java:847)
E/Database(  334):      at android.app.ContextImpl.openOrCreateDatabase(ContextI
mpl.java:572)
E/Database(  334):      at android.database.sqlite.SQLiteOpenHelper.getWritableD
atabase(SQLiteOpenHelper.java:118)
E/Database(  334):      at android.database.sqlite.SQLiteOpenHelper.getReadableD
atabase(SQLiteOpenHelper.java:187)
E/Database(  334):      at com.android.providers.telephony.TelephonyProvider.que
ry(TelephonyProvider.java:382)
E/Database(  334):      at android.content.ContentProvider$Transport.bulkQuery(C
ontentProvider.java:174)
E/Database(  334):      at android.content.ContentProviderNative.onTransact(Cont
entProviderNative.java:111)
E/Database(  334):      at android.os.Binder.execTransact(Binder.java:320)
E/Database(  334):      at dalvik.system.NativeStart.run(Native Method)
E/SQLiteOpenHelper(  334): Couldn't open telephony.db for writing (will try read
-only):
E/SQLiteOpenHelper(  334): android.database.sqlite.SQLiteException: attempt to w
rite a readonly database
E/SQLiteOpenHelper(  334):      at android.database.sqlite.SQLiteDatabase.native
_setLocale(Native Method)
E/SQLiteOpenHelper(  334):      at android.database.sqlite.SQLiteDatabase.setLoc
ale(SQLiteDatabase.java:1987)
E/SQLiteOpenHelper(  334):      at android.database.sqlite.SQLiteDatabase.<init>
(SQLiteDatabase.java:1855)
E/SQLiteOpenHelper(  334):      at android.database.sqlite.SQLiteDatabase.openDa
tabase(SQLiteDatabase.java:820)
E/SQLiteOpenHelper(  334):      at android.database.sqlite.SQLiteDatabase.openOr
CreateDatabase(SQLiteDatabase.java:854)
E/SQLiteOpenHelper(  334):      at android.database.sqlite.SQLiteDatabase.openOr
CreateDatabase(SQLiteDatabase.java:847)
E/SQLiteOpenHelper(  334):      at android.app.ContextImpl.openOrCreateDatabase(
ContextImpl.java:572)
E/SQLiteOpenHelper(  334):      at android.database.sqlite.SQLiteOpenHelper.getW
ritableDatabase(SQLiteOpenHelper.java:118)
E/SQLiteOpenHelper(  334):      at android.database.sqlite.SQLiteOpenHelper.getR
eadableDatabase(SQLiteOpenHelper.java:187)
E/SQLiteOpenHelper(  334):      at com.android.providers.telephony.TelephonyProv
ider.query(TelephonyProvider.java:382)
E/SQLiteOpenHelper(  334):      at android.content.ContentProvider$Transport.bul
kQuery(ContentProvider.java:174)
E/SQLiteOpenHelper(  334):      at android.content.ContentProviderNative.onTrans
act(ContentProviderNative.java:111)
E/SQLiteOpenHelper(  334):      at android.os.Binder.execTransact(Binder.java:32
0)
E/SQLiteOpenHelper(  334):      at dalvik.system.NativeStart.run(Native Method)
I/Database(  334): sqlite returned: error code = 1, msg = no such table: android
_metadata

I'm not certain they are all exactly the same, nor how many of these flash by since it all happens within a second of clicking the "Send MMS" button, and my scroll buffer is not set to hold that many lines.
Click to expand...

Coincidentally, I noticed this today too, when I tried to send a MMS from the gallery. They stick on "sending..."

I tried changing the state of the radios around but to no avail.
 
kizniche said:
Coincidentally, I noticed this today too, when I tried to send a MMS from the gallery. They stick on "sending..."

I tried changing the state of the radios around but to no avail.
Click to expand...

This is strange. I'm on my latest personal build (CM7-Triumph-09222011) and I was able to send 2 MMS messages in a row, but the other person receives them as blank text messages (which I don't think is a problem with CM7-MT).

EDIT: The blank messages only occur when I capture a picture.
 
I have another bug to report. This has happened twice to me so far since using the 9/20 build.

Symptoms: Pressing the power button to turn the screen on brings the softkeys and the screen on, but instead of seeing the lock screen, I see my phone's state as it was before it went to sleep (as though it is residual memory of the video before sleep). I know it is not the current state of the phone because I just received a txt and was turning on the screen to view it. There would normally be a notification in the top bar, but the screen I see isn't updated.

From this point I can touch the screen and long press the power button (power button gives hepatic feedback when held down) for input. The first time this happened I was able to put in my visual lock pattern even though I could not see any feedback of my input (I have the muscle memory down). I tried this latest time to unlock it this way but could not because the area of the visual pattern was much smaller and hence harder to get right (there was an alarm set and the app "contact owner" was displaying contact information on the lockscreen, making the 3x3 dots for the visual lock much more condensed). I was able to long-press the power button and after a couple tries get the phone to reboot by pressing in the general direction of where I remember the reboot menu to be.

My setup includes running on battery, app "contact owner" running (which I don't believe to be affecting this), an alarm (although the alarm was not set the first time this happened), and these system settings that affect the lockscreen:

Location and security/Screen unlock/Use visible pattern [X]
CyanogenMod settings/Lockscreen/Disable lockscreen on security [X]
CyanogenMod settings/Lockscreen/Delay and timeout/Security Lock/Screen timeout delay [5s]
CyanogenMod settings/Lockscreen/Delay and timeout/Security Lock/Screen turned off delay [1m]
CyanogenMod settings/Lockscreen/Delay and timeout/Slide Lock/Enable slide lock delay [X]
CyanogenMod settings/Lockscreen/Delay and timeout/Slide Lock/Screen timeout delay [15s]
CyanogenMod settings/Lockscreen/Delay and timeout/Slide Lock/Screen turned off delay [15s]

I more frequently, when bringing the phone from sleep, have a similar symptom of the screen displaying the state before it went to sleep, but after a second or two it "unfreezes" to display the lock screen or, if it has been less than 15 seconds per the above settings, bypasses the lockscreen,
 
mantera said:
Yeah, basically, it seems that if I let my phone sleep for a long time (I haven't exactly pinpointed how long "a long time" is yet but seems to be at least > 30 min so far)
Click to expand...

I think its < 30 mins. I have a Tasker profile set up to check for internet (it tries loading Google's logo) and if it fails, it toggles the 3g. I also have it notify me when this happens, and it happens *every time it runs* aka every 30 mins.

BTW, get Tasker. It rocks. ;-)
 
I have a minor bug to report. When I go to listen to google music, the app crashes. I'm not saying this to complain, but to help with the process.
 
I'd like to help in any way I can.

Sleep problems have escalated today. It takes over 5 seconds to wake out of sleep, and about 1 out of 3 times I try to wake, the phone restarts. Today, for the first time, my phone vibrated and wouldn't stop until it restarted.
 
PsyberS said:
I think its < 30 mins. I have a Tasker profile set up to check for internet (it tries loading Google's logo) and if it fails, it toggles the 3g. I also have it notify me when this happens, and it happens *every time it runs* aka every 30 mins.

BTW, get Tasker. It rocks. ;-)
Click to expand...


Nice. Got me thinking, so I made a similar profile that you mentioned but then am having it append a txt file with the amount of minutes since it lost connection last to see if there is a trend. It's set to check every 2 minutes. I'll have some usable data in a little while.
 
Metaroid/Kin, I cannot reproduce this (or any problems with sending pictures for that matter - I do this a dozen or more times a day as a matter of course.) Did you clear data when you loaded the latest version? If not do the following from a plugged in PC or Linux machine with adb on it:

adb shell

# cd /data/data/com.android.providers.telephony/databases
# ls

There should be three files in there starting with "telephony". If there are, remove the other two EXCEPT "telephony.db". Do not kill the "mmssms" ones or your messages will disappear!

Now reboot.

That will clear the APNs - I suspect what's happening is that you have an old APN related telephony database file in the SHM and WAL files, and the forced overwrite on boot is replacing the DB file but the others are out of sync as they're not the same. That may be causing this problem, although IN THEORY it shouldn't happen as the script does attempt to remove all of those suffixed files on boot.

The WAL and SHM files should get re-created on boot.

The errors you're getting are errors related to an internal problem with the APN database.

(I sent MMS images from the gallery and also directly from the camera all the time and just tried it again - no problems)
 
tickerguy said:
Metaroid/Kin, I cannot reproduce this (or any problems with sending pictures for that matter - I do this a dozen or more times a day as a matter of course.) Did you clear data when you loaded the latest version? If not do the following from a plugged in PC or Linux machine with adb on it:

adb shell

# cd /data/data/com.android.providers.telephony/databases
# ls

There should be three files in there starting with "telephony". If there are, remove the other two EXCEPT "telephony.db". Do not kill the "mmssms" ones or your messages will disappear!

Now reboot.

That will clear the APNs - I suspect what's happening is that you have an old APN related telephony database file in the SHM and WAL files, and the forced overwrite on boot is replacing the DB file but the others are out of sync as they're not the same. That may be causing this problem, although IN THEORY it shouldn't happen as the script does attempt to remove all of those suffixed files on boot.

The WAL and SHM files should get re-created on boot.

The errors you're getting are errors related to an internal problem with the APN database.

(I sent MMS images from the gallery and also directly from the camera all the time and just tried it again - no problems)
Click to expand...

I cleared everything before loading the 9/20 version.

The files I have in that dir are mmssms.db, mmssms.db-shm, mmssms.db-wal, telephony.db

Also for testing sake, I tried sending a MMS from the camera and still hangs on "sending..."

Edit: If there isn't anything else you can suggest I try to resolve this, I'll recheck the md5, format and reflash the latest version tonight to see if it resolves this.
 
Aha. The first update-cm-7.1.0-AP0.3-Triumph-KANG-signed.zip I downloaded and I have flashed now has a MD5 of bcd853e3fa1f7938176e5a12bc0d78fb.

I just redownloaded the file twice (to be sure of the true MD5) and the two I just downlaoded have a MD5 of 0013bb40189cc0c9d08b82f0d004e1b3. The new files are also over a half meg larger.

I'll reflash and see if things improve.

Edit: Strange... actually I just checked the MD5 of the file on my SD card (the file that was flashed) and it is 0013bb40189cc0c9d08b82f0d004e1b3. So that was a dead end, but I'll be reflashing anyway to see if it allows me to send MMS.
 
If that doesn't fix it let me know - it's not difficult to shut off the offending things that are doing the clear, although it should not hose you - the script that does the forced load is no longer necessary, but I left it turned on so as to prevent vestiges from screwing people (from previous loads.) It will go away in the next version.
 
Hey I just wanted to let everyone know that I installed a CM7 theme from the market and it ended up erasing my pictures and videos. The theme worked but I just wanted to let everyone know that if anyone else was thinking of doing this a backup wouldn't hurt. And THANKS ticker, this rom is great.
 
Triumphfeen said:
Can anyone please help me my 3g n service stay white thry wont change i tried factory resests reinstalling the rom and all fails
Click to expand...
Go to your app drawer....open up "Theme Chooser"....you can pick from 3 different themes....your 3g-Wifi, Battery color, will change...:D
 
Triumphfeen said:
Tried that! Notification bar goes away nd comes back woth the same color only the battery color changed
Click to expand...

Did you clear "system" before flashing the ROM?
That happened to me once. Seems like there were leftovers from an on build in the system and I had to format it to get the colored icons.
 
Triumphfeen said:
Tried that! Notification bar goes away nd comes back woth the same color only the battery color changed
Click to expand...

Guys, I think triumphfeen is talking about something completely different than theming.

He says:

Can anyone please help me my 3g n service stay white thry wont change i tried factory resests reinstalling the rom and all fails
Click to expand...

This behavior you described is when you're not logged into your Google account. That's why your 3G icons and signal bars stay white.
 
Thanks iisac but p-men was actually right once I formatted system then reinstalled it work! THANK YOU P-MEN! N THANK YOU TICKER AND IISAC FOR YOUR WORK!
 
Status as of right now:

1. My repo sync is hosed, but I am correcting that. I am also updating all the CM7 changes in the (approximately) last month. Until that's corrected, I'm stuck. Getting to a current codebase sync with CM7 is important and as I noted that was going to be where a lot of my time was focused this weekend.

2. I may have working HDMI video; I have been able to get the daemon to come up, which is the gating factor on detecting the fact that it is plugged in. I have no expectation that I have working HDMI audio as the audio path switch is explicitly commented out in the code and framework support, which we do not have, is required for that to work. I have no HDMI cable, however, so I have no means to test it. I will release a build with it, but if it doesn't work I have no means to debug it and I have insufficient interest in this to be worth buying a cable and debugging is going to be lots of fun given where my HDTV is in relationship to my working area. So......if people want this to work, someone has to send me a cable and essentially "gift" that to me, and there's no promise it will work even if you do - I may not be able to get it to function, especially on the audio side. This is one of those things that is either pretty easy or extremely difficult. In short HDMI video may magically appear and it may not, but until I have a cable I've gone as far as I can with it.

3. I'm working on an automatic detection of the data drop problem, which will allow me to patch the code to automatically detect and deal with it. That's not a fix, it's a kludge workaround, but it will at least restore connectivity when it drops. It appears that turning off data (NOT going into airplane mode and out) is all that is necessary to restore connectivity. If you're having this problem and can reproduce it (I cannot reproduce it on demand, although it happens once in a while) when it occurs please go into the settings menu and turn off data, then turn it back on and see if the problem clears. I am very interested in reports of whether that fixes it or not for others - it does for me - because there's a big difference between cycling the entire radio and cycling just the data side in terms of power consumption and disruption to whatever else you're doing. Detecting the state of the unit when it happens will require that I do quite a bit of digging as there is (once again) zero documentation as to where the state of things (e.g. whether a call is in process, where the entry points are to turn on and off the data connection is, whether data is actually active as Wifi is off, etc) are stored and can be easily queried and accessed.

4. I may be able to (partially) alleviate the interference between 3g data and Wifi, although doing so will hurt Wifi performance, perhaps unacceptably so. This is being actively looked at. This interference problem appears to be a hardware problem and it also appears there's a lesser degree interference between bluetooth and data when it is actively transmitting, which really sucks.

5. First stab at USB tethering was unsuccessful. I'm not at all sure why as of yet as there are no obvious errors being thrown (unlike Wifi tethering) but it shuts itself off on start. The problem may be a driver issue in my PCs; I'm going to leave that turned on in my next build in the event it helps someone. Bluetooth tethering if initiated from the PC using nothing more than "Connect using Access point" works, but requires you set the MTU on the PC end to less than 1470 (on Windows 7 it defaults to 1477, which fails.) There allegedly are "munge" options in the Linux kernel that will allow repackaging packets to resolve the mismatch but I've not yet found them and the ones someone else sent me did nothing (if someone knows where they are please shout!) Bluetooth tethering, incidentally, is native in CM7 and as such it will remain up when the unit sleeps and is very low-overhead. Given the known bad interaction between Wifi and 3g data you might prefer to use Bluetooth anyway. If you want tethering in CM7 "right now" and the other device has bluetooth this is the solution that works right now, does not rely on external software and works well.

6. I know Isaac is efforting the FFC; I've taken a few stabs at it this week as well without success. I suspect that a true fix will require using the CM7 camera interface solely, which should be entirely possible since the camera (both of them) is exposed in the device class structure. Getting what we have now working was a kludge that involved using bits of the cherry shared libraries; the likely problem with the FFC is similar to the GPS in that I suspect the FFC has to be poked to wake up, and isn't being in the cherry libs. Using the class structure should fix that, if we can figure out how to tell the code to do that.

6. I'm having severe trouble getting a recovery image that will mount an EXT4 filesystem. The odd part is that the kernel in the running copy here and my local kernel copy both have ext4 in them, but attempting to use the same kernel in the recovery_kernel location builds a recovery that blows up when trying to format /system with ext4. This is severely pissing me off as it is precluding moving to ext4; I'm still working it and it's probably something stupid on my part, as Isaac has a working ext4 recovery.

That's it for right now.
 
tickerguy said:
Status as of right now:

1. My repo sync is hosed, but I am correcting that. I am also updating all the CM7 changes in the (approximately) last month. Until that's corrected, I'm stuck. Getting to a current codebase sync with CM7 is important and as I noted that was going to be where a lot of my time was focused this weekend.

2. I may have working HDMI video; I have been able to get the daemon to come up, which is the gating factor on detecting the fact that it is plugged in. I have no expectation that I have working HDMI audio as the audio path switch is explicitly commented out in the code and framework support, which we do not have, is required for that to work. I have no HDMI cable, however, so I have no means to test it. I will release a build with it, but if it doesn't work I have no means to debug it and I have insufficient interest in this to be worth buying a cable and debugging is going to be lots of fun given where my HDTV is in relationship to my working area. So......if people want this to work, someone has to send me a cable and essentially "gift" that to me, and there's no promise it will work even if you do - I may not be able to get it to function, especially on the audio side. This is one of those things that is either pretty easy or extremely difficult. In short HDMI video may magically appear and it may not, but until I have a cable I've gone as far as I can with it.

3. I'm working on an automatic detection of the data drop problem, which will allow me to patch the code to automatically detect and deal with it. That's not a fix, it's a kludge workaround, but it will at least restore connectivity when it drops. It appears that turning off data (NOT going into airplane mode and out) is all that is necessary to restore connectivity. If you're having this problem and can reproduce it (I cannot reproduce it on demand, although it happens once in a while) when it occurs please go into the settings menu and turn off data, then turn it back on and see if the problem clears. I am very interested in reports of whether that fixes it or not for others - it does for me - because there's a big difference between cycling the entire radio and cycling just the data side in terms of power consumption and disruption to whatever else you're doing. Detecting the state of the unit when it happens will require that I do quite a bit of digging as there is (once again) zero documentation as to where the state of things (e.g. whether a call is in process, where the entry points are to turn on and off the data connection is, whether data is actually active as Wifi is off, etc) are stored and can be easily queried and accessed.

4. I may be able to (partially) alleviate the interference between 3g data and Wifi, although doing so will hurt Wifi performance, perhaps unacceptably so. This is being actively looked at. This interference problem appears to be a hardware problem and it also appears there's a lesser degree interference between bluetooth and data when it is actively transmitting, which really sucks.

5. First stab at USB tethering was unsuccessful. I'm not at all sure why as of yet as there are no obvious errors being thrown (unlike Wifi tethering) but it shuts itself off on start. The problem may be a driver issue in my PCs; I'm going to leave that turned on in my next build in the event it helps someone. Bluetooth tethering if initiated from the PC using nothing more than "Connect using Access point" works, but requires you set the MTU on the PC end to less than 1470 (on Windows 7 it defaults to 1477, which fails.) There allegedly are "munge" options in the Linux kernel that will allow repackaging packets to resolve the mismatch but I've not yet found them and the ones someone else sent me did nothing (if someone knows where they are please shout!) Bluetooth tethering, incidentally, is native in CM7 and as such it will remain up when the unit sleeps and is very low-overhead. Given the known bad interaction between Wifi and 3g data you might prefer to use Bluetooth anyway. If you want tethering in CM7 "right now" and the other device has bluetooth this is the solution that works right now, does not rely on external software and works well.

6. I know Isaac is efforting the FFC; I've taken a few stabs at it this week as well without success. I suspect that a true fix will require using the CM7 camera interface solely, which should be entirely possible since the camera (both of them) is exposed in the device class structure. Getting what we have now working was a kludge that involved using bits of the cherry shared libraries; the likely problem with the FFC is similar to the GPS in that I suspect the FFC has to be poked to wake up, and isn't being in the cherry libs. Using the class structure should fix that, if we can figure out how to tell the code to do that.

6. I'm having severe trouble getting a recovery image that will mount an EXT4 filesystem. The odd part is that the kernel in the running copy here and my local kernel copy both have ext4 in them, but attempting to use the same kernel in the recovery_kernel location builds a recovery that blows up when trying to format /system with ext4. This is severely pissing me off as it is precluding moving to ext4; I'm still working it and it's probably something stupid on my part, as Isaac has a working ext4 recovery.

That's it for right now.
Click to expand...

Here's a link to a list of hdmi to hdmi micro cables on ebay.
http://www.ebay.com/sch/i.html?_nkw...w=&_osacat=149974&_trksid=p3286.c0.m270.l1313http://rover.ebay.com/rover/1/711-5...tref=1&amp;_fln=1&amp;_trksid=p3286.c0.m283:D
 
tickerguy said:
Status as of right now:

1. My repo sync is hosed, but I am correcting that. I am also updating all the CM7 changes in the (approximately) last month. Until that's corrected, I'm stuck. Getting to a current codebase sync with CM7 is important and as I noted that was going to be where a lot of my time was focused this weekend.

2. I may have working HDMI video; I have been able to get the daemon to come up, which is the gating factor on detecting the fact that it is plugged in. I have no expectation that I have working HDMI audio as the audio path switch is explicitly commented out in the code and framework support, which we do not have, is required for that to work. I have no HDMI cable, however, so I have no means to test it. I will release a build with it, but if it doesn't work I have no means to debug it and I have insufficient interest in this to be worth buying a cable and debugging is going to be lots of fun given where my HDTV is in relationship to my working area. So......if people want this to work, someone has to send me a cable and essentially &quot;gift&quot; that to me, and there's no promise it will work even if you do - I may not be able to get it to function, especially on the audio side. This is one of those things that is either pretty easy or extremely difficult. In short HDMI video may magically appear and it may not, but until I have a cable I've gone as far as I can with it.

3. I'm working on an automatic detection of the data drop problem, which will allow me to patch the code to automatically detect and deal with it. That's not a fix, it's a kludge workaround, but it will at least restore connectivity when it drops. It appears that turning off data (NOT going into airplane mode and out) is all that is necessary to restore connectivity. If you're having this problem and can reproduce it (I cannot reproduce it on demand, although it happens once in a while) when it occurs please go into the settings menu and turn off data, then turn it back on and see if the problem clears. I am very interested in reports of whether that fixes it or not for others - it does for me - because there's a big difference between cycling the entire radio and cycling just the data side in terms of power consumption and disruption to whatever else you're doing. Detecting the state of the unit when it happens will require that I do quite a bit of digging as there is (once again) zero documentation as to where the state of things (e.g. whether a call is in process, where the entry points are to turn on and off the data connection is, whether data is actually active as Wifi is off, etc) are stored and can be easily queried and accessed.

4. I may be able to (partially) alleviate the interference between 3g data and Wifi, although doing so will hurt Wifi performance, perhaps unacceptably so. This is being actively looked at. This interference problem appears to be a hardware problem and it also appears there's a lesser degree interference between bluetooth and data when it is actively transmitting, which really sucks.

5. First stab at USB tethering was unsuccessful. I'm not at all sure why as of yet as there are no obvious errors being thrown (unlike Wifi tethering) but it shuts itself off on start. The problem may be a driver issue in my PCs; I'm going to leave that turned on in my next build in the event it helps someone. Bluetooth tethering if initiated from the PC using nothing more than &quot;Connect using Access point&quot; works, but requires you set the MTU on the PC end to less than 1470 (on Windows 7 it defaults to 1477, which fails.) There allegedly are &quot;munge&quot; options in the Linux kernel that will allow repackaging packets to resolve the mismatch but I've not yet found them and the ones someone else sent me did nothing (if someone knows where they are please shout!) Bluetooth tethering, incidentally, is native in CM7 and as such it will remain up when the unit sleeps and is very low-overhead. Given the known bad interaction between Wifi and 3g data you might prefer to use Bluetooth anyway. If you want tethering in CM7 &quot;right now&quot; and the other device has bluetooth this is the solution that works right now, does not rely on external software and works well.

6. I know Isaac is efforting the FFC; I've taken a few stabs at it this week as well without success. I suspect that a true fix will require using the CM7 camera interface solely, which should be entirely possible since the camera (both of them) is exposed in the device class structure. Getting what we have now working was a kludge that involved using bits of the cherry shared libraries; the likely problem with the FFC is similar to the GPS in that I suspect the FFC has to be poked to wake up, and isn't being in the cherry libs. Using the class structure should fix that, if we can figure out how to tell the code to do that.

6. I'm having severe trouble getting a recovery image that will mount an EXT4 filesystem. The odd part is that the kernel in the running copy here and my local kernel copy both have ext4 in them, but attempting to use the same kernel in the recovery_kernel location builds a recovery that blows up when trying to format /system with ext4. This is severely pissing me off as it is precluding moving to ext4; I'm still working it and it's probably something stupid on my part, as Isaac has a working ext4 recovery.

That's it for right now.
Click to expand...


Ticker, I am also still having the issue where the screen never turns off. The screen is black, but in low light you can see that the backlight is still on.
 
Check your options eb - the only time I've seen that is on a cold boot when plugged, and it's a one-time deal (on boot it gets out of sync and will do this if on charge.) I've never had it happen with the phone off-charge in normal use, or once the phone has come up and is running.

If you've customized things in the CM7 settings its possible that's the cause, or if you're using something like Tasker. If all else fails clear /data and /cache (cold boot) and then clear /system and reload.

Sam: I'm aware that HDMI cables are available on eBAY and Amazon and they're not terribly expensive. I guess you didn't understand what I said - I have no use for this option personally, and thus working on this would require me to spend money (even if just a bit) on something that has zero value to me. I'm frankly astonished that you would think that I'm going to not only spend lots of time and give away what I work on (which I'm doing now) but also spend money on something that has zero marginal value to me.

Some of you folks ought to think before you speak. This is getting to be rather "not fun any more" considering that this is the third or fourth time something like that has happened in the last couple weeks. I do have a tolerance limit for that sort of thing and you might want to consider that rather carefully.
 
ticker - other people and myself have a bug where the backlight and softkey lights turn on when trying to wake the phone, but the screen won't actually show. seems like it happens if you leave wifi on, i haven't had any problems when i manually turn off wifi
 
You must log in or register to reply here.
Back
Top Bottom