Android OS Forum banner

1 - 20 of 197 Posts

·
teh Jax
Joined
·
2,677 Posts
Over the past ~2 weeks, many people have been seeing this bug. We often write it off as being the bug that requires you to flash a Sense ROM, get a lock, then flash back to AOSP, but this is something different. Rather than discussing this separated in half a dozen different threads, I'd like for us to tackle this bug here in a thread focused to it.

I've seen what I believe to be this bug manifest with two different symptoms:
  1. Full GPS works fine while network location works in some apps but gives you an obviously bogus location in other apps.
  2. Full GPS works fine while network location works in some apps and simply fails in other apps.

NOTE: If Full GPS fails for you (such as in Google Maps), then go tackle that problem first before discussing here. If full GPS fails, then you probably have the bug that requires you to flash a Sense ROM. Please do NOT confuse that bug with this one!

I have seen this bug reported for all functional AOSP-based ROMs that I've followed, specifically CM7, OMFGB, OMGB, and Liquid Thunderbread. The funny thing is that some people get this bug while others do not, and there is not a clear pattern.

As I said before, this is clearly happening on pretty much every AOSP-based ROM. What is not clear, however, are what causes this. So it sure would be nice to figure this out. I'm not sure exactly what we need to do to find this but since you all are smart cookies, we'll still figure this out...

So go!

UPDATE (7/24):
This bug appears to happen only when you're using carrier-based network location and NOT when using Wifi-based network location. A known work-around is to connect to a Wifi network. We are still trying to fully understand this, however, to come up with a real fix.​
 

·
teh Jax
Joined
·
2,677 Posts
Discussion Starter · #2 ·
Two theories I have:

1. A ROM out there causes this in a way that it stays persistent through data/system/dalvik wipes and Nandroid restores. If this is the case, I suspect it's a Sense-based ROM, perhaps specifically Sense 3-based. My reason for this theory is not based on hard facts and is entirely circumstantial: I never experienced this bug when it was beginning to be reported yet I was flashing every AOSP ROM in sight, including many nightlies. However, I took a few days to play with the newest Sense ROMs and then came back to AOSP and immediately saw this bug. Also, prior to playing with these Sense-based ROMs, I had never ran the Sense 3-based ROMs before yet I know they're very common.

2. A kernel, or common piece of code in many kernels, are actually at fault for this bug. I know most of our AOSP kernels kang code from one another, so perhaps this is our common factor in a way that seems fairly random across many ROMs. And it could be that this bug is really no direct fault of a kernel but just an unfortunate conflicting sequence of events due to some performance/power optimization in a driver inside a kernel or something else like that.

Anybody else have any theories for this? And any ideas, plans, and time to execute said plan to confirm or rule out my two theories?

Again, I'm not claiming I have the answers here, I'm just trying to get this conversation going in a focused way so you smart folks can figure this out. :)
 

·
Average Android
Joined
·
80 Posts
Honestly, I was having just network location issues and I didn't do a damn thing and it just sorted itself out. It seems to be especially good after Chingy's latest leak. In my opinion, I think perhaps it was tower maintenance or something. I have no explanation as to why it started working for me after doing nothing.

Sent from my ThunderBolt using Tapatalk
 

·
Average Android
Joined
·
80 Posts
Jaxidian said:
Two theories I have:

1. A ROM out there causes this in a way that it stays persistent through data/system/dalvik wipes and Nandroid restores. If this is the case, I suspect it's a Sense-based ROM, perhaps specifically Sense 3-based. My reason for this theory is not based on hard facts and is entirely circumstantial: I never experienced this bug when it was beginning to be reported yet I was flashing every AOSP ROM in sight, including many nightlies. However, I took a few days to play with the newest Sense ROMs and then came back to AOSP and immediately saw this bug. Also, prior to playing with these Sense-based ROMs, I had never ran the Sense 3-based ROMs before yet I know they're very common.

2. A kernel, or common piece of code in many kernels, are actually at fault for this bug. I know most of our AOSP kernels kang code from one another, so perhaps this is our common factor in a way that seems fairly random across many ROMs. And it could be that this bug is really no direct fault of a kernel but just an unfortunate conflicting sequence of events due to some performance/power optimization in a driver inside a kernel or something else like that.

Anybody else have any theories for this? And any ideas, plans, and time to execute said plan to confirm or rule out my two theories?

Again, I'm not claiming I have the answers here, I'm just trying to get this conversation going in a focused way so you smart folks can figure this out. :)
I never ran a Sense 3 based ROM. The last sense ROM I ran was BAMF 1.8.6 before coming over to CM. I didn't have the issue in BAMF.

Sent from my ThunderBolt using Tapatalk
 

·
teh Jax
Joined
·
2,677 Posts
Discussion Starter · #5 ·
ProgHouse said:
Honestly, I was having just network location issues and I didn't do a damn thing and it just sorted itself out. It seems to be especially good after Chingy's latest leak. In my opinion, I think perhaps it was tower maintenance or something. I have no explanation as to why it started working for me after doing nothing.
I'm confused. Did you do nothing or did you flash the new radio?
 

·
Average Android
Joined
·
80 Posts
Jaxidian said:
I'm confused. Did you do nothing or did you flash the new radio?
Lol, my bad. This started happening to me on the last official MR2 OTA radio and I just left it alone and it sorted itself out. After flashing Chingy's leak, network location and GPS are just generally faster in finding me. When I was having the the issue it kept putting me in Arcade, NY. I live in Texas and I thought it odd that it kept putting me in a very specific town in upstate NY without fail.

Sent from my ThunderBolt using Tapatalk
 

·
Android Lover
Joined
·
405 Posts
I'm wondering if it has to do with weak signal. I notice that if I'm in my building (at work) I occasionally get this bogus location. What's odd is that it's the SAME bogus location every time (in another state). If I try to refresh location in my building, it still thinks I'm in my bogus location. As soon as I walk outside and refresh, the location fixes itself.

The above is also while forcing 3g while on the edge of 4g service.

The weirdest thing is that its either bogus all day or accurate all day.

edit:

I think it would be interesting to see if the effected users are in 3g/4g areas or if they are forcing 3g if they are in a 4g area.

Happens to me when I'm in my large work building, forcing 3g in the edge of 4g.
 

·
Premium Member
Joined
·
7,690 Posts
I randomly went from not having this issue to my Beautiful Widgets telling me I was in Estonia (I'm just north of San Fran lol). As soon as I'd hop on WiFi everything was grand. I tried GPS on and all sorts of stuff to no avail. I changed ROMs to a Sense ROM as I noticed I no longer had GPS working. Fixed GPS and ever since my location has been accurate. No clue what to think of any of this info but something weird is up.
 

·
Android Lover
Joined
·
405 Posts
Mustang302LX said:
I randomly went from not having this issue to my Beautiful Widgets telling me I was in Estonia (I'm just north of San Fran lol). As soon as I'd hop on WiFi everything was grand. I tried GPS on and all sorts of stuff to no avail. I changed ROMs to a Sense ROM as I noticed I no longer had GPS working. Fixed GPS and ever since my location has been accurate. No clue what to think of any of this info but something weird is up.
I'm pretty sure that if you're near a wifi hotspot and wifi is on, you won't have the problem. It seems as though when Android wants a coarse location, it will try wifi first, then carrier network (towers). If you have a widget that allows you to force GPS ONLY (don't use wireless or network location), I'm sure you'll see that the location is accurate.

I guess what I'm trying to say is that I believe the issue occurs when you get your location from the carrier network.
 

·
Android Apprentice
Joined
·
240 Posts
My location bug happens when in Google + and not connected to wifi. The verizon triangulation isn't working obviously since it thinks I'm in a different country.
 

·
teh Jax
Joined
·
2,677 Posts
Discussion Starter · #11 ·
skinien said:
I'm pretty sure that if you're near a wifi hotspot and wifi is on, you won't have the problem. It seems as though when Android wants a coarse location, it will try wifi first, then carrier network (towers). If you have a widget that allows you to force GPS ONLY (don't use wireless or network location), I'm sure you'll see that the location is accurate.

I guess what I'm trying to say is that I believe the issue occurs when you get your location from the carrier network.
Interesting theory. I don't know enough about this to know if that's how it works or not. Can somebody who knows for sure confirm this?

psufan5 said:
My location bug happens when in Google + and not connected to wifi. The verizon triangulation isn't working obviously since it thinks I'm in a different country.
And this anecdote appears to support that theory.
 

·
Android Lover
Joined
·
405 Posts
Jaxidian said:
Interesting theory. I don't know enough about this to know if that's how it works or not. Can somebody who knows for sure confirm this?

And this anecdote appears to support that theory.
My first Android phone (Samsung Captivate) had three location settings (Under Menu -> Location & security).

The three settings were something along the lines of:
-Use wireless networks
-Use carrier networks
-Use standalone GPS

CM (and I'm guessing it's all AOSP ) combines the first two:
"Use wireless networks: Location determined by wi-fi and/or mobile networks"

I'm wondering if we had the ability to separate the two, and un-check "Mobile Networks", we could prevent our phones from using network triangulation and giving us bad locations. Or, we could find out which of the three location "tools" is giving us bad data.
 

·
Supporting Member
Joined
·
74 Posts
Just wanted to give some input, i had the location bug only in the aosp weather widget and google +. I did a fix permission and cleared dalvik, no luck. The fix permission actually deleted all my apps. I performed a full wipe. Did not restore ANY apps or settings. Upon first boot i had the bug still but after about 15 min it magically started working. Been working perfect for 2 days now? I wish i could say there was something specific that fixed it but it seemed to have fixed itself after i almost gave up. Im running liquid tb 2.3.
 

·
Supporting Member
Joined
·
74 Posts
"Jaxidian said:
Two theories I have:

1. A ROM out there causes this in a way that it stays persistent through data/system/dalvik wipes and Nandroid restores. If this is the case, I suspect it's a Sense-based ROM, perhaps specifically Sense 3-based. My reason for this theory is not based on hard facts and is entirely circumstantial: I never experienced this bug when it was beginning to be reported yet I was flashing every AOSP ROM in sight, including many nightlies. However, I took a few days to play with the newest Sense ROMs and then came back to AOSP and immediately saw this bug. Also, prior to playing with these Sense-based ROMs, I had never ran the Sense 3-based ROMs before yet I know they're very common.

2. A kernel, or common piece of code in many kernels, are actually at fault for this bug. I know most of our AOSP kernels kang code from one another, so perhaps this is our common factor in a way that seems fairly random across many ROMs. And it could be that this bug is really no direct fault of a kernel but just an unfortunate conflicting sequence of events due to some performance/power optimization in a driver inside a kernel or something else like that.

Anybody else have any theories for this? And any ideas, plans, and time to execute said plan to confirm or rule out my two theories?

Again, I'm not claiming I have the answers here, I'm just trying to get this conversation going in a focused way so you smart folks can figure this out. :)
Im wondering if there is something memory wise that needs to wiped that we are missing, maybe some location data is stored somewhere and not wiped when we perform a factory reset. ever notice that sometimes your wallpaper is still there even though you do a full wipe? I know they say to wipe multiple times but it seems pointless, after cvpcs kinda debunked that myth.
 

·
Android Apprentice
Joined
·
17 Posts
I have experienced this bug. I mostly notice it when using News and Weather widget. It began on CM7 and continued until the release of CM7 RC1.2. Sometimes, on RC1.2 it could find the correct location sometimes not. It seems that after flashing a ROM it would initially determine my location correctly and then sometime later put me somewhere in Alabama (I'm in Hawaii).

I would be interested where it says other people are located in relation to where they actually are. Maybe there's some reason to it all.

I will try to take note of my signal strength when acquiring location to see if that makes a difference. I'm always on 4G and using baseband: 1.39.00.0627,0.01.69.0625r. Aloha.
 

·
Android Aficionado
Joined
·
412 Posts
Even on the newly leaked radio, I'm still having issues. I live in NC, but Weatherbug, Snowstorm, Gasbuddy, etc all think I'm in AL, even with GPS turned on. Only works when I'm connected to WiFi.

Network and GPS location seem to both work fine in Google Maps. At one point I did try a Sense 3 ROM, but I can't say if the issue was present prior, as that was some time ago.
 
1 - 20 of 197 Posts
Top