All Things GPS

[chrisinsocalif]

This may sound really stupid because I have No programming experience whatsoever, but I know the broadcom chip shares cycles with the CPU. Perhaps there is a buffering overflow, CPU data exchange or memory problem when trying to use the GPS which is fixed with a reboot since it clears any data in the CPU cycle or the memory. Maybe someone can give a more accurate technical perspective but just a thought.

 

[krichek]

krichek -

Thanks for the posts.

Are you running the stock ROM? Are you using the default GPS settings, or anything other than the XDA posted "fix"?

Curious since I haven't seen the GPS Test "In Use" # more than 3 except with the leaked Samsung firmware patch.

Sorry, it was late I totally forgot to add that heh..

I'm using the stock rom with RyanZA's lag fix. My wife's phone has the deodexed rom with the older ext3 to sdcard lag fix. Both are using totally stock GPS settings, neither has had any of the gps "fixes" because until I started reading some of the threads on various message boards I didn't know there was a problem that needed "fixing".

 

[wildblue]

I'm using the stock rom with RyanZA's lag fix. My wife's phone has the deodexed rom with the older ext3 to sdcard lag fix. Both are using totally stock GPS settings, neither has had any of the gps "fixes" because until I started reading some of the threads on various message boards I didn't know there was a problem that needed "fixing".

Interesting. Very hard to sort out all the variables at a distance, but it may be the lag fix that has a side benefit of optimizing use of the GPS chip. Not sure that's been a consistent result from folks who've installed the lag fix, but getting accurate data in a situation like this is nearly impossible.

End result is the same, we'll see what Samsung comes up with next month.

 

[SamsungVibrant]

Here is some internal T-mobile communication.

Dated the 18th.

Basically says:
Based on frontline and cust feedback, T-mobile and Samsung will partner to deploy a new software release to customers...expected to rollout in September.

Then it tells reps what to say to customers who inquir about the update, with a statement from Samsung.

The good news is, we might get this before the end of sept, if reps are being told "coming soon!"

 

[Robinelli]

I called customer service a couple of days ago b/c my speaker was going haywire. While on the phone I mentioned the GPS and they told me a fix was coming in 2 weeks and to hold tight.

 

[robo21]

I called customer service a couple of days ago b/c my speaker was going haywire. While on the phone I mentioned the GPS and they told me a fix was coming in 2 weeks and to hold tight.

Hey that's something, they are getting specific, that is a step in the right direction!

 

[Robinelli]

I should add, I decided to start leaving my GPS ON all of the time and that has solved the reboot issue. I don't need to reboot anymore to get the GPS to find me. It isn't get stuck anymore in other words. It is working really well. It takes less than 3 minutes to find me now. I'd say it takes 1 and sometimes up to 2. That's not bad though!

 

[robo21]

Here is some internal T-mobile communication.

Dated the 18th.

Basically says:
Based on frontline and cust feedback, T-mobile and Samsung will partner to deploy a new software release to customers...expected to rollout in September.

Then it tells reps what to say to customers who inquir about the update, with a statement from Samsung.

The good news is, we might get this before the end of sept, if reps are being told "coming soon!"

Excellent find Sammy!

 

[tweak89]

Ugh, this can't come soon enough, I pray it really does fix the problems that have plagued the gps.

I realized this weekend just how much I rely on it. I was hunting for a pop-up camper and found several to look at this weekend. Normally I would have used the GPS on my phone (formerly a G1) to look up and navigate to various addresses. Heck, I couldn't even use the atm locator to take out the cash to buy one because the BoA app couldn't determine my location using wireless networks only. Couldn't even get a lock with 11 sats in view. It seems to work on an entirely random basis.

Sheesh, luckily I had my trusty tomtom in the car!

 

[coolshades]

Dude! I wouldn't trust ANY GPS with my life. But to each his own. Both the GM Navigation in my Chevy C6Z06 and my Garmin have led me astray on more than one occasion. If "my life depended on GPS" I wouldn't be here. If you are placing your life in the hands of ANY GPS unit perhaps it is your head that needs to be extricated from your 4th point of contact.

Robo, did you use the "fix" here or simply use what came with the Vibrant? I take it your tests were only with satellites, and not with wireless network. Thanks for your info.

Cool

 

[robo21]

Robo, did you use the "fix" here or simply use what came with the Vibrant? I take it your tests were only with satellites, and not with wireless network. Thanks for your info.

Cool

No fix. Stock Vibrant. You're welcome.

 

[coolshades]

OK, here is my test result today. I cannot get any location if I left wireless data unchecked. When I checked it, Maps works instantly. GPS test shows 9-10 satellites under clear sky, but only 1-2 used. Those 1-2 used would appear green, orange, or red (they changed). The accuracy is almost the same as my factory installed gps that came with the car, except the Maps may be a little slow in recognizing that I did not follow its guidance. It is very accurate against my wife's iphone using Navigon.

My concern is that if satellites alone does not work, then when data become unavailable, the navigation would not work. Also, I thought maps would need wireless data to update the maps. Is that why satellites alone would not give me my location? Has someone tested the gps with commercial navigation software? Also, I wonder why GPS Test only showed 1-2 satellites used, while my wife's iphone using Navigon showed 4-5 locked?

Cool

 

[Robinelli]

Mine worked fine when I drove through areas with no cellular signal.

 

[coolshades]

Robinelli, were you using both satellites and wireless data, or just satellites? TIA for info.

Cool

 

[Robinelli]

I went on an 8 hour trip and had my GPS and cellular networks turned on. I lost my cellular connection for at least 2 hours and my GPS kept right on working. I don't know how, but it did. I drove back home a different way and had the same result again. I also drive in the mountains and I have no cell signal there either- maybe I should've gone with Verizon- and my GPS still works up there, too.

 

[Big-Dog]

I went on an 8 hour trip and had my GPS and cellular networks turned on. I lost my cellular connection for at least 2 hours and my GPS kept right on working. I don't know how, but it did. I drove back home a different way and had the same result again. I also drive in the mountains and I have no cell signal there either- maybe I should've gone with Verizon- and my GPS still works up there, too.

How were you getting maps? I was camping without signal and I could get sats to lock on, but since I didn't have a map, it was useless.

 

[wildblue]

If you're using Google Nav my understanding is it loads all the map segments for your route, so a connection isn't required once a route is set. What it can't do is re-calculate a route without a data connection. So, as long as you stay on the original route it should work fine.

For me this limitation makes Google Nav very unreliable. I've often missed turns or gone off the calculated route for numerous reasons when driving in strange cities. Being at risk of losing the re-calculate option at any moment is not a happy thought. It's made me seriously investigate 3rd party apps like CoPilot, which store the maps onboard. There are trade-offs either way, but I see this as a major flaw in Google Nav, and have told them so.

 

[krichek]

If you're using Google Nav my understanding is it loads all the map segments for your route, so a connection isn't required once a route is set. What it can't do is re-calculate a route without a data connection. So, as long as you stay on the original route it should work fine.

For me this limitation makes Google Nav very unreliable. I've often missed turns or gone off the calculated route for numerous reasons when driving in strange cities. Being at risk of losing the re-calculate option at any moment is not a happy thought. It's made me seriously investigate 3rd party apps like CoPilot, which store the maps onboard. There are trade-offs either way, but I see this as a major flaw in Google Nav, and have told them so.

It can recalculate from a missed turn etc. What it seems it can't do is recalculate if you go say 10 miles in a different direction and then try to get a new route without a data connection.

I've taken my old Droid up around Ft Bragg in Ca several times. The network service along the coast there is spotty to say the least. I've been on a backroad out in BFE with no cell service at all and still been re-routed when I would pull off to eat, get gas etc...

 

[wildblue]

It can recalculate from a missed turn etc. What it seems it can't do is recalculate if you go say 10 miles in a different direction and then try to get a new route without a data connection.

I've taken my old Droid up around Ft Bragg in Ca several times. The network service along the coast there is spotty to say the least. I've been on a backroad out in BFE with no cell service at all and still been re-routed when I would pull off to eat, get gas etc...

Interesting. I was using Google Nav on some local roads and made a turn off of the planned route. I noticed it was not re-calculating, and then realized my data connection had dropped. This lasted for a couple of minutes, then once the data connection was re-established the re-calc happened immediately.

When I got home I did some searching and found this statement:

Data Connection Loss - Google Mobile Help

• If you deviate from your route, you will not receive a reroute until you re-establish a data connection.

My experience, and this info was the basis for my post.

I was no where near 10 miles from my original route. Curious you've had a different experience. Maybe it's determined by the specific map segments that have been downloaded. If so, it's still hit or miss, which can make navigating unknown roads an "adventure" I'd prefer not to have.

 

[coolshades]

That was my concern with aGPS phones. I do not know exactly how GPS works, but I though 3 satellites are required to have any accurate positioning. I guess as long as I have wireless network data plus a couple of satellites locked, it will work just fine. I wonder if Robinelli had at least 3 satellites locked when he was driving without wireless data?

I hope this gps issue can and will be resolved by Samsung and Tmobile.

 

[SamsungVibrant]

That was my concern with aGPS phones. I do not know exactly how GPS works, but I though 3 satellites are required to have any accurate positioning. I guess as long as I have wireless network data plus a couple of satellites locked, it will work just fine. I wonder if Robinelli had at least 3 satellites locked when he was driving without wireless data?

I hope this gps issue can and will be resolved by Samsung and Tmobile.

You actually need a minimum of 4 satellites locked for an absolute precise three dimensional fix that includes height above earths surface, not 3 as others have been saying in this thread. 3 Satellites locked will only get you two dimensional position.

Thats why, every GPS I've ever used gets more than 4 satellites locked, except for the vibrant.

 

[SamsungVibrant]

This may sound really stupid because I have No programming experience whatsoever, but I know the broadcom chip shares cycles with the CPU. Perhaps there is a buffering overflow, CPU data exchange or memory problem when trying to use the GPS which is fixed with a reboot since it clears any data in the CPU cycle or the memory. Maybe someone can give a more accurate technical perspective but just a thought.

Chris, you aren't stupid, you are actually very smart. Some folks including developers have suggested that the Vibrant actually has a hardware flaw that has to do EXACTLY with what you stated! Buffer problems!

Most likely the hardware is flawed, Samsung will not be able to fix this, instead the firmware software update will just optimize GPS, to make the hardware flaw seem less apparent.

You can follow the buffer hardware problem theory at the xda forum.

 

[Robinelli]

That was my concern with aGPS phones. I do not know exactly how GPS works, but I though 3 satellites are required to have any accurate positioning. I guess as long as I have wireless network data plus a couple of satellites locked, it will work just fine. I wonder if Robinelli had at least 3 satellites locked when he was driving without wireless data?

I hope this gps issue can and will be resolved by Samsung and Tmobile.

If the # of satellites locked is correct on the phone it only had 2. It has never locked more than 2. However, it was accurate down to the second on telling me what turns to make and when to make them.

 

[SamsungVibrant]

Hey everyone, well it appears the Epic 4G is going to ship with the GPS "optimized," being very careful not to use the word "Fix." However, sadly, the accuracy is still horrible. Seems like the issue may actually be hardware in my opinion, and the software/firmware is just optimizing the GPS, and not actually fixing the GPS. Optimize is the same verbage samsung used.

What does this mean? Means that for all those people who want to use GPS on foot, hiking, getting around the city on foot, whatever, with horrible accuracy, you may be out of luck. Seems like we will be limited to using the GPS in the car, where accuracy is not as crucial as being on foot.

Also at 1:07 he talks about parallel roads. Some people argue, who cares about accuracy if it gets you from point A to point B. Well accuracy is very important, just like the guy in the video says, parallel roads can get you in trouble. I've posted my experience with poor accuracy before. When I was on the 405 near sepulveda, where the 101 runs parallel for a brief second to 405. My Vibrant actually jumped over and placed me on the 101, even though I was on the 405 the entire time. It got confused when the two freeways ran parallel for a second. The stupid GPS had me driving on 101 for 10 mins, even though I was on the 405 the entire time. It appears even when the GPS optimization is released, mistakes like this may continue, due to poor accuracy.

Then again, this is my opinion only, we won't truely know how well the update optimizes the GPS until it is truely released.

here is a clip from briefmobile.com
YouTube - Samsung Epic 4G for Sprint - GPS Test (vs. Captivate)

 

[robo21]

Hey everyone, well it appears the Epic 4G is going to ship with the GPS "optimized," being very careful not to use the word "Fix." However, sadly, the accuracy is still horrible. Seems like the issue may actually be hardware in my opinion, and the software/firmware is just optimizing the GPS, and not actually fixing the GPS. Optimize is the same verbage samsung used.

What does this mean? Means that for all those people who want to use GPS on foot, hiking, getting around the city on foot, whatever, with horrible accuracy, you may be out of luck. Seems like we will be limited to using the GPS in the car, where accuracy is not as crucial as being on foot.

Also at 1:07 he talks about parallel roads. Some people argue, who cares about accuracy if it gets you from point A to point B. Well accuracy is very important, just like the guy in the video says, parallel roads can get you in trouble. I've posted my experience with poor accuracy before. When I was on the 405 near sepulveda, where the 101 runs parallel for a brief second to 405. My Vibrant actually jumped over and placed me on the 101, even though I was on the 405 the entire time. It got confused when the two freeways ran parallel for a second. The stupid GPS had me driving on 101 for 10 mins, even though I was on the 405 the entire time. It appears even when the GPS optimization is released, mistakes like this may continue, due to poor accuracy.

Then again, this is my opinion only, we won't truely know how well the update optimizes the GPS until it is truely released.

Sammy, when I think of your posts on this subject, a few terms come to mind: Contentious, drama mongering, negative, alarmist, etc. Just look at the alarm you sound in the title to this post! "...we can forget about GPS on foot with our phones." Kind of a questionable claim given that the "fix" hasn't arrived yet. Kind of unfounded given that the review you cite as your "evidence" relates to a SINGLE unit.

What makes you think that the Epic 4G is any different than the Vibrant in unit to unit variation? We have seen that there are wildly different experiences not just on this forum but all forums are reporting that some people have no problem ACCURATELY navigating from point A to point B.

Whoever argued that "accuracy doesn't matter as long as it gets you from point A to point B?" Please name one person. I would argue that accuracy does, in fact, matter a great deal in gettting you from point A to point B. Also, while your Vibrant can't precisely locate your position, mine will pinpoint my location to within a few feet. And I've been saying this from the beginning.

Yes, we already know your Vibrant and many others are suffering from serious GPS problems. Somehow you seem to think by flooding the forum with posts about how bad the GPS is you will get a fix faster or won't be forgotten by Samsung. What happened to your statement earlier wherein you said this thread would get quiet until the fix appeared in September?

How can you draw valid conclusions from a sample of ONE GPS review of ONE Epic 4G? What would have happened if back in May I (or anyone else who has a Vibrant with working GPS) had posted a video review of my GPS results with the Vibrant? Could we conclude that all Vibrants would have working GPS? Of course not. Why not get some results that are statistically significant, as in wait for the release of the Epic 4G and then set up a poll over on the Epic forum? Even then keep in mind that many people who never have an issue DO NOT VISIT FORUMS. So I would question the results even then, but still they would certainly have more merit than a sample of ONE!

The optimization in the "optimized GPS" in the 4G probably is not the "fix" that is forthcoming. That may very well be still in the works and apply to all Galaxy S including the Epic 4G - we just don't know yet.

Calm down, relax and see what comes next month. If you still aren't satisfied by the end of next month you will be joined by lots of voices clamoring for better GPS, meanwhile take it easy.