boomerbubba
Android Enthusiast
It turns out that the Epic apparently does not even use the same GPS chip as its GSM cousins. See details at my post at xda.
-
After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.
All Things GPS
- Thread starter SamsungVibrant
- Start date
-
- Tags
- samsung epic 4g
Seddy
Android Enthusiast
I did run the test while in the house. I haven't ran one while outside yet but I will do that tomorrow before I go into the workplace. Like I said, I didn't know if my indoor results were bad or good, but now that I look closely at the other results... I see theres plenty of room for improvement.
Yeah, the only consistency has been the moderate to slight attenuation problems which makes sense as the reception hardware is identical.
The difference in the decoding hardware could be why we see the main problems manifested differently.
I see you note you are with Sprint ATS, just so you know Sprint ADP (the top device group) has found a number of GPS issues and say it is at the tip of their problem list.
And why on earth is ATS telling people "use wireless networks" has anything at all to do with GPS when they are unrelated?
Seddy
Android Enthusiast
I dont know what the other ATS agents are telling people. I haven't gotten any calls regarding the Epic and GPS as of yet... most of my calls deal with cheap ass Palm Pre's and Pixi's and syncing Blackberries with Outlook... oh.. and those damn connection cards. (excuse my french). You also have to understand, there are about 15 different call centers with ATS agents that work inside of them. They are run by different companies and have different policies, therefor the information you get from one agent may be completely different from the info you get from another. I get calls about mis-information given to them from other agents, and they all think we all know eachother and we're all in the same building. LOL
This is what I am finding also. I start a GPS app with GPS off, go into settings to turn it on, pop back to the GPS app and almost always get a lock within a few seconds, sometimes before I can even back-button to the app screen. Moreover, though the accuracy seems to be hard-coded at 98ft/30m, it seems to me to be much more accurate than that, maybe no more than 15-20 feet typically.
I have had three GPS-capable phones before, a Moto Q, a Q9c, and an HTC TP2. This phone's GPS performance is as good as any of them in accuracy and better than all of them in TTFF.
You are cycling the gps on and off multiple times and that may defeat the cache bugs. The bugs we are talking about are widely observed.
Also the latest update tweaks the use wireless networks, it has no no affect on aGPS.
In terms of why the Epic performs worse than other smartphones on GPS, and way worse than Touch Pro 2, agianst which I did extensive testing, it mainly comes down to Time to First Fix being skewed by the cache bug which Sprint now is working on.
that is why applying the cold start workaround has such a positive impact on users who use GPS normally (don't unnecessarily cycle like you do)
That link was there and credit was given when the post originally published. It was never edited, nor added. I know that, because I wrote it. If you'd like to take issue, you're free to, but I'm completely comfortable with the credit as it stood since I felt that there was a team effort among the users on this forum to identify the symptoms and possible fixes as I followed it along through the postings.
I don't know why you didn't care to actually read through the article and find the link yourself originally and it doesn't matter, but you were wrong.
I don't intentionally cycle GPS; it's just that I keep it off until I need it, and usually start the app before I start it, since I know the app will just take me to the settings page when I start it if GPS isn't active at that time.
When you say the Epic performs way worse than the TP2, do you mean in terms of accuracy, TTFF, or both? My TP2 locates to about the same accuracy as my Epic but is way slower to get a fix, especially the Sprint Nav/Telenav app, which will stall literally for minutes unless I run the ephemeris update utility every time I start it.
I'm not calling into question your results; you seem to have done a lot more research on this specific issue and to have a lot more knowledge about GPS in general than I.
I only chimed in on the thread because I'd heard about the GPS issues in this phone and its siblings, and I just haven't seen them in my everyday use at all.
In my particular experience, the phone's GPS performance is quite good, better overall than any of my three previous phones.
Stringyquark
Member
I'm currently in denver coming home from california. Great vacation! For all my complaining earlier my gps worked pretty well. Got a decent lock while in flight read out 500 mph once we reached altitude.was able to get a pretty good lock in downtown san francisco. even with the tall buildings. Still needs improvement but works for my basic purposes. Like I'm drink how do we get home.
Epicurean
Android Enthusiast
How did you get a lock? Were you in airplane mode? I have not been able to get a lock in airplane mode - figured it was just part of the bugs.
Stringyquark
Member
Got my lock before turning on airplane mode. I haven't been able to get a lock as a standalone unless it already got all the info, but once you get a good lock then it's fine for awhile.
There are some making observations/claims on standalone function, but the people offering them are too long time posters and/or their language doesn't indicate full understanding of testing needed.
I have not been out of the country since getting the phone, and really we need someone to do couple of clear sky one hour attempts outside of cdma and no wifi (and with our cold start workaround disabled of course).
I use TomTom and google maps on Touch Pro 2, rock solid. Never used sprint maps/nav on my TP@ as it is on a sero. I do feel, and many others here note, that sprint nav is kinda slow and hinky on the Epic.
As far as your other observations I am not challenging them, but rather pointing out that the main bug is heavily influenced by user habit. We know that the proper method of ejecting and refreshing cache is not there, but that other actions may do this, some very often, but with cold start workaround doing it every time.
Let's face it, Engadget dropped the ball and did a disservice to their readers. That they are feeling defensive and don't want to give credit is their business.
Their current article mentioning the workaround still contains errors which they refuse to fix, misinforming their readers once again.
EG How on earth is engadget, a tech blog/site, still passing off Samsung's conflation of unrelated*, "Use wireless networks" as "agps". It makes them sound like a mouthpiece for damage control instead of a serious outlet.
from engadet:
really? Where pray tell is the user able to turn on and off "network assisted GPS" from any normal user menu on Epic? Answer: NOWHERE
How is engadget even printing an answer they know to be spurious? Not subjectively questionable but non existent on any menu on Epic?
Stringyquark
Member
My apologies, standalone isn't the word I should have used. I was referring to running gps while in airplane mode. I got a lock and a number of sats locked before switching into airplane mode. I still haven't tried bringing my phone into the back 40 in a nice big dead zone to see if it will lock, though while driving from san francisco to silicon vally where I data roamed it worked fine but had all ready hooked up and was running. (good 4g in the vally by the way) as for every thing I know about gps, it isn't much except what I've learned form wikipedia or aero1.
Yeah, there's a checkbox that says "use wireless networks." I don't know why you'd decide to not know that's what is being referred to or not, it doesn't matter. Now you can quote the part of the article where I disagreed with their assertion. I'll wait.
It matters, it matters a lot, since it has nothing to do with aGPS, it is not an assist at all. It is an alternate method that uses no satellites. Sorry but engadget got p0wned by Samsung throughout this entire issue. When you got that response from Samsung, why would you, if you are a serious publication not follow up and tell them you are not going to print an objective falsehood?
After the "cold start" fix I hadn't had any issues aside from the fixed 30m accuracy. Today, however, on my way to work I started the Google navigation map to check on traffic and plugged my phone in, and could not get a lock for the first time ever. I have an hour commute and usually don't plug my phone in, today I did, using the "official" sprint charger and noticed that as soon as the led light turned blue on the phone, the GPS immediately got a lock on, so it appears the sprint official charger causes too much interference and it results in further reduced snr 
I came from an Evo, which had near-instant GPS locking all the time. The Epic works sometimes, but quite often it takes forever to get a lock, and Google Maps will just say my location is unavailable.
I have tried both with and without the cold start fix. I'm wondering if anybody has had their GPS work as well as the Evo's had? Perhaps if I swapped my Epic for a different one, it would work better.
I have tried both with and without the cold start fix. I'm wondering if anybody has had their GPS work as well as the Evo's had? Perhaps if I swapped my Epic for a different one, it would work better.
boomerbubba
Android Enthusiast
Every Epic has the same GPS firmware bugs. They just manifest themselves differently according to the usage pattern. There is no evidence of difference among units, which makes sense because they all share the same firmware. Officially, Spint and Samsung remain in denial about the bugs.
dblankenburg
Lurker
I use my Epic's GPS on a daily basis and have lightning fast lock and no problems with any GPS based apps I've tried so far.
boomerbubba
Android Enthusiast
Intensive daily GPS use is one of the usage patterns that avoids the Epic's GPS bug, which is caused by obsolete cache data not being refreshed properly after a period of inactivity. It is time-sensistive. That does not mean your GPS firmware is not buggy. Nor does it mean your GPS hardware unit is superior. Your usage has just not triggered the bug.
So your anecdotal report is just another confirmation of the previously documented properties of the bug.