A few experiences from my old Xperia devices running official (fully Jolla-made) SFOS images (maybe it’s still relevant for the JP2026, but better take it with a grain of salt): Sometimes my devices don’t get a fix at first attempt, even when waiting for a very long time under the open sky with GPS Info permanently running and showing a non-zero satellite count. This usually happens when I have not used positioning since a long time (missing or outdated almanac data I suppose). To get an initial fix, I usually try out the things below. Once the device has successfully acquired its position, it has no problems reacquiring it when I temporarily close all GPS-consuming apps (location icon in status bar disappears). Probably relevant: I have configured Location to “Device-only mode” in Settings.
Close GPS Info > Wait for the pulsating location icon in the status bar to disappear > Wait a few more seconds > Restart GPS Info
Close GPS Info > Wait for the pulsating location icon in the status bar to disappear > Disable Location from the top menu > Wait a few more seconds > Re-enable Location from the top menu > Restart GPS Info
The location icon in the status bar should disappear immediately when the last app which accesses positioning data is closed. Sometimes this isn’t the case. The icon either disappears with a delay or not at all. Maybe an instable background process in the OS?
After it successfully got a fix, the device maintains it in a stable manner. However, if I enter a building while GPS Info or some navigation app is running and the device therefore loses the signals from the satellites, it can happen that I don’t get a fix anymore, even when I was inside the building for just a few minutes. If I close all GPS-consuming apps before I go in, it’s no problem however. I also get a fix again afterwards.
My best guess: If the device is still downloading almanac data from the satellites, but gets interrupted in the process (i.e. loses the connection or user closes all GPS-consuming apps), it is unable to properly resume the download. Instead, it tries to use the incomplete dataset, which will fail of course, even though it sees plenty of satellites.
Question to the power users: Can the entire almanac data be deleted somehow to force the device to completely download it again from scratch?
The following should be taken with another grain of salt, because it could have been pure coincidence as well:
Even though I have set location acquisition to “Device-only mode”, I once needed to disable Flight mode to get a fix.
Another time I even needed to shortly switch on mobile data to get a fix.
My feeling as a layperson: Maybe the routines for acquiring, proper download-resuming, checking for completeness, checking for errors, and updating of the almanac data should be investigated.
This should not matter even in car you can drive into tunnel for few minutes and then get fix shortly after you get out. Xperia10ii xperia10v and jp2026 here
And that would also be my interpretation, and – as I understand it – @Moripeluka’s view as well. The second value is the value of the almanac data that he was supposed to see. But that’s just a guess, not a fact.
After connecting Location, I get “0/29” nearly immediately. Per visible timestamps, it took 3 seconds to find my exact location (I still had the “screenshot” notification displayed), thanks to beaconDB.
After 10 minutes passed, I am at “0/41”, I guess satellites are moving in and out of my horizon so I should see 41. I am still at 0 used because I am inside.
I get close to a window. Within a second, it uses 23 satellites:
If I get far from the window, the number decreases from 23 to a low number within seconds.
I am back at my chair and it still reports 6/43. It still can detect a few satellites now that it saw them at the window, and a weak signal can still be received. This part is a bit surprising to me, why I wouldn’t get the 6 satellites earlier if the signal can be received even at my chair inside.
I can see in the 4th image you added in the first message that there is signal strength for at least one satellite so it can see that. The yellow bar after the satellites in the list in CSD app. How many of those with a yellow bar do you see when you are outside?
Yes, in perfect weather – hardly any clouds at all.
I repeated the test today with the new OS version under heavy cloud cover and with just a few small raindrops.
This time, it managed to connect up to 5 out of 45 satellites – but only briefly.
Even when I held my hand over the phone to take a screenshot, they were gone again. Most of the time, the count was 0 out of 40+. If it doesn’t rain, I’ll run another test tomorrow with both Sailfish phones.
Admittedly, the weather was bad, but navigating was virtually impossible despite the satellite signal we were receiving. The signal was too weak and there where not much satellite.
Short interruptions of, say, 5 or 10 minutes are not a problem IIRC, but longer interruptions of, say, 20 or 30 minutes are.
I never tested anything GPS-related in a scientific way. It’s just that sometimes in daily life I go like “What the heck? It was working nicely just before.”, without having an immediate link to some root cause.
Xperia X and Xperia XA2 Plus in my case. The XA2 Plus might already suffer from the alleged antenna hardware problems reported in various threads on this forum. For the brave ones: An apparent fix for this issue with lots of technical background information can be found here, but I didn’t try it so far.
The GPS in the Xperia X still works reliably - once you manage to get an initial fix and from there on avoid too long satellite signal interruptions (at least in Device-only mode, which I prefer over all other modes).
What often helps for me: Apps like Jolla Camera and IIRC even Pure Maps (when currently not navigating) stop accessing GPS data when they are minimized or when the screen is turned off. GPS Info on the other hand continuously keeps fetching data, thus keeping the relevant subsystems (almanac download, etc.) active. So when I need positioning data, I always start GPS Info first, wait for the initial fix, and only then start the actual productive app. As a “keep alive” measure, I often keep GPS Info running in the background.
EDIT: From the linked repair topic:
So this almanac (and ephemeris) datasets have different “expiration dates”. Maybe the respective subsystems in Sailfish OS just handle these validity periods in a wrong or ineffective way?
Not sure the almanac/agps/beacondb discussion etc has anything to do with it, a properly working gps chip in J2 has graph like this in gpsinfo within seconds, seems like a barely working chip that can get some signal sneaked through QA, but waaaay off the standard and should fall under warranty
The standard high-accuracy positioning option in settings, again online sources should not inhibit getting signal strength reported from sats, they help speed up calculations as data transfer straight from sats is slow, barely any sats displaying signal strength seems like a borked gps chip
Factory reset and comparison with the XA2+:
Settings: Mobile Data: on, GPS: on, GPS setting: high-accuracy positioning.
J2: Test duration approx. 1 hour, XA2+: Test duration 15 minutes.
Although the J2 acquired a fix after 20 seconds, it took 10 minutes before more than 2 satellites were available for positioning. After those 10 minutes, I saw for the first time a period during which it could have navigated. For 3–4 minutes, it was able to maintain a continuous connection to between 3 and 5 satellites. The rest of the time, the device no longer had sufficient connections. Sometimes one satellite was green, sometimes two, then just one again. I must have switched the mobile on at least five times during that hour. The result was always 0/approx. 45.
The next Image are 20 Sek after start after the first fix.
