Discussion and testing on features and known&potential issues

[John Veness 57]

Very often, for me the Home button (middle circle) seems to act as if I pressed the Overview button (top/right square).

Also opening/closing the keyboard often brings up the Overview screen.

Anyone else seeing this?

Edited February 8, 2020 by John Veness

[EskeRahn 5,451]

1 hour ago, John Veness said:

My Sony Xperia Pro worked like this, so it's certainly doable in Android.

Oh certainly doable, but the question is how much tampering with the stock android that would be needed.

I think they now let Android do all the keyboard handling, only having made the mapping files to handle the key mapping, but no 'logic'

[divstar 164]

1 hour ago, Wikiwide said:

Australia, New South Wales, Sydney

Have already installed Sailfish OS on it. My phone seems to have unreasonably quiet speaker during phone calls (quieter than Geeksphone Keon); and when I switch on loud-speaker mode, the other party gets echo on their end of the line - but that's likely just my device having some mis-configuration. Music plays fine, loud and clear.

Thank you. Best wishes.

~~~~~~~~~~~~~~~~~

Per aspera ad astra...

It's a common bug from what I read on here (I've got it, too, when using WhatsApp and listening to voice messages via loudspeakers). I also heard a fix is in the making if I am not mistaken, as it probably is a software error - rather than a hardware one.

[lameboyadvance 98]

17 hours ago, MickH said:

https://f-droid.org/en/packages/com.jarsilio.android.waveup/

 

Quote

Known issues

Unfortunately, some smartphones let the CPU on while listening to the proximity sensor. This is called a wake lock and causes considerable battery drain. This isn't my fault and I cannot do anything to change this. Other phones will "go to sleep" when the screen is turned off while still listening to the proximity sensor. In this case, the battery drain is practically zero.

...Looks interesting, but before I use it does the Pro1 suffer from this issue at all?

[Craig 1,435]

17 hours ago, MickH said:

Was looking for an app that could activate the screen when opening the front cover

Probably best to have phone turned off while opening front cover anyway.

sorry, sarcastic comment cuz I can't imagine what front cover you're opening

[J_P 23]

On 2/6/2020 at 5:55 PM, kontakt said:

There is an unofficial list: https://docs.google.com/spreadsheets/d/15uE12Yv5nMvIF42U33cY5FQoF6M66QIn00cC876uf7Y/edit#gid=0

The first issue listed  - 'no sticky modifiers' is not fixed IMO as changing the keyboard type to ALPHA only makes shift sticky, and does not make the function key sticky.
I think we still want a proper fix for all the modifier keys. (Not sure if I should be editing the sheet.)

Also, 'speakers - left and right switched' - workaround: 'invert head'. 🤣

[EskeRahn 5,451]

2 hours ago, Craig said:

I can't imagine what front cover you're opening

Flip-case....

[EskeRahn 5,451]

20 hours ago, MickH said:

Was looking for an app that could activate the screen when opening the front cover (like on my XPeria X, which uses a magnetic sensor).
Found WaveUp on the F-Droid Store. The app uses the proximity sensor and works excellently in both portrait and landscape.

https://f-droid.org/en/packages/com.jarsilio.android.waveup/

Personally I would be more interested in the opposite: Something turning Off the screen (including touch sensibility) when the proximity sensor is activated. E.g. closing a flip-case or simply putting it in a pocket.

[EskeRahn 5,451]

1 hour ago, J_P said:

The first issue listed  - 'no sticky modifiers' is not fixed IMO as changing the keyboard type to ALPHA only makes shift sticky, and does not make the function key sticky.
I think we still want a proper fix for all the modifier keys.

Someone posted that Shift, Alt and Meta gets sticky, So IF they change type to Alpha a fix could be to change so the yellow arrow keys to send right&left Meta rather than Fn, and obviously adjust the key-maps to reflect this. Still Ctrl would not be sticky though.

[MickH 211]

1 hour ago, EskeRahn said:

Personally I would be more interested in the opposite: Something turning Off the screen (including touch sensibility) when the proximity sensor is activated. E.g. closing a flip-case or simply putting it in a pocket.

I pretty much think your prayers have been answered, as the app does work in that way, I'm sure. It does have a so called "Pocket Mode" which is described in the documentation. It doesn't behave like when your making a call. In that instance the phone isn't locked when you put the device to your ear and the developer has taken that on-board so WaveUp doesn't lock the phone while a call is in progress.
As for the aforementioned battery drain, I haven't noticed anything untoward.
You can set a xx second warning which gives a short vibrate xx seconds before the screen turns off and locks. You can feel it vibrate whenever you put your fingers too close to the proximity sensor when using the phone.

[MickH 211]

😕😕How come the discussion about the WaveUp app has jumped across from the "Apps that works great with a real keyboard and in landscape" thread?😕😕

[EskeRahn 5,451]

30 minutes ago, MickH said:

😕😕How come the discussion about the WaveUp app has jumped across from the "Apps that works great with a real keyboard and in landscape" thread?😕😕

Try read the initial post in the thread. It was never intended as a discussion threads on apps. And as this is actually a feature request that is fixed by the app, I found it more suitable in here.

That apart it did not seem to work that well on my unit, maybe it is because I fetched it from the Android Market, where it is 3.2.0, the version is slightly lower in your link 3.1.8.

It does offer a lock mode I did not try, but simply turning off the display would have been enough for me. It seems a little odd that it would need more privileges to turn off the screen than to turn it on. But never mind.

[Craig 1,435]

8 hours ago, J_P said:

The first issue listed  - 'no sticky modifiers' is not fixed IMO as changing the keyboard type to ALPHA only makes shift sticky, and does not make the function key sticky.
I think we still want a proper fix for all the modifier keys. (Not sure if I should be editing the sheet.)
 

I think/hope anyone is welcome to edit the sheet. 

I think we want a proper fix for everything listed.  Just cuz there's workarounds or partial workarounds for some issues, doesn't mean we don't want it fixed in stock, for best out of the box experience for new users and reviewers and stuff, to help it keep selling, to help the company succeed, etc....

 

[wilhelmfitzpatrick 47]

@david curious what speedtest app you are using?

[EskeRahn 5,451]

@david Speedtest works fine here https://www.speedtest.net/my-result/a/5736094493

[david 929]

37 minutes ago, EskeRahn said:

@david Speedtest works fine here https://www.speedtest.net/my-result/a/5736094493

Try the following:

- Turn on wifi and cellular

- Unplug phone from any charger

- Put phone in landscape orientation, with keyboard out

- Reboot phone by holding down the power button and selecting restart

- Immediately when android starts, launch speedtest

- The app should orient landscape and then go back to portrait

- Immediately after it goes to portrait, press the start circle

I am not saying all these variables are part of the issue, but I have been causing it to force reboot for the past 30 minutes, and haven't narrowed down all variables yet, so the above steps are the best way to replicate what I am doing.

It is definitely related to wifi and may be related to battery level.

Edited February 10, 2020 by david

[david 929]

40 minutes ago, wilhelmfitzpatrick said:

@david curious what speedtest app you are using?

https://play.google.com/store/apps/details?id=org.zwanoo.android.speedtest

[EskeRahn 5,451]

1 minute ago, david said:

- Immediately when android starts, launch speedtest

Can not do that, it request password on boot. But that things might fail if it is not finished booting, I really only see as an oddity, and not anything worth chasing, let alone use any resources fixing. But sure it ought not to happen, and if this is the only way we can reproduce the boot behaviour, the recipe might help them fix that.

[david 929]

6 minutes ago, EskeRahn said:

Can not do that, it request password on boot. But that things might fail if it is not finished booting, I really only see as an oddity, and not anything worth chasing, let alone use any resources fixing. But sure it ought not to happen, and if this is the only way we can reproduce the boot behaviour, the recipe might help them fix that.

Enter the pwd and launch after.  It happens even after waiting, but I am trying to give you the best chance of reproducing it.

I do have security disabled too, however.  Only mentioning it as a variable.

[EskeRahn 5,451]

6 minutes ago, david said:

Enter the pwd and launch after.  It happens even after waiting, but I am trying to give you the best chance of reproducing it.

I do have security disabled too, however.  Only mentioning it as a variable.

Tried your recipe both with and without autorotate, no problems....

A GUESS, as you have had other WiFi concerns, could it be that you are doing it while it is not completely finished negotiating with the WiFi router?

[david 929]

To be clear, it happens even after multiple minutes of being booted.  In looking at other reported crash/reboot situations, @Waxberry has wanted to see if wifi is involved, because he is trying to chase down some reboots possibly related to wifi.  This is a way to reproduce the problem so they can track down what is going wrong.  (assuming others can replicate...I can make it happen at will)

I will continue to narrow down the variables and update with the details.  Right now it is definitely wifi related and might also involve battery charge state.

[david 929]

8 minutes ago, EskeRahn said:

Tried your recipe both with and without autorotate, no problems....

A GUESS, as you have had other WiFi concerns, could it be that you are doing it while it is not completely finished negotiating with the WiFi router?

No.  As I said, it can be minutes after booting, with wifi completely up and running.  This isn't related to the ping latency issue.  Well, not unless it is somehow. lol

[david 929]

I will test this for another day and then do a factory reset to see if it resolves the issue.  If it does, then there is a good chance something got left over between OTA updates that the new firmware doesn't like.

If @Waxberry or anyone else wants any type of logs before I do that, let me know.  If a factory reset fixes the issue, then chance to learn from it will be lost.

[EskeRahn 5,451]

2 minutes ago, david said:

I will test this for another day and then do a factory reset to see if it resolves the issue.  If it does, then there is a good chance something got left over between OTA updates that the new firmware doesn't like.

If @Waxberry or anyone else wants any type of logs before I do that, let me know.  If a factory reset fixes the issue, then chance to learn from it will be lost.

Could well be. Or perhaps the rooting mechanism is not perfected and have some odd side effects?

[david 929]

1 hour ago, EskeRahn said:

Could well be. Or perhaps the rooting mechanism is not perfected and have some odd side effects?

Definitely possible.  The only semi unique things about my setup might be:

- Magisk root (no modules used and no changes made by me to any system files)

- fastboot was used to enable charging when phone is powered off

- I do not have boot security enabled

- I did do one factory reset on the 20191203 firmware that my phone shipped after unlocking the bootloader.  I then allowed the OTA update to happen to get me to the 20191210 firmware.  I then used Magisk manager to patch the boot image and gain root.  I then did the final OTA to bring me to 20200106.  I did undo root and redo it shortly afterward, on that same firmware.

- I have hundreds of apps from my old phone restored onto the Pro1 via Titanium Backup (no system backups/restores were done, however).  I can't use safe mode to test whether those apps are to blame, as I won't be able to run speedtest in safe mode.  I can use other tools to convert speedtest to a system app, but I don't want to modify my /system partition.  There does appear to be a magisk module to run a user app as a system app in a systemless way, so I might consider that as a way to isolate speedtest from all my other apps in safe mode, to rule out some other app causing this issue.  Alternatively, I can use TitaniumBackup to freeze or uninstall all my apps except speedtest and TitaniumBackup.

UPDATE:  I froze all my user apps except for speedtest and TitaniumBackup and the problem still exists.

Oh, and keyboard out/in and screen orientation don't seem to affect it.  It happens with it in portrait orientation with the keyboard in as well as in landscape with keyboard out.

Also, for a while, I could avoid the problem if I enabled airplane mode for a few seconds (which disables wifi) and then disabled airplane mode (if I did it too quickly, the problem would sometimes still happen).  That trick doesn't seem to work anymore.  I thought the behavior might have changed due to being low on battery, but even after I've charged the phone back up, and even though I'm plugged in, that work-around does not work anymore.  Only guess is that some state is being changed in something that the wifi uses when these reboots happen, and it is is now in a state where cycling the wifi off and on doesn't help anymore.  

The problem has not happened a single time when on cellular (with wifi disabled).  If the app is able to get to the stage where it starts transferring data for a test (even if the test is cancelled), then all other tests will work until the phone is rebooted.  For instance, if wifi is turned off, a test is run on cellular, and then wifi is turned back on, then it will work fine on wifi.  Or, if through some series of steps I do manage to get it to work on wifi (like the wifi restart method I was using for a while, that I mentioned above), then it will always work from then on, until rebooted.

CORRECTION!!:  It now happens on wifi *after* it has run on cellular.  Once again, some state somewhere has changed and that has changed the behavior.  In other words, it happens on wifi every time now, no matter if wifi has been cycled on and off or if the test has been run on cellular first.

NARROWING IT DOWN:  It apparently only happens on 2.4GHz, not 5GHz.  My router has 3 access points, 1 at 2.4 and 2 at 5.8.  I was on 2.4 the whole time today (the first day I've been on that frequency with my Pro1).  I just switched to each of the 5.8GHz connections and the problem didn't happen.  Switching back to 2.4GHz, after successful 5.8 GHz speed tests, caused it to crash right away again.  It crashes at the beginning of the test, when the app is in the "Connecting..." phase.

It does *NOT* happen with another router I have that has no authentication on its 2.4GHz access point.  The access point where it crashes has the authentication set to WPA2-Personal and encryption set to AES.  If I set it to no authentication and connect to it with the Pro1, the crash still happens.  So it appears to be a combination of the 2.4GHz band and certain routers.  I thought it might have something to do with which 802.11 protocol was being used, but I forced both routers to 802.11n and the problem still exists with only the one router.  The router where it doesn't happen is set for 20 MHz channel width.  The one where it is crashing was set to 20/40 MHz.  I changed that to 20 MHz and it still crashed.

If I force the router to use legacy protocols on the 2.4GHz access point, then the problem doesn't happen.  The link 54 Mbps, so it is either connecting at 802.11a or 802.11g.

So it has been narrowed down to 802.11n, on 2.4GHZ (didn't try 802.11n on 5GHz, only 802.11ac), from this particular router.  But what makes this router special vs the other router is that it has advanced features -- beamforming, etc.  So it could be one of those things that is being exercised in the phone, under certain conditions, that is causing it to crash.  It could merely be that it has a much better signal strength to the router it crashes with, hence has a higher link rate, hence is trying to send/receive data faster over 802.11n (but not as fast as it is able to send/receive with 802.11ac on the 5 GHz access points) to the problem router than the non-problem router.

Anyone wanting to try to reproduce this can try to connect to a 2.4GHz 802.11n access point.

For reference, my router is an ASUS RT-AC5300.

As I mentioned before, I have uninstalled and reinstalled the app several times.  The problem happens whether I allow the app the permissions it wants or if I don't give it those permissions.  

SPECIAL NOTE:  The boot process did hang once a the fxtec logo.  I let it go for several minutes, thinking maybe it had enabled the boot security again and was doing some sort of encryption, but it never came out of it and I had to force it to reboot by holding down the power button.  It is unknown whether the forced reboot from the crash caused this or if 1 out of N boots will result in a hung boot. I'm rebooting dozens of times, so I might uncover something like that more often than others.

Edited February 11, 2020 by david