Tasker crash/restart during navigation and tasker zip files not recognized by Windows 8.1

[Steve Rothkin]

For the last couple of weeks I've been seeing spontaneous random crashes/restarts of the Tasker monitor while driving with Google Maps navigation. I've attached a log and system dump of one of these along with a backup of my tasker setup.

Also, the zip file was created using the tasker Zip command on the directory that the task copied the files too. The zip file is not readable by the built in Windows 8.1 zip file handling (but it is readable by WinRAR and the extracted files do match the source files).

[Pent]

For the last couple of weeks I've been seeing spontaneous random crashes/restarts of the Tasker monitor while driving with Google Maps navigation. I've attached a log and system dump of one of these along with a backup of my tasker setup.

OK, will look after 4.7 is out (soonish)

Also, the zip file was created using the tasker Zip command on the directory that the task copied the files too. The zip file is not readable by the built in Windows 8.1 zip file handling (but it is readable by WinRAR and the extracted files do match the source files).

Works also fine with linux unzip.

I looked at the java zip API but didn't find any options that would make it incompatible with anything.

No error message from windows ?

Pent

[Steve Rothkin]

Double clicking on the file in windows explorer shows an empty directory.

Right-clicking on the file and clicking Extract gives
The compressed (zipped) folder (name.zip) is invalid.

In WinRAR the info button shows version to extract = 2.0 and host OS is DOS.

[Steve Rothkin]

3 random crashes/restarts of monitor today while running navigation. I wonder if it's got something to do with cellular dead spots and play store license verification.

[Pent]

Tasker doesn't use cell data connection for a licence check.

Are these logs from 4.6u3 ?

Could you in any case update to 4.7 please if you havn't yet.

Thanks,

Pent

[Steve Rothkin]

On 6/9/2015 3:21 AM, Pent wrote:
> Tasker doesn't use cell data connection for a licence check.

I meant the Google API for license checking might. It's curious that
I've only noticed the restarts while navigating.

> Are these logs from 4.6u3 ?

Yes.

>
> Could you in any case update to 4.7 please if you havn't yet.

I noticed in the release notes that I need to go through my tasks to
manually reselect some actions (not happy about that -- Tasker updates
should not break existing code or require manual changes to them). So I
need to defer the upgrade until I have the time to go through everything.

[Pent]

manually reselect some actions (not happy about that -- Tasker updates
should not break existing code or require manual changes to them). So I
need to defer the upgrade until I have the time to go through everything.

Agree. It was a choice between

- reducing cases where data created in 4.7 was loadable into earlier versions
- users needing to possibly update some relatively rare parameters
- not adding the extra functionality

Someone was going to be unhappy whatever.

Pent

[Steve Rothkin]

On 6/9/2015 8:00 AM, Pent wrote:

manually reselect some actions (not happy about that -- Tasker updates
should not break existing code or require manual changes to them). So I
need to defer the upgrade until I have the time to go through everything.

Agree. It was a choice between

- reducing cases where data created in 4.7 was loadable into earlier versions

Why should loading data from a newer version into an older version even be a concern? From what I've seen in the past this usually didn't work anyway.

- users needing to possibly update some relatively rare parameters

Maybe I'm misunderstanding the release notes.  My impression is I need to search out all my tasks that use the array squash function, or test file system and manually fix those steps. And I'll need to use 2 devices in parallel since the upgraded probably one won't show what used to be selected. That's a major PITA.

This would be comparable to if I after an upgrade of SQL Server blanked out some portions of stored procedures in the database such I wold then have to look comb through all the stored procs and manually fix them without any report or other automated help to guide me to the specific procs/lines that needed to be changed and what they were before the upgrade. That's really bad.

The tasker update process should apply these changes for me automatically rather than require me to search them out one by one and manually fix them.

[Steve Rothkin]

Today with 4.7u1m it crashed 3 times while navigation was running -- twice in the morning and once on the way back in the afternoon.

The 1st and 3rd crashes were geographically close to each other (not more than a few miles apart) and not far from a location context area so maybe there's a problem relating to location contexts (though that wouldn't explain the 2nd crash).

[Pent]

It's our old nemesis, the connectivity change broadcast.

If you remember, I improved the efficiency of that piece of code and it got
better, perhaps you have now more profiles that listen for such broadcasts
(e.g. wifi connected) so it's become a problem again, or when something else is going on that
slows the UI thread or device in general it comes up again.

I will add some time stamping to the code that runs to try and find which part is taking so
long. If you can send me another log from 4.7u2 when it's out, that would be great.

Pent

[Pent]

But could you first try turning off logging (SD and system) and see if that
helps ?

Thanks,

Pent

[Christopher Dyrlch]

I can confirm that Android kills the Monitor while using Google Maps.

Since i'm currently building a logbook i also run into the issue, that Android kills all of my tasks and the Monitor, precisely while using some Heavy Load Tasks to save the Location. And it also happend with no Tasks running at all. I always thought that it was the Memory Leak Bug in Android 5 that caused it. But now i think that Maps is just using to much Memory

[Steve Rothkin]

FWIW, I just switched to a Nexus 6 (lollipop 5.1.1, Google Fi) and the
problem didn't happen in 3 hours of driving yesterday. Not sure if its
because of a faster processor, newer version of android, or being a
stock rom.

[Pent]

Well, pleased you don't have the problem any more, but shame we won't be able to fix it :-)

Pent