• Replies 50 replies
  • Subscribers 21 subscribers
  • Views 7477 views
  • Users 0 members are here
Options
Related

ERA/Fenix6???

Alan.raceQs
over 2 years ago

For six months, I have had no ERA reports from my previous versions. None. Zilch. Nada!

On June 6 I released a new version(4.1.3)

Today I find 4 ERA reports for the new version.

This app version 4.1.3 has been

  • running since June 6
  • on over 300 watches globally
  • across 48 different products,
  • but has only reported crashes on 3 products: fenix6, Fenix6 S and fenix6 Pro,
  • at four code points,
  • all of which are within the highly trafficked regions of the app.
  • There is far too little detail in the crash report to enable any debugging.

device

Occurrences

first

last

fenix6

5

2022-06-24

2022-06-24

fenix6 S

1

2022-06-24

2022-06-24

fenix6

4

2022-06-28

2022-06-29

Fenix6 Pro

1

2022-06-30

2022-06-29

As a mathematician, but not an epidemiologist, I find this result highly questionable.

Can anyone offer a similar experience? 

Is there an incompatibility between the F6 and ERA? 

  • over 2 years ago in reply to FlowState

    OK, thanks - a Forum feature I was unaware of.

  • over 2 years ago in reply to flocsy🤠

    How? Whose policy? As a rough estimate each of those dots covers about 1,000 square kilometres, so they certainly don't identify individuals.

  • over 2 years ago in reply to Alan.raceQs

    makeWebRequest does such error

  • over 2 years ago in reply to Alan.raceQs

    The GDPR in Europe is very strict and prosecuted with high penalties. I think this is why people are very sensitive about it. I can see your wish to get some statistics about the use of your app, and I think it's OK as long as you have informed the user that his location is stored on your server, but if I was you I'd have them explicitly consent to that.

    Sorry for the off-topic.

    On-topic: I agree with _psx_ that your problem might lie in the background process. Since it crashes at different lines in the code (all of which are executed on a regular basis), it's probably a specific line in the background. Maybe even only with iPhones or Android, so really hard to debug!

    The question would be, what has changed for the Fenix 6? 

  • over 2 years ago in reply to BobMiles

    This error is easy to find in dev space but in system is more difficult because you have to find all function with different number of parameter like onRecive(....) - two version.

    And first look into jungle because you have mentioned you put a lot of annotation so for f6 family can be something special.

    And one more thing, resign from type checking maybe compiler choose bad function symbol.

  • over 2 years ago in reply to _psx_

    Can you explain your last sentence? This in monkey c, not java or some other language with function overloading... How can type checking cause the wrong function to be called?

  • over 2 years ago in reply to flocsy🤠
    _psx_ said:
    And one more thing, resign from type checking maybe compiler choose bad function symbol.
    flocsy🤠 said:
    Can you explain your last sentence? This in monkey c, not java or some other language with function overloading... How can type checking cause the wrong function to be called?

    I think the issue is sometimes the type checker will check a different symbol than the one which is actually called at runtime, as in this bug report:

    forums.garmin.com/.../the-type-checker-s-symbol-lookup-doesn-t-match-the-runtime

  • over 2 years ago in reply to flocsy🤠

    Overloading is impossible strict by you can use if/case like system does in case of webRequest.

    In compilation time all names are changed into symbol so maybe compiler choose bad one because think that any parameter is unnecessary. Of course I'm not sure but if description of error is good it's problem with number of argument when you/system call function.

  • over 2 years ago in reply to BobMiles
    BobMiles said:
    but if I was you I'd have them explicitly consent to that.

    I am relying on the fact that Garmin provides a link to the developer's privacy policy on the download page, and that I am complying with my published privacy policy. Full Stop.

    BobMiles said:
    your problem might lie in the background process

    There is no "background process" - and certainly nothing to do with Android or iPhone. 

    BobMiles said:
    The question would be, what has changed for the Fenix 6? 

    Yes, yes, that's the question I'm asking, but what's the answer?

  • over 2 years ago in reply to Alan.raceQs

    Sorry I wasn't of much help! I thought makeWebRequest is executed in the background and the result comes back at a variable point in time.