Some information on how fitbit test watch faces after submission (request)

[TKCMuzzer]

Hello fitbit devs. It is possible you could give an insight on how you test watch faces and apps. Info like,

Android or IOS, which browsers you use, is it an actual developer who tests or do you just use testers. Not super detailed, just an outline, i'l explain why.

There is no doubt as time has gone on the review process has got longer on fitbits end. Now, i personally test my watch faces over a five day period. I install them on a Versa and an Ionic, I also install them using Android and IOS. This is to cover any issues during the review process.

It seems recently that the fitbit review team are finding issues that for the life me I can not recreate whatever equipment I am using. When my latest got rejected, i instantly checked the issue they brought up and guess what, everything was fine. So, I went away and tested it again on all devices, all browsers over a 3 day period and no issues whatsoever. I resubmitted and again today a different issue to do with the credentials I supplied not working. So, I jumped straight on a browser and tested the credentials and they worked fine, there has been no issues with the website etc.

This is why I am requesting extra information as to recreate the issue, people like myself developing need to know what is involved in the test environment as the whole process is becoming laborious due to the 7 day turnaround after resubmitting. This will ultimately start to turn people away as its pretty much impossible to solve an issue that can not be recreated without all the information.

(Please note: Fitbit have already cleared one of my watch faces using the exact same credentials, so its even more baffling)

Any help will be much appreciated. 

[rc26]

This is what I do:

• Initial testing is done in the simulator as the code is written. PC, Mac using Chrome browser mainly.
• More testing is done using the developer mode / bridge with phone and watch to observe console output...look for errors.
• Then build and post private build to gallery, send links to family phones with Fitbit watches paired and test the installation and operation there. We're all Android now, so dusting off an old iPhone and pairing a watch to it might be worth it. Usually I don't. Only if one of my group members points out an anomaly.
• Start a Facebook group for the project and build a community of testers/users and encourage feedback. They'll find every bug and issue.

[johnboy00]

I don't know how FitBit tests clock faces, but I sure wish they would do it more quickly.

As for how I test, I use the simulator to test on all devices, then I test on my Versa.

[TKCMuzzer]

So, bit of an update. As my latest watch face got rejected I did get more info on the issue. I was not surprised that one of them was Android related. This was my fear, the Android install process can be very inconsistent as in my testing can require many wipes of the fitbit app cache in order to get the watch faces installed properly. This means that issue is not with the watch face itself but primarily the app and its connection to Android phones.

I assume this will produce random issues in testing were one watch face works and the other doesn't. I have only been able to replicate the issue fitbit highlighted once and that was whilst using the Android app. I cleared the cache tried again and the watch face installed fine. 

This happened on my last watch face, I resubmitted and it worked even though I altered nothing in the code. 

I'm not surprised the issues raised where using an Android device. Unfortunately I can not effect the fitbit app and its connection with the Android system so 50% of my watch faces will probably have to just be resubmitted just because of Android issues ( I test them many times on my Android device before hand). Very frustrating. 

[JonFitbit]

The review team test on iOS and Android, for each supported Fitbit device. The queue size, testing complexity, number of devices, and the quantity of apps/clocks being submitted are all factors in the time taken for reviews.

There are indeed sometimes problems which manifest for iOS or Android specifically and the review team aren't there to debug issues which may or may not be caused by the platform, or a developer's code. If you're having specific issues like the one you described, please reach out to me on Discord or here, and I'll try to help diagnose. Then we can log support tickets for the iOS and Android teams to help identify any platform discrepancies.

Thanks for the feedback, and persistence!