Closed bernardsilenou closed 3 years ago
@bernardsilenou Tested it on my system by ugrading from 1.50.0 to 1.51.0 and had no problems with synchronization or the upgrade itself. Could you please verify it with another device (and ideally on another system as well)?
Regarding your first point, this was done on purpose to not disrupt users in their work. If a user is currently in an area with very bad internet connection, they might want to delay the app update. Getting the prompt to download and install the new update every time they synchronize would be very annoying, which is why it only appears on startup. We could however think about adding a "Check for updates" button to the settings that would allow users to manually trigger the update process without having to close the app.
@MateStrysewske Thanks for the information. I will test with the coming release. It is a good decision that we should not force the user to upgrade each time they sync. However it will be good to inform them of the new update as you suggested as long as the app is on. We can also add a warning message that good internet connectivity is needed to do the upgrade.
@MateStrysewske I did dug further by testing with another device with Fresh installation of 1,51.
@bernardsilenou Looks like the test server contains at least one additional test that does not have a sample_id assigned, which leads to a database insertion problem in the mobile app which then in turn leads to the synchronisation problem. I.e., there is corrupt data in the database. I don't know how that happened, but it's definitely not a general problem with the mobile app.
@MateStrysewske I got it. So what is the logic we need to put in place to correct this king of issues?
In a workshop with the Nigerian users last week, they mentioned synchronization challenges as their main challenge. Things like this may appear in their systems also.
How can i test further?
@bernardsilenou I guess we'll have to investigate how it's possible that an additional test is created that doesn't have a sample ID. I'll reopen this issue and relabel it accordingly so we can have a look at it in the current sprint.
@bernardsilenou I made a few adjustments that MIGHT have an impact on this although I think it's unlikely. I can't produce an additional test without a sample ID on my system. I'll close this issue for now.
@MateStrysewske Super, I would test with 1,53 and give you feedback
@fhauptmann @MateStrysewske @kwa20 We discussed to further investigate the cause of this issue in the first place
We still need to figure out, how it happened to corrupt the database. Can this happen on production environments, too?
I reopened because we want to keep track and test further
I got a dump of the database from test1401: I didn't find a sample without UUID.
@fhauptmann The problem is not that there's a sample without a UUID, but that there are additional tests without a sample_id.
This ticket was solved by @MateStrysewske, merged and closed to the released version
Bug Description
Steps to Reproduce
1. 2.
Expected Behavior
Screenshots
System Details
Additional Information