Read on Twitter
There is a potential large-scale bug in the Scottish #ProtectScotland COVID app. No further updates from #TestandProtect since 30 Dec. Let me summarise in a thread what we know so far, why this is very worrying and why it must be taken seriously. 1/
Five days ago I wrote about a potential problem with the #ProtectScotland Android app: it does not seem to sync with the COVID database unless users explicitly open it. The result: potentially many users may be unaware they have been exposed to COVID https://twitter.com/migorinova/status/1343971813387337732
2/
2/
Thanks to people who helped amplify the post, #TestandProtect and @scotgov responded. The TLDR of the statement:
1) More than 1750 exposure notifications have been sent over 7 days
2) They believe it affects only select # users
3/
1) More than 1750 exposure notifications have been sent over 7 days
2) They believe it affects only select # users
3/
The full statement can be found in this article by @ChrisMusson https://www.thescottishsun.co.uk/news/scottish-news/6468440/protect-scotland-app-covid-coronavirus-snp/ 4/
Re 1) Assume that there is an equal number of Android vs iPhone users (should be a reasonable estimate, e.g. see https://www.statista.com/statistics/262179/market-share-held-by-mobile-operating-systems-in-the-united-kingdom/) 5/
Assume also that about 10% of Android users keep their COVID app open, or re-open it often (this is an unreliable estimate based on what people have shared with me: only 2 out of 30 people who have shred information about the app with me would open it regularly) 6/
If what we observe really is a large-scale bug, it means that the 1750 exposure notifications reported by @scotgov were generated by the data checked against iPhones and 10% of Android phones only. In other words, it is reasonable to assume the notifications should have been 3182
That is, there could be, potentially 1432 people in Scotland who were *not* notified they should self-isolate in the week leading to New Years.
Need to stress this is a *very rough* estimate due to lack of reliable data. It also doesn't take into account other tracing method 8/
Need to stress this is a *very rough* estimate due to lack of reliable data. It also doesn't take into account other tracing method 8/
But hopefully it puts things into perspective. There could be 1000+ people who should self-isolate but don't know it. With cases raising like mad, we can't afford this, and I hope #TestandProtect, @scotgov, and the developer http://nearform.com are taking this seriously 9/
I'm writing all of this, because of @scotgov's statement that they believe this is affecting only a select number of users: this statement is worrying. It is clear that there is a problem. I'm not aware of a single positive example of the app working as expected on Android. 10/
Please @scotgov, @scotgovhealth, @NicolaSturgeon, push the developer to fix this ASAP. And tell people to open their app so that it syncs! If only 10% of 1432 exposed ppl develop COVID, and the R-rate is 1, notifying them to self-isolate means ~143 direct cases prevented! 11/11
Oh and btw... Disabling the battery optimization as the gov's statement suggests, didn't do anything for me....... So everyone, just keep opening your app once daily 
