janpf / ctt Goto Github PK

I noticed that keyCount in iOS log is diverging from the keyCount in the filehashes.json file, with 2020-10-27 being different. The difference is in the order of about 100 users. After investigating I found that it is due to cwa querying the files from the endpoint https://svc90.main.px.t-online.de/version/v1/diagnosis-keys/country/EUR/ instead of previously DE instead of EUR. Also the computed hashes from iOS match with those of the EUR files explaining the issue we had last week.

This raises two points:

• for your dashboard: do you want to keep it displaying the uploaded keys from german user or do you want to switch to european users, consistent with the CWA behaviour.
• for the purposes of ena_log, I would be keen on having the counts updated in the filehashes.json file to those of the files from EUR endpoint.

Thanks and best regards
Felix

I hope this question is appropriate for the project. I have read that keys are only valid for a given day. Looking at the stats, I notice increasing published keys, but from 9/23 an strongly decreasing number of valid keys. Does that mean keys have been delivered to late (aka "over-stress at health agencies/labs")? I think this discrepancy is important to understand the value of CoronaWarnApp.

Thx!

On the 22nd August something weird is happening again:

@micb25's graph is showing 86 users, while your graph is showing 66 and the output of @mh-'s parser shows "70 users" with some unparsed keys and a warning regarding the multiplier. This might be related to corona-warn-app/cwa-server#723, not sure tbh 🤷.

It seems that the data for 2020-08-18 is not quite right:

• https://ctt.pfstr.de/users/2020-08-18.txt shows a detected padding number of 1, resulting in 379 users
• The graphs for number of users and number of keys show 98 users and 952 keys, which doesn't match the numbers from approx. users file
• The dashboard from @micb25 https://micb25.github.io/dka/ shows 75 users, 749 keys and a padding of 5 for 2020-08-18, which seems to be about right.

Is there a problem in the downloaded source data? If so, why does https://micb25.github.io/dka/ show more reasonable values?

I am about to create a small application that can match log files of the exposure notification framework and display a nicer report than what Apple/Google are providing. In this report, I am displaying the dates of the diagnosis key file for matches.

As the log file of the exposure notification framework just has the hash of the file inside, I would need a mapping hash -> file. This is currently hardcoded.

Since you are already maintaining all keys for download and a list, I wondered if you could add a generation of a json object with such a mapping.
Python stub to get the hex representation of a diagnosis key file is as follows:

zip_data = zipfile.ZipFile(diagnosis_key_file)
infolist = zip_data.infolist()
with zip_data.open([f for f in infolist if '.bin' in f.filename][0], 'r') as bin_file:
    hash = hashlib.sha256(bin_file.read()).hexdigest()

Thanks in advance for consideration and best regards
Felix

in the long run I will need a new plotting library with the following main improvement points:

1. better interactivity
2. easier to implement interactivity
3. easier to display all data of a hovered datapoint/date
4. synchronize interactivity between graphs? (this is possible with the current vega lite usage, but not done yet. See point 2.)
5. ??

Happy new year! 🎉
With first data for the 2021-01-01 the charts started plotting the 2021 data as if it was for 2020, disregarding the year:

TL;DR: Does anybody have the collection of hourly files for the last few days?

Explanation why I'm missing those: my script didn't catch them, as I'm not archiving the hourly keys daily and the server stopped providing the hourly files for all days but yesterday:

available dates:
svc90.main.px.t-online.de/version/v1/diagnosis-keys/country/DE/date

hourly files available for yesterday (all hourly files available):
svc90.main.px.t-online.de/version/v1/diagnosis-keys/country/DE/date/2021-01-29/hour

the day before (no longer available):
svc90.main.px.t-online.de/version/v1/diagnosis-keys/country/DE/date/2021-01-28/hour

Thanks :)

Can you explain what valid keys means in the graphs? What would be invalid keys?