Please add the used protocol ( ipv4/ipv6) to recordings and later to the filters statistics and opendata

[grzchr15]

https://www.netztest.at/doc/#_measurement_data

The following information is transmitted to the Control Server for each completed test:

• Latency
• Uplink data rate
• Downlink data rate
• Test UUID
• Date and time

• The number n of concurrent TCP connections actually used

  (optionally) the individual tuples (t(j)k,b(j)k)(tk(j),bk(j)) measured during uplink and downlink measurement. In order to keep the data size small, not all tuples might be transmitted. In these cases an exact recalculation of the measured data rates based on the transmitted tuples might not possible.

Note	Further parameters might be collected and transmitted additionally to these information. These additional parameters are out of scope for this specification.

Maybe the used protocol ( IPv4/Ipv6) can be added to statistics and so shown which ISP or Mobile ISP uses already IPv6 in Austria and to which quality

[cproof]

Thank you for raising this issue.

While there is no "ipv4/ipv6" field in Open Data, the anonymized IP is contained in the Open Data files (see https://www.netztest.at/en/Opendata for documentation), where it can easily be inferred if a test took place over IPv4 or IPv6. For some tests, also the connection type is also recorded, showing if network adress translation took place (e.g. nat_local_to_public_ipv4)

We found that generating rollout statistics in general based on the type of connection used for testing does not yield correct results, as clients/Browsers may use IPv4 for conducting the speed test even though IPv6 is available.

[grzchr15]

Browsers may use IPv4 for conducting the speed test even though IPv6 is available.

That should only happen when happy eyeball fails on ipv6.

And that would be the missing open data anyway.