Jittered violin plots on nrc.nl (https://www.nrc.nl/nieuws/2020/04/27/waarom-rutte-zo-voorzichtig-is-de-epidemie-laat-zich-lastig-meten-a3996722): Would they have chosen these because they are much more intuitive than boxplots?
I tested this using the invalid prefixed announced by cloudflare (https://blog.cloudflare.com/is-bgp-safe-yet-rpki-routing-security-initiative/) and did pings from the consumer networks using RIPE Atlas (results: https://atlas.ripe.net/measurements/24907077/#!probes)
I just checked what Dutch ISPs use RPKI validation (https://isbgpsafeyet.com/) and was amazed: Some do (XS4ALL, SURFnet), those of the majority of users do not (e.g. KPN, Ziggo, T-Mobile Thuis, …)
I made a quick chart based on stichting NICE data (with pandas/altair/jupyter lab) that I could not find for the Netherlands: cumulative outcome after n days.
…but there is much more potential for visualizations.
Just found out [parts of] the dutch government are working on a shared user experience structure and will likely open it up. Sounds (re-)usable! https://designsystem.gebruikercentraal.nl
A real world example: properly salted hashes that leak (https://www.rtlnieuws.nl/tech/artikel/5095321/covid19-alert-datalek-crypto-digibyte-coronavirus-ministerie-app) can be leaked if users have bad passwords
After 3 clicks I'm no longer sure open-source would give me more confidence in the apps…
It feels very ad hoc. When the adoption rate of a system is essential, trust in the systems implementation is crucial. This concept deserves better/a do-over, such as called for by https://twitter.com/WilmaH/status/1251137532751941633 et al
I thought I heard the COVID-19 tracking apps in the Netherlands would be open source but I can not spot it in the tender (https://www.tenderned.nl/tenderned-tap/aankondigingen/192421)... Did I make it up, or am I overlooking it?
This project's build status is almost poetic… https://github.com/softwaremill/retry
Good thread on the DP-3T protocol for privacy-perserving proximity tracking https://twitter.com/PowerDNS_Bert/status/1247617988368441344 .
I do not see malicious client (database of MAC -> ephemeral IDs) in the threat model. But that is a disadvantage of any broadcasting scheme/could be solved by MAC randomisation.
SURFnet (in Dutch) on the changes on their network at the start of the network. Traffic patterns inverted, traffic links to consumer isps close to congestion (and traffic being rerouted), research networks perform fine due to being over dimensioned
Vega Lite now contains more compact grids of charts. Nice change, makes that more usable on small screens.
I just realised that the market cap for KLM is less than the value of the guaranteed loans - so getting stock in return for downside risk may not be possible? https://ycharts.com/companies/AFLYY/market_cap
Besides zoom's end-to-end encryption not actually being end-to-end on my calls, it is not enabled by default.
If you see a healthcare capacity monitoring app that integrates into electronic healthcare records mentioned in the news in the Netherlands (https://2twnty4.nl): it has a business model where the user is the product.
Healthcare providers need a premium license to see detailed capacity information.
A social network for the University of Twente community. Anyone with an @utwente.nl or @*.utwente.nl email address can create an account here and participate in the global fediverse with millions of other users. This means that students, staff and alumni can create an account here. Content does not reflect the opinions or policies of the University of Twente.
We support \( \LaTeX \) formulas: Use
\) for inline LaTeX formulas, and
\] for display mode.