Rainer Müller @raimue@mastodon.social

HelloFresh delivery via DPD. #mpller

Best part during debugging: IPv6 always worked, because in ip6tables the FORWARD chain has an ACCEPT policy by default. Looking forward to nftables everywhere.

Writing this iptables rule feels stupid, but it solves the problem:
iptables -t filter -A FORWARD -i br0 -o br0 -j ACCEPT
https://serverfault.com/questions/963759/docker-breaks-libvirt-bridge-network

Okay, the tutorials I followed forgot to mention this. I just found out I also had to tell Wireshark to decode as HTTP on the custom 8443 port in the protocol preferences. Now it already looks quite usable with the "http" filter.

I know about using wireshark with SSLKEYLOGFILE to decrypt TLS traffic, but analyzing it this way is very unstructured because you can still only follow the TCP streams and not see individual HTTP requests directly.

What should I use to log HTTPS traffic to a service I host myself? I have a valid certificate and requests come from anywhere I do not control. I need something like a reverse proxy, but I want to inspect the full requests including headers and body. Any recommendations?

Waking this old MacBook from hibernation one last time before giving it to recycling. Everything not saved will be lost.

And most importantly, by using this advice when composing a thread, you make it possible to read threads on Twitter comfortably. Thanks.

By structuring a Twitter thread like this, you allow others to reply to a specific point you make. It also makes it possible to quote and retweet, because each tweet in itself has value.

The compose window on Twitter allows you to draft a whole thread at once. If your client cannot do this, switch to a better one. You can move stuff around and insert new tweets in between. When you are done, you publish it all at once.

If there isn't enough space left to finish your sentence, rephrase it or simply move it to a new tweet. You do not have to fill all of the 280 characters every time.

If you were writing a longer text, you would use paragraphs. In a Twitter thread, each tweet is like a paragraph. You use it to make one point and it should be self-contained.

Threads are inconvenient for me when people break mid-sentence into a new tweet. This is what makes it hard to read in the Twitter web interface or in a client. Just don't do this. Use short, coherent sentences for each thought you have.

You know what annoys me most about Twitter threads? It's the replies asking some bot account to "unroll". Here are some hints on how to structure threads that are meant to be read in a Twitter client.🧵

Was soll das eigentlich, dass ich im Online-Banking für die Suche in Umsätzen älter als 90 Tage eine TAN brauche? Dagegen kann ich problemlos ohne TAN alle Kontoauszüge der letzten Jahre als PDF herunterladen. Da ist halt dieselbe Information drin. #PSD2

Was soll das eigentlich, dass ich im Online-Banking für die Suche in Umsätzen älter als 60 Tage eine TAN brauche? Dagegen kann ich problemlos ohne TAN alle Kontoauszüge der letzten Jahre als PDF herunterladen. Da ist halt dieselbe Information drin. #PSD2

If you use #Firefox Android with an ultra-wide 21:9 display, this is the best solution to fix fullscreen on YouTube. By default top/bottom of the video will be cut off. https://stackoverflow.com/a/66488398/67498

The data was most probably acquired by systematically querying for phone numbers. I checked my privacy settings and "Who can look you up using the phone number you provided?" was set to "Everyone". The other values in the record are from my public profile only.

The last few days I have received SMS spam. I know my phone number was in the #FacebookLeak and the faked senders also are from the data dump as well.
If this is possible, why did spammers not just use random numbers before? The chance to hit an assigned number seems high?