Show newer

UOMF: Logging With #Emacs #OrgMode

I really love this method of notifying me when something is not running properly in the background such as a #cron job.
It provides easy to read instructions on my agenda.


Me, a unixy old person: "... and before end to end encryption was standardized, that meant that something called replay attacks were possible. To mitigate this, kerberos authentication was invented, which..."

A young unixy person: "can you pet it though."


"Kerberos was named for the three headed dog that guards the gates of hell, right?"

"... yes?"

"Can I pet the dog, is my question."



"Ok, no, but file that bug immediately."

After more clicking and searching I have to admit defeat. I cannot find the report that was presented to me by reportbug through the web interface 😰

In the meantime the Debian BTS replied with a mail containg a link to the report:

If I understand the thread correctly then the bug was already discovered in unstable and should be fixed upstream and in 5.15. But still, this version entered testing and will thus be presented to all testing users, so take care!


Show thread

Now I want to see the same bug report in my browser so I head over to the Debian Bug Tracker:

And it shows .... No reports found!

Aha. Heading to the Package info for this package and then clicking on "Fehlerberichte" also finds no reports:

Where do I file a bug report for the bug tracker? 🤔


Show thread

The Debian package overview shows that this package entered unstable on November 4th and was migrated to testing just yesterday (Nov 15). So ok, let's file a bug report.

reportbug presents me with 917 bugs on that package 😱 but looking through them it seems those are all bugs filed against the Linux kernel package, not only the one I am interested in.

And indeed there is a bug report already showing the symptoms I see. So I add a comment on it and send the mail.


Nov 16 09:24:43 krikkit kernel: amdgpu 0000:08:00.0: amdgpu: failed to write reg 28b4 wait reg 28c6
Nov 16 09:25:03 krikkit kernel: amdgpu 0000:08:00.0: amdgpu: failed to write reg 1a6f4 wait reg 1a706
Nov 16 09:25:23 krikkit kernel: amdgpu 0000:08:00.0: amdgpu: failed to write reg 28b4 wait reg 28c6
Nov 16 09:25:43 krikkit kernel: amdgpu 0000:08:00.0: amdgpu: failed to write reg 1a6f4 wait reg 1a706


Show thread

A new Linux kernel entered testing (Bookworm) and so I decided to test it: linux-image-5.14.0-4-amd64. Turns out it does not boot correctly on my AMD Ryzen 5 2400G with AMD graphics.

Initially I thought the system is dead but it does make progress after half a minute and after another minute even gets into GNOME but the kernel log shows real errors:


In die Richtung lässt sich einiges machen. Manches sogar noch einfacher?
Warum muss ich als Radfahrer beim linksabbiegen wenigstens zwei Ampelphasen warten, als Auto aber nur eine?
Warum müssen Radfahrer und Fußgänger überhaupt einen Knopf drücken und bekomme für die jeweilige Richtung nicht einfach grün mit?
Warum ist bei größeren Kreuzungen die Fahrradgrünphase nur 5 Sekunden (erste Ampel, bei in Reihe stehenden Ampeln für jede Verkehrsinsel.)

Die Pressemeldung von Luca ("Bars und Clubs sind #Corona-Hotspots") bezüglich Clubs und Bars führt ganz gerne in sehr falsche Richtung.

Diese Locations sind im Luca-System Warnungshotspots.

Show thread

Bevor ihr die Pressemeldung von #lucaApp wieder falsch lest und da jetzt ein Infektionsrisiko an Orten vermutet und andere vergesst, möchte ich auf beispielweise diese Grafik hinweisen:

Auch mal eine gute Idee - warum nicht im Normalfall Grün für die Fußgänger und Radfahrer statt immer für den Autoverkehr?

I do find that I do not bother enough to actually do this and thus stop using extensions altogether.

This end result is kind of sad but it is the outcome of adding up little inconveniences until the payoff is less than the costs...

Show thread

For example this happened to me on the audio-switcher extension a while ago. I then found that there are even multiple forks of the same extension registered on the gnome extension website differing only by the supported versions in the manifest file 😭

Knowing the core of the problem I can quickly clone the git repository and fix the metadata myself but to make things work I then need to log out and back in.

Show thread

Since the switch to Wayland, working with extensions has become a pain. It is no more possible to restart gnome-shell with "Alt-F2 r Enter" to dynamically adapt to changes for the extensions. A complete logout / login cycle is required now and I simply do not find this acceptable.

Debian Bookworm is so close to top of tree that it is not uncommon for extensions to stop working as they simply did not bump the supported version in their manifest files yet.

Show thread

So I tried to adapt to the horizontal arrangement of workspaces for two weeks, but I find it a digression on two monitors. Pushing a window onto another workspace from the overview has become more clumsy and difficult.

Given that this is my only use for the overview mode I now reverted back to the old behaviour through this gnome-shell extension:

Wir wissen ziemlich genau, was uns und unsere Kinder als Gesellschaft und als Spezies bei dem aktuellen Prognosekorridor der Erderwärmung erwartet. Ist wie Corona oder bei der Nebelbank auf der Autobahn: Wir meinen vollgasgewohnt, es reiche auch noch nach dem Aufschlag zu bremsen


The Chrome team just landed a patch that lets sites block "View Source" _right in the middle_ of the Chrome Dev Summit.

You can't make this stuff up. I owe my career for large part to View Source and tinkering with a webpage's building blocks.


Die Prüfergebnisse zum Einsatz von #Microsoft365 an Schulen in BaWü sind nun online über FragdenStaat abrufbar. Aufgrund zahlreicher rechtlicher Risiken (#Datenschutz) rät der LfDI vom Einsatz ab.


Show older | Hostsharer im Fediverse ist die Mastodon-Instanz für Mitglieder der Hostsharing eG. Sie wird seit 2018 von der Hosting-Genossenschaft betrieben.