blog.crox.net

The 2006 Dürrenast train disaster - "trolley problem" for real

nospam@example.com (crox) — Sun, 24 Aug 2025 11:17:00 +0000

The 2006 Dürrenast train disaster occurred on May 17 when a runaway BLS maintenance train with failed brakes sped downhill from Frutigen towards Thun, ultimately crashing and killing three people. As the train gained uncontrollable speed, dispatchers faced critical choices, similar to the theoretical "trolley problem": routing the train towards populated or obstructed lines would potentially cause greater casualties. Ultimately they opted for a crash against stationary work trains in Dürrenast to limit wider harm.

Max S wrote a detailed account of the events on Medium.

On SWI swissinfo.ch you will find an article about the events, as well as a link to an SRF film (in German with subtitles in several languages).

Disable WiFi Power Management on Raspbian 12 (Network Manager)

nospam@example.com (crox) — Fri, 15 Aug 2025 22:48:00 +0000

8 years later there are still issues on Raspberry Pi / Raspbian with WiFi power management causing disconnects and generally bad performance.

I've found a reliable method to disable WiFi Power Management for the on-board WiFi chip (brcmfmac) here: NetworkManager WiFi Power Saving.

Ensuring the EFI partition is mirrored on an Ubuntu install with root on software RAID 1

nospam@example.com (crox) — Sun, 04 May 2025 19:28:00 +0000

I found a working procedure here:

- Make sure both EFI partitions are mounted from /etc/fstab, eg. using /boot/eficopy in addition to /boot/efi. (Both partitions need to have the proper partition type GUID, i.e. C12A7328-F81F-11D2-BA4B-00A0C93EC93B or "EFI System".)

- Run dpkg-reconfigure grub-efi-amd64 and include both partitions when prompted.

libheif1 >= 1.18 on debian 12 bookworm (installing packages from backport in debian stable)

nospam@example.com (crox) — Fri, 02 May 2025 18:08:00 +0000

Add this line to /etc/apt/sources.list:

deb http://deb.debian.org/debian bookworm-backports main

Run:

# apt-get update
# apt install -t bookworm-backports libheif1

This solves the issues encountered when decoding HEIF (.heic) files produced by iPhone / iOS 18, causing for instance rendering in Nextcloud to fail.

This page has more details on using Backports.

Nextcloud behind reverse proxy - WOPI URL changed from https to http

nospam@example.com (crox) — Wed, 05 Mar 2025 22:45:00 +0000

I'm trying out the "Nextcloud Office" app together with "Collabora Online - Built-in CODE Server". That Nextcloud instance is behind a reverse proxy.

The problem I encountered was that regardless of the 'overwritehost' and 'overwriteprotocol' settings, the WOPI URL kept being automatically changed from https to http.

From other posts regarding similar issues I figured out that the WOPI URL is obtained by querying https://[yourinstance]/apps/richdocumentscode/proxy.php?req=/hosting/discovery

Looking at the proxy.php source code, I found the reason for the issue:

// URL into this server of the proxy script.
if ((isset($_SERVER['HTTPS']) && $_SERVER['HTTPS'] !== 'off')
        || (isset($_SERVER['HTTP_X_FORWARDED_PROTO']) && $_SERVER['HTTP_X_FORWARDED_PROTO'] === 'https' )
        || (isset($_SERVER['HTTP_X_FORWARDED_SSL']) && $_SERVER['HTTP_X_FORWARDED_SSL'] === 'on')
) { 
    $proxyURL = "https://";
} else {
    $proxyURL = "http://";
}

I then modified my reverse proxy (Apache) config to add the missing header:

RequestHeader set X-Forwarded-Proto "https"

Now everything works as expected

Ubuntu Server 24.04, software raid as boot device (with EFI)

nospam@example.com (crox) — Sun, 09 Feb 2025 21:59:00 +0000

Everything you need in this excellent article: https://alexskra.com/blog/ubuntu-20-04-with-software-raid1-and-uefi/.

Website to test HTTP requests

nospam@example.com (crox) — Mon, 13 May 2024 21:34:00 +0000

And many other things: https://webhook.site/.

BackupPC - binary garbage in XferLog.z / XferLog.z getting huge

nospam@example.com (crox) — Fri, 30 Dec 2022 18:32:00 +0000

I run a BackupPC instance that is still on Debian 10 / buster. The latest rsync package for Debian 10 has version number 3.1.3-6.

I recently noticed infrequent issues when backing up hosts that are either on Debian 11 / bullseye, or on Ubuntu 22.04 (which both ship rsync 3.2.3). The symptoms are as follows:

- backups take much longer than usual
- XferLog.z starts "normally", but after a certain point contains a lot of binary garbage, and gets much bigger than usual (hundreds of MB, or even in one case up to 12 GB)

After investigating and looking for information online, I came across these bug reports, which contain the explanation as well as a workaround: Debian Bug report #969463 and BackupPC issue #369.

=> The issue is caused by a combination of a change of default behaviour introduced with rsync 3.2.3, and a bug in File::RsyncP.

The solution that works for me (pending an update to Debian 11 and BackupPC 4) is to add the following line in the individual server config for each of the affected hosts:

$Conf{RsyncArgsExtra} = ['--no-msgs2stderr'];

3D-Fuel Pro PLA+ material profile for Cura LulzBot Edition / LulzBot Mini printer

nospam@example.com (crox) — Sun, 18 Sep 2022 13:11:00 +0000

For some reason, Cura LulzBot Edition does not include a material profile for Pro PLA+ by 3D-Fuel used on a LulzBot Mini printer.

Here are the settings that work for me: 3d-fuel-pro-pla.xml.fdm_material (2.85 mm).

Split a multi-object STL file into separate files containing each one object

nospam@example.com (crox) — Sat, 20 Nov 2021 14:03:00 +0000

slic3r --split input.stl

This will create input.stl_01.stl, input.stl_02.stl etc. in the same directory.

This worked perfectly for me on a file created with OpenSCAD containing different parts that I wanted to print individually with Cura Lulzbot Edition. I found the info here.

Find large files in a BackupPC transfer log (sort files in XferLOG by size)

nospam@example.com (crox) — Sat, 13 Mar 2021 14:40:00 +0000

When your backups suddenly takes much longer to complete, it could be because large files that were previously excluded from the backup were renamed or relocated elsewhere. In order to identify those files, or just to sort the list of files by size, I use the following code (bash):

cd /tmp/
BackupPC_zcat /var/lib/backuppc/pc/mypc/XferLOG.99.z > xlog
for S in `cat xlog | sed -e 's/^[^/]*\/[0-9]*[ ]*//' | cut -f1 -d\  | egrep '^[0-9]+$' | egrep '[0-9]{9,}' | sort -n | uniq ` ; do fgrep " $S " xlog ; done

This will show the list of files bigger than 99999999 bytes (100 MB). Remove "| egrep '[0-9]{9,}'" to just list all instead.

Note: on Debian, BackupPC_zcat is in /usr/share/backuppc/bin/. I've added a symlink in /usr/local/bin/ so that I don't need to look for it every time.

Configure Garmin Express to download files to a different partition

nospam@example.com (crox) — Sat, 28 Mar 2020 23:30:00 +0000

Short answer: you can't. Files (map data, firmware updates etc.) are always downloaded to %ProgramData%\Garmin\CoreService\Download (typically C:\ProgramData\Garmin\CoreService\Download)

The only solution I could find is to move that folder to another partition (make sure to completely exit Garmin Express first), then create a link to it:

c:>cd C:\ProgramData\Garmin\CoreService

c:\ProgramData\Garmin\CoreService>mklink /J Downloads "d:\Garmin.Downloads"
Junction created for Downloads <<===>> d:\Garmin.Downloads

c:\ProgramData\Garmin\CoreService>

Add HEIF support to Ubuntu Linux 18.04 (thumbnails / gimp) - preview and open HEIC files from iPhone directly in Ubuntu

nospam@example.com (crox) — Mon, 16 Mar 2020 22:10:00 +0000

Since Apple moved from JPEG to HEIF as default format about two years ago, there has been a lot of progress.

There is a very straightforward way to add minimal HEIC/HEIF support to Ubuntu:

sudo add-apt-repository ppa:strukturag/libheif
sudo apt-get install heif-gdk-pixbuf heif-gimp-plugin heif-thumbnailer

That's it! You don't even need to logout or restart, the new functionality is available immediately.

More details here and here.

Geohashing Wiki Down

nospam@example.com (crox) — Tue, 07 Jan 2020 22:13:00 +0000

The geohashing wiki hosted at wiki.xkcd.com/geohashing had been down since September 2019.

It is now back at a new address https://geohashing.site/geohashing/.

[updated with the latest info on 2020-03-09]

All-black maps with staticmaplite (OSM static map generation - Access control configuration prevents your request from being allowed)

nospam@example.com (crox) — Sun, 04 Aug 2019 00:21:00 +0000

I've been using the staticmaplite library for a while to generate static maps based on OSM tiles.

Recently I started getting the following error when downloading tiles:

Access control configuration prevents your request from being allowed at this time. Please contact your service provider if you feel this is incorrect.

The OSM team was very responsive when I contacted them and they pointed me to https://operations.osmfoundation.org/policies/tiles/.

Based on that info, I checked how staticmap.php was fetching tiles and noticed that the User-Agent string was hardcoded to "Mozilla/4.0". I fixed this and now I can fetch tiles again. You can get the updated version from here: https://github.com/crox-net/staticmaplite.