The link shows a list of firmwares. The actual sources are at https://github.com/Freetz-NG/freetz-ng - but it's not the original FritzBox firmware sources:
> Freetz is a toolbox for developers and experienced users to build a modified firmware based on the original firmware for the DSL/LAN/WLAN/VoIP-Routers AVM Fritz!Box and T-Com Speedport (identical hardware) and to transfer this firmware to the device. There are many extension packages available, along with options to remove unwanted functionality from the original firmware.
> yes, ZFS can be configured to use a WAL too for durability
ZFS always uses ZIL for sync writes. You can optionally set `sync=disabled` for a dataset to switch it off. I would not describe it as "can be configured to use WAL".
Actually, using DEBUG for creating anything besides one or two simple DOS/BIOS calls was a PITA due to the lack of symbolic labels. After that, the MASM/TASM boilerplate started to feel bearable.
The article suggests using NFC normalization as a simple solution, but fails to mention that HFS+ always does NFD normalization to file names, and APFS kinda does not but some layer above it actually does (https://eclecticlight.co/2021/05/08/explainer-unicode-normal...), and ZFS has this behavior controlled by a dataset-level option. I don't see how applying its suggestion literally (just normalize to NFC before saving) can work.
Looking at your search results, there's just one recent ZFS corruption case with SEEK_DATA/SEEK_HOLE (in several HN reflections), a 2-year old Ubuntu-only buggy patch story, and some 2008 [Open]Solaris corruption.
Although being faster in some cases, ripgrep lacks archive search support (no, transparent decompression ignoring the archive structure is not enough) which works great in ugrep.