| Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
`ffmpeg` fails as it is not run with an output file, and we don't really
want to transform the input in any way. We just want the nice metadata
format. `ffprobe` provides just that and is thus better suited for our
usecase.
|
|
The printer can be offline when starting the system, or it can be
otherwise unreachable.
|
|
|
|
|
|
`--top-level` is now default.
|
|
|
|
|
|
I started using `iamb` when it was really not-finished yet (and used a
json config).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Otherwise, the two of them will have different seek parts in the files
and override each other.
|
|
|
|
`rec` can't handle the quoted reference to the attribute set
(`<LEFT_SUPER>`).
|
|
|
|
|
|
|
|
This ensures the same spacing for each block.
|
|
|
|
This happens on apzu, as she has only an integrated intel GPU.
|
|
|
|
|
|
|
|
Otherwise, the prompt for the new screenshot is overshadowed by the
normal key mappings in the `run` mode.
|
|
This is hard to type
|
|
|
|
|
|
Otherwise, builds are performed on-disk.
|
|
That just avoids the visual clutter, caused by the nix store paths.
|
|
This avoids bit-rot in apzu's key mappings and maybe helps me to
actually learn all the possible key mappings.
|
|
Both (lix and cppnix) have a fix for some rather serious CVEs. But lix
also introduced a rather serious system breaking bug. So we need to
update to something that works now, and I'm not really motivated to
apply the patches for the lix bug manually.
|
|
This might help remedy the incredible long start-up time.
|
|
The `eprintln` was previously supplied (impurely) by my shell library,
and without it, the command-not-found handler goes into infinite
recursion.
|
|
It is fun, but not worth the extra complexity.
|
|
|
|
|
