firefox = {
      profiles."user" = {
        bookmarks = [
          {
            name = "Toolbar folder";
            toolbar = true;
            bookmarks = [
               {
               name = "nix folder";
               toolbar = false;
               bookmarks = [
                  {
                    name = "NixOS Search";
                    url = "https://search.nixos.org/packages";
                  }
              

On phone so testing or reading source is a bit tough, but maybe nesting folders like the above work?

Make this the toolbar directory. Note, this does not mean that this directory will be added to the toolbar, this directory is the toolbar.

https://github.com/nix-community/home-manager/blob/master/modules%2Fprograms%2Ffirefox.nix#L425-L427

If you open the bookmarks utility in firefox, does it show as a folder as expected? I think maybe toolbar should be false if you expect to see your folder on the toolbar

I think I share your view about course correction needed but if anything this seems to be a step in the right direction, I really do not see how Mozilla is being unreasonable (with this feature). To have a chance at the market outside of core geeks some telemetry is needed, and if we truly think that anonymoused (not connected to session, IP, or user, only distinguishable from others by timestamp of when data was recieved) counters based on a few narrow categories, then I think the userbase is a part of the problem.

should we really protest when they are doing it right? are we not being unreasonable if we take the stance that no telemetry is allowed?

the only step in a more privacy-direction I can see this feature taking is if it is opt in by default, but this being so incredibly trivial non identifiable info, I think maybe it's better for all parties for those to just use Librewolf then

I think those are all "yes"-boxes, that shows the nuances of different types of abuse and the wether it's ongoing or not or not yet realised

This measure is not linked with specific individuals and is further anonymized using a technology called OHTTP to ensure it can’t be connected with user IP addresses.

it's kinda ridiculous to see them emphasise this but get called out as if they are doing so. if anything they seem to be taking the most privacy focused approach I've come across, going forward as a decent example. Also makes me appriciate that they look for other feedback than user comments cause that seems like a notoriously unreliable source of info for data-driven decisions

I (not who you are asking) responded to your original comment about cheat.sh, but feot kinda fitting here too,

in the case if cheat.sh it does so much more than the manpages that I would definitely say that manpages should keep doing it's thing, because these tools strive to do more, which both makes them valuable and makes the manpages the right tool for it's thing.

I think cheat.sh has the upper hand over tldr, it retrieves the DBs from tldr and others, gives far better results imo. rsync is a decent exampøe where cheat.sh does better than tldr imo: https://cheat.sh/rsync

as for cheat.sh vs manpages: each has their uses. As someone who uses rsync once every .. two months, maybe, cheat.sh gives me the info i need much quicker. ie: -avz, but maybe -c if you want to verify file integrity, that's 8 lines/2 examples in, but reading the manpage of rsync then checksumming is almost something you need to know to look for, which is fine for what the manpages are intended for. these cheatsheets gives you common use cases, and are more of a quick reference.

also cheat.sh gives a lot more functionality than man, I can recommended skimming over the github page https://github.com/chubin/cheat.sh

I've had a lot of gains from setting ingredient ranges on workbenches really low, iirc that didn't really register in the analyzer but made a real noticable impact