rework mark_nonblock so it sets the cloexec flag too and use it in control.c to avoid these flags. (which are expected to become available on a future revision of POSIX and are already widely available.) It's not an issue for telescope to do the socket/accept + fcntl dance because there aren't threads that can fork(2) (there are no threads at all!) reported by @sikmir at github https://github.com/omar-polo/telescope/commit/59ef79dd19611c7846b00427e6f2267c748ae290#r74498414

and drop every mention of GPG...

it's currently unused (the net process unfortunately can't sensibly make use of it: breaks the DNS...) In the future I'd like to move the parsers in their own (sandboxed) process, so don't delete the landlock code yet.

it's just a typo: instead of looking for the end of the string pointed by path, if look for the end of uri. reached that point, path is NULL. reported by a gcc 11 warning: "argument 1 null where non-null expected"

in 0677399e i've re-added scan_scaled even if unused to reduce the diff with /usr/src. This causes a warning on platform that lacks that functions (everything not OpenBSD) because its prototype is missing.

can conflict with sys/prctl.h, spotted while trying to build on alpine linux.

add a mechanism that will signal a regeneration of the history file once it reaches the 150% of the HISTORY_CAP. while here refactor save_session, now it's somewhat readable.

was broken after the fs/ui merge

the recent commit to rework handle_save_page_path changed the tab id *before* setting it to the download, so the ui and net process went out of sync and downloads hanged.