Surprised they missed follow! It’s a bit odd to use, but once you get used to it it’s better than tail in many circumstances IMO. `less +F` starts less following stdin or whatever file argument you’ve provided. <C-c> breaks following, allowing you to search around a business-as-usual `less` session. Hitting `F` (that’s uppercase) starts following again. Yes, you can just start following within a session with `F` too if you forgot to add +F to the `less` invocation.
If you're following a pipe (such as `kubectl logs | less +F`), <C-c> is sent to all processes in a pipeline, so it stops less from following and it stops the other process entirely. Then you can't start following again with F, or load more data in with G.
Less provides an alternative of <C-x> to stop following, but that is intercepted by most shells.
Funnily enough, it literally tells you right there on the bottom line: “Waiting for data... (^X or interrupt to abort)”. No shame in not noticing, just another case of blindness to long-familliar messages I guess.
By the shell or by the kernel’s terminal discipline or by the terminal emulator? AFAIU the shell is basically out of the picture while `less` is running.
> I can <C-z> while less is running to background that process using the shell, so the shell is clearly not completely gone.
The shell isn’t gone, but it isn’t active either from what I understand. The function of converting the user’s typing ^Z on a terminal (or a ^Z arriving on the master end of a pseudoterminal) into a SIGTSTP signal to the terminal’s foreground process group is[1] a built-in function of the kernel, much like for ^C and SIGINT or ^\ and SIGQUIT. (The use of ^Z resp. ^C or ^\ specifically, as well as the function being active at all, is configurable via a TTY ioctl wrapped by termios wrapped in turn by `stty susp` resp. `stty intr` or `stty quit`.) So is the default signal action of stopping (i.e. suspending) the process in response to that signal. The shell just sees its waitpid() syscall return and handles the possibility of that having happened due to the process stopping rather than dying (by updating its job bookkeeping, making itself the foreground process group again, and reëntering the REPL).
I am not saying that doing job control by filtering the child’s input would be a bad design in the abstract, and it is how terminal multiplexers work for instance. I admit the idea of kernel-side support for shell job control is pretty silly, it’s just how it’s traditionally done in a Unix system.
Whew! Advanced Unix system programming level stuff. I've dabbled a bit in that field, in C, on Unix, some older versions on PCs. It was fun. Any recommendation for a tutorial style book or site or blog on the subject, other than man pages and the Kerrisk book (TLPI, which is more of a reference), for Linux?
With `tail` you can press enter a few times to put some empty lines after the last line. This is useful e.g. when you trigger a function multiple times and want to easily see line groups from each attempt. It's the only reason I still use `tail` for following when `less` is available.
A visual mark would be nice, agreed. I haven't tried it, but I wonder if you could approximate it with the bookmarking feature that less(1) does have. It wouldn't be visible, but it would scroll to a consistent mark.
I usually use tail when I need to do some ad-hoc log following.
Having to set bookmarks and remember them is a PITA I can usually do without. If I'm looking at "normal" log output, it's usually set up in a nice aggregator somewhere, where I can easily exclude noise and otherwise uninteresting output.
Maybe OT, but I thought for a long time that "follow" was some sophisticated file descriptor trickery that required you to somehow "stream" the file while reading and would therefore be incompatible with opening a file "normally".
My mind was blown when finding out its really just "keep on polling after EOF". Meaning there is absolutely no difference between opening a file normally and "following" a file - and software could easily switch between the two "modes" on the fly.
It would be nice to have a mode that follows in the sense of automatically picking up new output, but that simultaneously would let you navigate around, similar to how terminals behave. Then you’d only need an autoscroll toggle for when you’re at the bottom.
To elaborate on this, lnav (https://lnav.org) is always polling files to check for new data and will load it in automatically. It does not require the user to do anything.
As far as following the tail of the file: if the focused line is at the end of the file, the display will scroll automatically; otherwise, the display will stick to the current position. Also, if there is a search active, matches in the new data will be found and highlighted.
I would say it's a bad UX and not just odd. I can't see any benefit to making it modal. It should just load new data as it becomes available without making the user do anything.
