[Toybox] [PATCH] tail: implement -F (and its companion -s).

[Rob Landley]

On 7/1/21 11:40 AM, enh wrote:
>     This is mostly a "code style" thing, I need to work out the preferred style for
>     dealing with the context switch, and document the issue. (The compiler is
>     forgiving, but I don't want to confuse people reading the code...)
> 
> yeah, i did scratch my head about that one for a bit before deciding to punt the
> problem to you :-)

In _this_ case an anonymous struct works fine because the only user of the name
is sizeof and we can just sizeof(*TT.F). But I should still write up a general
whatsis.

Sigh, I have like 15 todo items that are all reasonably _easy_ to do, and I'm
emotionally exhausted by job hunting and uncertainty. I'm doing the tabsplosion
thing where stuff gets interrupted by other stuff and I have a dozen open tabs
with half-finished stuff.

By the way, those "file went away, file came back" messages are to stderr, not
stdout. Does that count as an error exit? No idea: when I ctrl-c and echo $? it
tells me 130 because it's saying it exited due to a signal. SIGH.

> pointer to an incomplete type seemed better than void* but, yeah, moving the
> struct definition below GLOBALS would have been more in keeping with the rest of
> the tree.

I need to update the docs and do explanatory videos, but right now even the
level of jetlagged "I did not finish editing the material into a coherent order
that fits in the timeslot" of https://www.youtube.com/watch?v=Sk9TatW9ino or
https://ia803008.us.archive.org/34/items/OhioLinuxfest2013/24-Rob_Landley-The_Rise_and_Fall_of_Copyleft.mp3
is more than I've got in me because I don't know if I'll be able to take my cat
to the cardiologist apointment I just made or if I'll be in New Mexico to start
a new job by then.

Rob

P.S. No I don't know why https://www.bilibili.com/video/BV1gp4y1Q7To/ either.