The Mysterious Case of the Probable Bit-Flip

I use StumpWM, a window manager written in Common Lisp, on my primary work machine, and have added a
number of customizations to it.  It's similar to GNU Emacs, in that everything's a binding which can
be changed, although there are some changes I never bothered making to its internals after attempts.

StumpWM provides a feature called the mode-line, in part controlled by the *SCREEN-MODE-LINE-FORMAT*
variable; that list of strings or code is evaluated on a timeline defined by the *MODE-LINE-TIMEOUT*
variable, set to one second on my system, and the mode-line makes many simple features rather simple
to provide.  The following list element reports the current battery capacity, with very little work:

 (:EVAL (IGNORE-ERRORS (WITH-OPEN-FILE (*STANDARD-INPUT* #p"/sys/class/power_supply/BAT0/capacity")
                         (READ-LINE))))

I use similar code to report other characteristics in my mode-line.  One day I noticed the mode-line
to be mangled, with blanks where these characteristics ought to be.  StumpWM provides the ability to
evaluate expressions via the C-t : chord by default, and I soon enough saw that all such expressions
were entering the debugger with a curious reason: The system told me that WITH-OPEN-FILE was integer
in type and therefore not suitable with how I was using it.  I believe this originally happened when
trying to re-evaluate my configuration file.  After my confusion dissipated, I realized myself to be
witness to a likely bit-flip error, which my Common Lisp system was dutifully ignoring as I'd asked.

After logging into the same system via a virtual terminal, a measure to maintain my GNU Emacs state,
the only program state that matters much to me, I killed StumpWM and it was restarted automatically.
I thereafter saw none of this oddity, and it's not occurred since.  I've long known runtime checking
to be defense against hardware failures, but it's nice to have a story in which this happened to me.
.