You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
$ type -a hist
hist is a shell builtin
hist is /Library/Frameworks/Python.framework/Versions/3.6/bin/hist
hist is an undefined function
in other words, if I try to run hist I will get the shell builtin first. I can circumvent that by using the env program, but this is not obvious right away.
The text was updated successfully, but these errors were encountered:
What is likely to be builtin really depends on what shell whoever is using, and there are enough shells there so changing API just because one of them does not seem to be good enough reason.
POSIX sh builtins (hist is not one of them, according to posix manpages on Fedora 30) would get exception as most of them is almost guarranteed to be in many shells, but then again, POSIX itself defines command so you can work around your problem by calling
command hist
(for interactive usage, you could do something like, alias bpl_hist='command hist' could be better.
That said, if bashplotlib would want to avoid conflicts altogether, renaming all commands to eg. bashplotlib-hist, etc. could be one solution but it seems like overkill though... :)
Well, it's not like ksh93 is one of the new shells with a small user base. It's also not like it changes a lot -- it's been there since at least 1993.
I respect whatever decision you make, I'm just pointing out that the current failure mode in one of the older, bigger shells is a terrible user experience. If it's not fixed, it should at the very least be documented.
My shell (ksh93) informs me that
in other words, if I try to run
hist
I will get the shell builtin first. I can circumvent that by using theenv
program, but this is not obvious right away.The text was updated successfully, but these errors were encountered: