![](https://seccdn.libravatar.org/avatar/1d1201a6224c340b67e1a7ee29f57504.jpg?s=120&d=mm&r=g)
Anton, et al -- ...and then Anton Aylward said... % % On 2021-03-18 7:09 a.m., David T-G wrote: % >I don't like that code at all. It's guaranteed to succeed, whether .alias % >exists with a non-zero length and parses happily, or .alias doesn't exist, % >or .alias exists AND HITS A FAILURE, which doesn't make sense. % % Oh, it DOES make sense. % It HAS to succeed since this is in .bashrc. [snip] I respectfully disagree :-) I would as a distribution ensure that all of the stock code provided will under normal circumstances get to a login prompt, and I would as an SA ensure that any code added would not lock up the system, but the user is on his own, and I as a user would much rather actually see an error than blindly miss a failure. This is just a bad practice, IMHO. But we all have opinions, and these days mine is even less weighty than it might ever have been ... I also would love to see an example of something critical to a system that would go in a user's run commands that should be allowed to silently fail rather than holding up a shell login. HAND :-D -- David T-G See http://justpickone.org/davidtg/email/ See http://justpickone.org/davidtg/tofu.txt