Admitting Defeat on K&R in LCTHW

He is sadly correct. I've made the same experience with Perl. Most of the popular and wide-spread Perl books are written in the 90s, and since then we've learned as a community that many of the things they teach are outright dangerous to newbies and should never be found in anything but the worst throwaway code, because they're both harder to read and sources of errors.

There is a number of great, recently written books, that teach Modern Perl. A Perl style that makes code easier to read and more correct.

Yet time and time again there are people who will recommend Learning Perl 4th, because they read it, and when told about the new ones and the bad properties of the books they recommend, they'll dig in and argue, often ending in: "They teach Perl in an entertaining manner, so they're fine to recommend."

You simply cannot "help" such people. To "help" would amount to destroying large parts of their world view in a "you have been doing things wrong for decades" manner. Many minds are neither able nor willing to take that and will reject it as strongly as possible.

Indeed. I think I fit into that category, first used UNIX(TM) in 1978, first programmed in C in 1980-1 ... and I've never thought K&R was a good book. It was Lions' Commentary on UNIX 6th Edition, with Source Code (https://en.wikipedia.org/wiki/Lions%27_Commentary_on_UNIX_6t...) that was the key, and since its first edition, I've used C: A Reference Manual by Harbison and Steele (as in Guy Steele), initially written for a CMU spinoff compiler company because K&R was an insufficient reference for implementing a compiler.

Arrogant, no. Frustrated with close minded people and hero worship? Yes. Though, he does tend to bring it on himself even when he's correct.