It looks like that sub was completely removed recently but the DB::Color module hasn't been updated since 2015.

Edit: of course that was in bleed so that specific change is probably not exactly your problem unless you're using a bleed compiled perl. That being said though looking in the 5.34.0 source I don't see a sub _cmd_l defined therein.

Oh and the sub looks to maybe have been renamed _cmd_l_main (haven't grobbled through the entire history, but as a guess . . .). Prossibly if you patch DB::Color to use that instead of the plain _cmd_l it may just work (in which case you probably should file an issue with the author through the github tracker).

bleed compiled perl

By this, do you mean a development version of Perl that has not yet been officially released???

Correct; the "bleed" branch is the name for the bleeding edge development line in the git repository. If one were to compile from a git clone with that checked out you'd get whatever the 5.35.tobe version currently looks like. I'm guessing that's not the case here as the module's similarly unworky for me with 5.34.0 (however its tests pass); I merely pointed to that change as the first sign I saw of what might be going on with the sub in question.

Edit: me no spal gud at 1 in morning obviously . . . but I'm consistent, so I've got that going for me.

The cake is a lie.
The cake is a lie.
The cake is a lie.

> This is only a proof of concept. In fact, it's fair to say that this code sucks. It's not very configurable and has bugs. It's also going to possibly be a memory hog, as if the debugger wasn't bad enough already.