What are you suggesting is fixed in 5.30? I was not aware of any change in behaviour around this (which I understood was unfixable due to back-compat), and indeed I see the same output for the OP's script with 5.30, 5.32, 5.34 and 5.36.

perl -wE 'BEGIN { print $] } my $x if 0'
5.037003
This use of my() in false conditional is no longer allowed at -e line 
+1.
[download]

map{substr$_->[0],$_->[1]||0,1}[\*||{},3],[[]],[ref qr-1,-,-1],[{}],[sub{}^*ARGV,3]

Ah ok, that appears to be a quite specific case being caught, I would guess during constant folding:

% /opt/v5.36.0/bin/perl -wlE 'BEGIN { print $] } my $x if $ARGV[0]' 0
5.036000
% /opt/v5.36.0/bin/perl -wlE 'BEGIN { print $] } my $x if $ARGV[0]' 1
5.036000
%
[download]

not the same thing

This my $x if 0 is definitely a bug, because constant folding shouldn't allow a lexical to be declared.

But the OP has a condition which is only checked at run-time, while the declaration happens at compile-time.

This means the new variable will not be overwritten with an init-value if the condition is false.

The deeper issue is that postfix-conditions have no own scope, the problem is non-existent for if ($bool) {my $x = ... }

Not sure if this qualifies as a bug or a weird feature.

But a warning should happen in any case if a declaration happens before a postfix condition, because of the broken symmetry to prefix conditions.

Cheers Rolf
_{(addicted to the 𐍀𐌴𐍂𐌻 Programming Language :)
Wikisyntax for the Monastery}

But thanks a lot for the article link!
The article makes things very clear! :-)