deleting a disallowed key should croak too

[XSven]

You have mentioned an alternative implementation way that could be based on Hash::Util::lock_keys(). But such kind of implementation would be different because your DELETE() override does not croak when applied to a disallowed key. From my perspective the DELETE() override should be changed so that the assertion

...
eval { delete $hash{four} };
ok($@ ne '');
...

passes.

[davorg]

I wanted the same behaviour as a standard hash - where deleting a non-existent key fails silently.

[XSven]

The important difference is that the non-existent key is a disallowed key (not part of the fixed key list) . Such a key should trigger the same behaviour when calling STORE() and DELETE(). That is also what Hash::Util::lock_keys() enforces!

[davorg]

To be honest with you, this module was written as a toy, demonstrating tying modules. I wrote it for a perl.com article over twenty years ago. I don't expect anyone to use it (especially not since lock_keys() has been available).

If you'd like to send me your CPAN ID, I'd be happy to hand the module over to you. I'll add you as a collaborator on the repo so you can do what you like with it.

[XSven]

Sorry I don't want to upset you. I know the perl.com article and that was the trigger to look into this module. One of the problems of CPAN is the growing number of toy modules. I think you have written an article about some other problem that CPAN suffers from. Unfortunately I cannot find it anymore. I try to take each module that I am looking at serious. So my CPAN id is SVW.

[davorg]

Oh, I'm not upset at all. I just don't really have the time or interest to maintain this module any more. So it's all yours :-)

I've sent you invitations to become a collaborator on the repo. And I'll now give you comaint on CPAN.