> There's been a lot of discussion about it which I unfortunately couldn't
> follow.  Now it's up to 0.04 and seems to have a lot of functionality, and
> I really don't see the point of all of it.  And even worse, it's gotten
> _more_ intimate the with DT::TZ internals, instead of less.

There was a lot of discussion. :)

If it doesn't show up in the archives I can paste a thread together for you.

DT::TZ::LINKS is still the only internal structure that is modified.

> Why do we need something like is_timezone()?  And if it's needed, it

So you can tell the difference between US/Hawaii and Pacific/Honolulu.  Which may seem 
perfectly clear to you but unless you look at the internal data structure is probably 
not to most people.

> almost certainly does not belong in a module called Alias!  Why does it

Probably not - but what ever module handles the aliasing should have a complete 
interface as it's all related functionality.  This was missing elsewhere.  Maybe we 
should talk about moving this into the DT::TZ dist.

> have the timezones() method which completely duplicates the functionality
> of DateTime::TimeZone::all_names()!

Because all_names() returns a direct reference to the data structure and I was worried 
about protecting the internals.  And as I stated above it's all related functionality.

> All of the stuff about dying if an alias is defined and so on seems like
> total overkill.  The purpose of this module, AFAIC, is for individual
> end-users to be able to define some useful _local_ aliases, in an

This is related to comments made about using this module under mod_perl.  I think it's 
clear that a lot of people will use this to handle localization of timezones (like 
what EST is defined as).

> environment _they_ control.  No one should ever be distributing a CPAN
> module that uses this, and distributing an app that does this by default
> would almost certainly be a bad thing too.

But they will...  And why shouldn't someone be able to define what timezone X means to 
them in a distributed app?  I this passes the 'is it useful' test.

-J

--

Reply via email to