Re: Perl 6 modules plan

Front page | perl.perl6.stdlib | Postings from August 2001

Thread Previous | Thread Next

From:

Jarkko Hietaniemi

Date:

August 11, 2001 12:50

Subject:

Re: Perl 6 modules plan

Message ID:

20010811143623.G23329@chaos.wustl.edu

> =over 4
> 
> =item Document namespace guidelines
> 
> This is a good opportunity to iron out some namespace wrinkles.  First
> step is to figure out what's already there on CPAN and can be documented
> as current "best practice" (eg. silly modules go in Acme::*, network
> protocols in Net::*).  

I must say I'm rather leery of the Net:: prefix, too.  So much
these days is network related that stuffing "net" in front is
fast approaching no information at all.

I would prefer to steer our thinking away from naming modules
as a strictly disclipined naming hierarchy

	Animalia::Chordata::Mammalia::Primata::Pongidae::Homo::Sapiens

towards a more to-the-point naming

	Man

(any sexist overtones are between reader's ears and a language artifact)

We don't need to repeat the full ancestry, the full logical "derivation"
of the module.  Most of the time it's unnecessary.

No, I'm not proposing a completely flat namespace.  The currently
suggested two-layer naming rule of thumb is still a good rule, it
helps in navigating, and sometimes to differentiate between otherwise
similar names (the namespace of TLAs is quite crowded, for example.)

That is, *if* one wants to navigate in a naming hierarchy.  I think
most of the time one doesn't want to "navigate", one wants to *find*.
Therefore I suggest having a very definite spot in the module metadata
for, ta-dah:

	keywords

The problem with hierarchical organization of data is that there are
probably as many, if not more, different hierarchies as there are
organizers (or users).  Another man's clean hierarchy is another man's
inscrutable mess.

With keywords there are as many *alternative* routes to the same
information as there are combinatorial pathways.  Which is a lot,
and much more flexible.

P.S. Sorry for this dropwise replying.

-- 
$jhi++; # http://www.iki.fi/jhi/
        # There is this special biologist word we use for 'stable'.
        # It is 'dead'. -- Jack Cohen

Thread Previous | Thread Next