Danny Brian skribis 2007-11-29 10:57 (-0700): > Perhaps a pro XML-er can weigh in. Unlike many others on this list, I > use XML for almost everything. I think the point is what you're > saying here above, Jim. The benefits you describe of a native XML data type > boil down to a) encouraging a common approach to processing, and b) > not having to import modules.
And it does, but it doesn't have to be "native". It can be just "standard", including "de facto standard". Yes, XML is essential for many programmers. Yes, having a standard XML data type can certainly improve things for many people. But why on earth would it need to be bundled with Perl? All you need to make something the de facto standard, is to be good and the first, or better than all existing options. DBI is Perl 5's primary database API. Very few people ever consider using anything else. I think as many people use DBI as use XML in some way. It is NOT bundled with Perl, and has never been. Yet it is extremely popular. Do the same with XML, please. If anyone else reading this, feels like building this data type, with the code to back it, by all means please start today. But putting it in the core only helps in the forcing-down-one's-throat area. It doesn't improve maintenance, flexibility, or usability one bit. > "Encouraging a common approach" is not an easy or necessarily smart > thing in the XML space. Still, though, XML is very probably flexible enough (with its roles) that a single XML data type can still be a good idea. Something representing an XML element with its children is incredibly useful when standardised. And if it doesn't do what you want, just add a role that makes it do that. > A native XML type would only serve to antiquate Perl 6 long before > it's time (!), and is therefore a ... nonstarter. Amen. -- Met vriendelijke groet, Kind regards, Korajn salutojn, Juerd Waalboer: Perl hacker <[EMAIL PROTECTED]> <http://juerd.nl/sig> Convolution: ICT solutions and consultancy <[EMAIL PROTECTED]>
