This page is no longer maintained — Please continue to the home page at www.scala-lang.org

Re: (fwd) the mayor of xmltown

No replies
Alex Cruise
Joined: 2008-12-17,
User offline. Last seen 2 years 26 weeks ago.
From: "Normen Müller"
Does this also entail to remove NodeSeq (and move the methods to Node)? What about attribute values then?

It's entirely possible that I've "gone native" after working with traditional XML APIs for too long, but surely it's not controversial that some kinds of Node might "have-a" sequence of child nodes, others might not, but no Node "is-a" sequence of anything. 

I agree with the majority opinion in scala-internals that NodeSeq and the relationship Node <: Seq[Node] need to go; they're just too confusing, no matter how well they might map to native XML constructs.  Implicits can mop up the grey areas and resugar any conveniences we once inherited from dear old Grandad.

As to the specific design of whatever comes next, Paul, if you have any work in progress in this area I'd love to see it.  Otherwise, let's get started hashing out the design criteria.

Are there any axiomatic design decisions that we need to get out of the way first?  I think there's consensus on:

- Node will not inherit from NodeSeq or Seq[Node]

-0xe1a

Copyright © 2012 École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland