| > So '---' is not a valid operator symbol, but '-->' is. A line
| > of hyphens of
| > any length introduces a comment.
| >
| >
| > ] I do not understand the example: if every lexeme consisting of two
| > ] or more hyphens begins a comment, `-->' begins a comment!
|
| No, '-->' does not consist of two or more hyphens; 'consist of' means
| 'contains only', and there's a '>' in the lexeme. Is that clearer?
Aah, yes, That's clearer. D'accord.
| > Allow a type and a class to have the same name
| >
| > Rejected. It's an un-forced change, and it allows even more
| > obscure programs
| > than now. Data constructors and type constructors can share
| > the same name,
| > but data constructors appear only in expressions, and type
| > constructors only
| > in types, so there's no confusion. But classes appear in types too.
| >
| > ] No, no, no! Why on earth should Haskell 98 dictate the
| > choice of names?
|
| Interesting! Several people have spoken up about this one, and I don't
| feel very strongly either way, so I'm open to persuasion. Just to
| articulate
| the alternative, I'm proposing that if types and classes can have the same
| name, then in export and import lists one would say
| class C
| for classes, but simply
| T
| for types (as now). I don't want to get into whether T is a type
| synomym, a data type, or a newtype; if we say 'type T' it might be
| misleading.
|
| How does that sound?
Sound's good. Needless to say I vote strongly for this alternative.
Cheer, Ralf
