This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
The casting to string and untypedAtomic still contains the two sentences "The canonical form of positive zero is "0.0E0". The canonical form for negative zero is "-0.0E0". " Since earlier rules handle these two cases, these two lines can be deleted.
I'm sorry but I disagree. The canonical form for float/double is defined in this paragraph. The forms for positive and negative zero are specially called out as these are special values. Smart users may be able to infer these forms from the preceding general rules but a bit of redundancy does no harm.
Sorry, but there is currently a contradiction. xs:string(-0e0) should return "-0" and not "-0e0" according to the earlier parts of the spec. Talking about the canonical form in that place is misleading and contradicts the earlier casting rules.
Now, with the fuller explanation, this makes sense. Fixed! Ashok Malhotra