4 Apr
2005
4 Apr
'05
8:25 a.m.
Decoding such malformed surrogate sequences does not affect the ability of an application to treat ASCII characters without decoding UTF though, which if I understood correctly was the only reason for the extra stringency in the decoder...