Should <start-of-line> match after a byte order mark?

[andrewhickman]

Consider these two documents:

• https://play.yaml.io/main/javascript?input=77u_IyBjb21tZW50
• https://play.yaml.io/main/javascript?input=77u_fApmb28=

The reference parser doesn't consider them valid unless a separating space is added after the byte order mark, which seems weird. I wouldn't expect the presence of a BOM to affect anything other than the chosen encoding.

Perhaps it should be specified that <start-of-line> matches after a BOM, so that no separator is required?