Tatu Saloranta

@bollmann Ah. I agree, requiredness is something that should be checked as per schema. I missed that part. A simple unit test would be good, it can initially be added...

No, Jackson simply binds `ByteBuffer`s to/from binary data. In theory polymorphic type handling could help, but I suspect that if this is something you really need to retain, custom (de)serializers...

Sounds like a bug, yes. If anyone has time, the first step could be to write a (failing) unit test to reproduce the problem in its smallest form, and help...

Thank you for the report! I suspect there may be an issue with linefeed handling within comments inside DTD; location counters need explicit handling (due to multiple LFs XML allows),...

(added "test-needed" label just because automated test would be great -- I realize instructions for writing test are included above)

@adamretter I do not remember how it works with Woodstox SAX API; I'd have to dig into sources to remember. But mechanism supported are just the basic SAX API (and...

Hmmh. Indentation mentioned was meant for for element indentation, and not for changing possible white-space between attributes. And `writeSpace()` specifically would only produce cdata sections, which inter-attribute white space would...

We are always open for patches: `writeRaw()` I think should not modify the state -- and should be usable for this usage -- for sure. Whether `writeSpace()` should work that...

Good point; it is bit of ambiguous case since without flushing like that possibly existing use cases could break. Not sure how to proceed here, however, since as a practical...

Thank you @dan2097.