Tyler Gregg

I've opened the above-linked feature request within the ion-java repository. This functionality isn't currently provided, but the use case is pretty easy to understand, so we (the Ion maintainers) should...

#22 adds coverage for presence of the DEL (0x7F) character in clobs.

#35 adds coverage for LST append. #33 and #34 add NOP padding coverage. #30 adds coverage for floats with underscores. #31 adds coverage for symbol zero. Removing those from the...

@toddjonker You should have access now. Can you try again?

All three variants are "correct", in that when the text is known, nothing else matters. Is this a concern about consistency across the reader options?

We need to clarify the spec. Overpadded zero values on type code 2 (e.g. 21 00) should be allowed. I've created https://github.com/amzn/ion-docs/issues/93.

Right. There are currently a few different test files that include shared symbol tables (which, as you say, would be interpreted as user data under the current test semantics) and/or...

> So is it embedded_document:: or $ion_embedded_whathever:: now? It's still `embedded_documents`. The linked ion-test-driver spec includes the implicit proposal to change it to `$ion_embedded_streams` (because reserved Ion annotations are supposed...

For the second-to-last sequence, the final string contains the UTF-8 code units `\xF0\x90\x80\x80`, and in the last sequence the final string contains the code units `\xF4\x8F\xBF\xBF`. Both of these are...

Right, the escapes represent code points. Once you are un-escaping the escape sequences, you are just calculating the Unicode (i.e. not UTF-8 or UTF-16 or any other encoding) code point,...