Release 1.1

[ballsteve]

It's time to prepare the next release - 1.1. The namespace redesign is pretty massive, so it needs to get out there. Also, we should keep the release cadence as short as possible.

Let's bring everything into the dev branch, make sure all warnings are eliminated, and documentation is complete.

[Devasta]

Sounds good. I need to check if we have a fix for issue 98 now that the namespaces are updated, will try to get to it ASAP.

[ballsteve]

There's one test failing in the dev branch - conformance::xml::ibm_valid::ibmvalid_p66ibm66v01xml

Once that's fixed we can get the release out.

[Devasta]

I only have access to a Windows laptop, but I can dig out an older box and put Linux on it for testing on that if needed.

[ballsteve]

I have a Mac laptop, so Unix O/S, and a Linux server so no problem with access to hardware & software.

[ballsteve]

I think I've found the problem: get_canonical translates \r\n into \n. This is being applied to the test document (P66/ibm66v01.xml) but not the output document (P66/out/ibm66v01.xml).

Why this is passing on a Windows box confuses me: it should fail everywhere.

Q: why is this translation being performed on output? Shouldn't it only be done on input?

[ballsteve]

Just checked the XML and Canonical XML specs.

XML (fifth edition), 2.11, says that CRLF should be translated to LF on input.

Canonical XML, 3.4, says that special characters, such as should be encoded as a character entity on output.

The change to the output is easy, and I'll do that tomorrow. I'd have to investigate the fix to the input parser.

We should add some additional tests for these, if there aren't already tests in the test suite.

[Devasta]

Grand, I'm on another business trip during the week but I can get to it at the weekend if needed.

[ballsteve]

Release 1.1 is now live!