[Math Processing Error] with empty mmultiscripts in Native MathML mode

[fred-wang]

 <math><mmultiscripts></mmultiscripts></math>

leads to a [Math Processing Error] with the Native MathML mode. This is invalid MathML but normally MathJax should pass it to the browser and let the native MathML code handle the error.

[dpvc]

The issue486 branch of my fork of MathJax includes a fix.

[fred-wang]

OK, looks good. A quick look at the source shows that a similar processing error will happen with

 <math><munder></munder></math>
 <math><mtable><mlabeledtr></mlabeledtr></mtable>

so that would be good to fix them too while we are there.

[fred-wang]

Crashtests/issue486.html

=> In testsuite

[dpvc]

OK, I'll look into the other cases.

[fred-wang]

=> Ready for release.

[fred-wang]

It seems that closing the issue increase the progress in the "Milestone" bar. Should we leave open issues marked "merge" (I don't remember why we waited the last minute before closing everything in the past)

[dpvc]

The reason we had left them before was that it was easier to see at a glance what was actually going to be part of v2.2. This was before we had the labels for versions. Now that we have milestones, that probably isn't necessary. I think it is OK to close the issues when merged, and will start to do so.

[fred-wang]

What about the "Merged" label. Is it still necessary?

[pkra]

I'll create a "final" issue for the release (increasing versions, packing etc). That would prevent the milestone from closing automatically after technical bugs are closed -- and remind us to do the versions and packaging ;) .

[dpvc]

I guess the merged label isn't really needed.

For the final issue, please include registering with Bower.

[pkra]

Done.