Progress of 5/6 IDNABIS documents & mappings consistency issue

[Paul Hoffman]

At 3:00 PM -0800 12/20/09, Lisa Dusseault wrote:
>On the topic of mappings, I've previously noted
>(http://www.alvestrand.no/pipermail/idna-update/2009-December/005962.html)
>that we had some IETF Last Call issues raised about how compatible our
>idnabis-mapping list of mappings is with the mappings of IDNA2003.  I'd like
>to be sure there's WG consensus on these issues before progressing the
>mappings document.  If some links out there contain characters that were
>required to be mapped in IDNA2003, and software is upgraded to IDNA2008
>in a way that makes those links fail, that seems to be something we should
>be concerned about.  The key issues seem to me to be what the WG opinion
>is on having one set of optional mappings for better interoperability, or two
>or more to encourage flexibility. Michel Suignard's recent message
>(http://www.alvestrand.no/pipermail/idna-update/2009-December/006389.html)
>seems like a good starting point for that discussion.

This may be a mischaracterization of the WG's intentions for the mappings document; or, it may be exactly right. Given that the document was created after the WG had already made many, many decisions on direction, it is really hard to say what the mappings document is supposed to be. My feeling (as co-author, but definitely not as WG chair or responsible AD) is that the main thing the WG wants the mappings document to say is "you can do mappings however you want". It feels like the WG also wants a simple example of a mapping scheme, but nothing definitive.

Before we start monkeying with the text in the document, we should have consensus on what the WG actually wants. If there is no rough consensus on that, we can't move forwards.