Re: XML 1.0 / XML Namespaces and ALM

I will let Henry and Philippe decide what to do. I would just make a few 
observations.

In 2022, Norm Walsh [personal communication] pointed out that the Fifth 
Edition's "NameStartChar production ... contains absurd things like 
#xFEFF, the zero-width no-break space. And #xFF1C, fullwidth less-than 
sign. Except parsers don’t accept those." For his work on InvisibleXML, 
Norm has made an exploration of this space documented at [1].

When we (the XML Core WG) opened up the character set for names in the 
Fifth Edition, we had push back from several quarters saying they 
weren't going to change their implementations to do that. Apparently, 
that sentiment continues.

I do note the (non-normative) Appendix J [2] does make a variety of 
suggestions for XML names; the first suggestion there relates to the 
issue of name start characters.

paul

[1] 
https://codeberg.org/InvisibleXML/ixml/src/commit/19bd1ed99b3568930c765cc675f7122a355e3e9c/misc/understanding-names.md

[2] https://www.w3.org/TR/REC-xml/#sec-suggested-names

On 2026-08-07 22:08, MURATA Makoto wrote:
> FYI: XML 1.0 fifth edition is still not supported by many 
> implementations, including Xerces-J and MSXML.
>
> Regard,
> Makoto
>
> 2026年8月7日(金) 23:20 Henry S. Thompson <ht@inf.ed.ac.uk>:
>
>     Addison Phillips writes:
>
>     > Some time ago (TPAC 2024??), I was tasked with reporting a gap
>     in XML
>     > 1.0 5e and Namespaces in XML 1.0 3e.
>     >
>     > The issue is documented here:
>     > https://github.com/w3c/i18n-activity/issues/1903
>     >
>     > The problem is that Unicode assigned a new bidirectional formatting
>     > character, U+061C ARABIC LETTER MARK, in 2013. This potentially
>     affects
>     > the grammar of NameStartChar or at least merits an errata mention.
>
>     Thanks to you for picking this up, and to Philippe for getting the
>     mechanisms refreshed.
>
>     I'm certainly available to draft an erratum, but will wait for
>     Philippe to suggest what process we should use to get review/approval
>     and to publish.
>
>     ht
>     -- 
>            Henry S. Thompson, School of Informatics, University of
>     Edinburgh
>                     10 Crichton Street, Edinburgh EH8 9AB, SCOTLAND
>                                e-mail: ht@inf.ed.ac.uk
>                           URL: https://www.ltg.ed.ac.uk/~ht/
>      [mail from me _always_ has a .sig like this -- mail without it is
>     forged spam]
>