Metadata-Based Naming?

[Scott Cantor]

> On Jul 6, 2026, at 8:38 AM, Rod Widdowson <rdw at steadingsoftware.com> wrote:
> 
>>> How the heck did XML validation let this pass? "xmllint" makes no complaint about it.
>> 
>> I doubt xmllint is even namespace aware, but I don't know offhand. It may require a special flag for that.
> 
> The more interesting question is why you didn't get huge warning on reload from the metadata resolver
> 
>    <saml2:AttributeValue>sn last_name</saml:AttributeValue>
> 
> That should have jumped right out before you had to turn up logging anywhere else...

Ah, I didn't even spot it. Yeah, I'd expect xmllint to flag that as not-well-formed, and I'm sure the reload would.

I suppose I haven't ever tried something like that with both prefixes defined (presumably to the same namespace). I can't imagine that's well-formed, but I wouldn't bet my life savings on it.

-- Scott