Daylight savings time change strangeness

[Cantor, Scott]

> I think I know the answer to this but I thought that the Shibboleth IdP would
> use UTC to send the IssueInstant and NotOnOrAfter timestamps. Isn't that
> supposed to be Daylight savings time immune?

It's not immune to a server that's operating an hour off, it's just an unambiguous way to express the time.

> We seem to have a vendor
> that change and hour behind us and it would not allow authentication for
> about an hour. Would they have to be translating the time if they wanted to
> use local?

Everything is always translating to compare to their current time.

-- Scott