OpenSAML user discussion

["Scott Cantor" <>]

> I debated whether the transforms should be directly manipulatable.  Per
> the SAML spec, the transforms SHOULD consist only of the enveloped
> signature transform and exclusive canonicalization transform (with or
> without comments), but it's not a MUST.  I suppose we might revisit and
> make the only transforms option a boolean toggle of with/without comments
> for the exclusive c14n.

I stuck with a fixed set, but made the c14n algorithm settable. If you
really need a custom transform set, that really should be a different
reference object, IMHO, but it doesn't really matter that much.

Theoretically, we should probably block the use of that object with SAML 1.0
objects, though. It will just lead to errors (per the other note).

-- Scott