Subflow input mappings are lost in transitions

[Misagh Moayyed]

Yes, true and it's precisely because of those "limitations" that input
mappings are used I guess. You are right that both the caller and the
callee flows need to declare input parameters that they pass and require.
I had made declarations on both ends, and yet I was still not receiving
any input on the callee side, because invocations specifically ignored
everything the caller had passed along.

...which is fine if I am to use the PRC anyway. 

Even if this were an actual issue, I very much doubt it could be solved
via pure XML config. You'd have to rewrite the entire flow declaration via
Java beans or something to be able to handle that type of dynamic!

> -----Original Message-----
> From: dev [mailto:dev-bounces at shibboleth.net] On Behalf Of Cantor, Scott
> Sent: Monday, August 24, 2015 9:09 AM
> To: Shib Dev <dev at shibboleth.net>
> Subject: Re: Subflow input mappings are lost in transitions
> 
> Also, note that flow scope does not span subflows anyway (conversation
> scope does).
> 
> -- Scott
> 
> --
> To unsubscribe from this list send an email to dev-
> unsubscribe at shibboleth.net