Shibboleth Developers

[Will Norris <>]

On May 16, 2006, at 11:37 AM, Brendan Bellina wrote:

> The difference is that v4 was built against the release version of  
> Shibboleth 1.3.  Version 3 was not.  There were no other changes.

to expand on that... I was building the prior version against cvs  
HEAD, which for the most part WAS Shib 1.3, but now that a bit of  
Shib 2.0 stuff is being checked in that is no longer the case.  In  
v4, I built it from a clean checkout of  Rel_1_3_FINAL_C.  As a  
result, the patch no longer includes any blackbox tests, since all of  
that stuff was added after 1.3 was released.

I was also generating the patch file from within eclipse previously,  
which meant it was a little more difficult to apply the patch from  
the command line (at least for me).   It should be more easily  
applied from the command line now.


> For right now this is a patch we are offering to the community.  I  
> don't know what is involved in getting it to be adopted into the  
> official release.  We are not averse to that happening though.   
> There has been some discussion between Will Norris, Walter Hoehn,  
> and myself as to whether there is benefit in altering the syntax to  
> distinguish it somewhat from attribute release syntax, just to  
> avoid possible confusion, but no decision has been reached.

Yes, we're working with Walter on checking this in to CVS for  
inclusion in Shibboleth 2.0.  We are still discussing the ARP syntax  
to try and find the simplest and least confusing method.  If we do  
end up changing the syntax at all, we may release a v5 of the patch  
to use whatever new syntax we decide on... we'll see.


-will

Attachment: smime.p7s
Description: S/MIME cryptographic signature