: Why are we using libverto again

Sam Hartman hartmans at MIT.EDU
Thu Jul 7 14:04:41 EDT 2011


>>>>> "Greg" == Greg Hudson <ghudson at MIT.EDU> writes:

    Greg> On Thu, 2011-07-07 at 10:44 -0400, Sam Hartman wrote:
    >> Now, for the KDC today, we could just use a specific event library and
    >> gain significant complexity savings.

    Greg> The reason I favor an intermediary is that the event loop will be part
    Greg> of the API for kdcpreauth plugins (and probably KDB and authdata plugins
    Greg> in the future).  Picking an event loop now carries costs not just for a
    Greg> future libkdc but also a plugin interface transition cost.


I understand that.  And based on your description of a simple
intermediary, that made some sense.  However the complexity of the
intermediary has far exceeded what I was expecting.  I'll admit that my
original simple vision wouldn't have worked, but I think we should all
carefully consider what we got and whether it's worth the cost.

If everyone else had a more accurate estimate of the complexity when
they read your original proposal and I'm the only one who has been
surprised, then we should move forward with this plan.



More information about the krbdev mailing list