Heimdal/MIT API convergence

Sam Hartman hartmans at MIT.EDU
Thu Aug 8 13:54:01 EDT 2002

>>>>> "Jason" == Jason Garman <jgarman at wedgie.org> writes:

    Jason> (Somewhat off topic...) Personally, I'd like to see better
    Jason> coordination between MIT and the Heimdal group on APIs,
    Jason> since its annoying to constantly encounter software that
    Jason> will compile with one and not the other.  There's also a
    Jason> few instances where error messages from the KDC on one will
    Jason> be misinterpreted by libraries for the other.
I think that for future development you're likely to see this;
certainly when adding functionality to MIT Kerberos I do look at
whether Heimdal already has an API and consider whether adoption of
that API would be a good idea.

We'd like to actually work on API unification; the Heimdal developers
have expressed a desire to do this in the past as well.

Neither team really has time for this.  Also, the developers working
on the different products seem to have a different idea about what the
API should feel like in certain areas.  There are several cases where
Heimdal has added helper APIs that perform the functions of multiple
MIT APIs; in some cases I'm not really I see the justification as
worth the divergence.

But if someone wanted to throw resources (qualified developer time
would be more useful than money) at the problem, you would likely see

MIT does see Heimdal interoperability at the protocol level as a
critical issue; based on responses to our concerns I believe the
Heimdal people do as well.

More information about the krbdev mailing list