[1972] in Kerberos-V5-bugs
Re: passing arg 3 of `gss_import_name' discards `const' from pointer target type
daemon@ATHENA.MIT.EDU (Theodore Y. Ts'o)
Fri Jun 7 18:38:14 1996
Date: Fri, 7 Jun 1996 18:38:01 -0400
From: "Theodore Y. Ts'o" <tytso@MIT.EDU>
To: Marc Horowitz <marc@MIT.EDU>
Cc: "Barry Jaspan" <bjaspan@MIT.EDU>, krb5-bugs@MIT.EDU
In-Reply-To: Marc Horowitz's message of Thu, 06 Jun 1996 17:44:22 EDT,
<9606062144.AA27769@opus.MIT.EDU>
Date: Thu, 06 Jun 1996 17:44:22 EDT
From: Marc Horowitz <marc@MIT.EDU>
This issue is based on how you interpret the gssapi c bindings spec.
Ted chooses to take the spec literally, and refuses to change
anything. I chose to insert const's where appropriate. These const's
in the function (but not constant) declarations were in the OV code,
but removed by MIT at some point.
Yes, I like to follow the spec, and not take liberties with it. The
extra const types also cuased all sorts of GCC warnings, which I didn't
like. The next time we do this, I think we need to take more care
about how the const types are sprinkled into the headers.
IMHO, the thing to do is fix the gssapi function prototypes, and make
sure const is allowed in appropriate places in the next c bindings
I-D.
We can argue this on the list, but I don't believe const should be
optional; it either should be always there, or never there.
- Ted
