[2182] in Kerberos-V5-bugs
Re: non-sized types in kadm5
daemon@ATHENA.MIT.EDU (Ken Raeburn)
Sat Aug 24 03:50:15 1996
To: "Barry Jaspan" <bjaspan@MIT.EDU>
Cc: krb5-bugs@MIT.EDU
From: Ken Raeburn <raeburn@cygnus.com>
Date: 24 Aug 1996 03:49:47 -0400
In-Reply-To: "Barry Jaspan"'s message of Wed, 21 Aug 1996 13:52:24 -0400
"Barry Jaspan" <bjaspan@MIT.EDU> writes:
> But the right solution is to used sized types everywhere in
> the kadm5 api that they matter
I agree.
> (and, IMHO, everywhere in every api,
> period).
Disagree. There are times when types appropriate to the architecture
are better. You might as well try to dictate integer endianness.
Use "int" when "int" is appropriate, and "int32" when "int32" is
appropriate. The library can do the translation, and detect
out-of-range values (which would include but not be limited to cases
where conversion from "long" to 4 octets could not properly represent
the value).
