[238] in Kerberos-V5-bugs
problems with KRB5 beta 2
daemon@ATHENA.MIT.EDU (Mark Frost)
Wed Oct 14 18:34:46 1992
From: mfrost@shield.eng.pyramid.com (Mark Frost)
Date: Wed, 14 Oct 1992 15:33:39 -0700
To: krb5-bugs@Athena.MIT.EDU
Hi there. I've been working on building the kerberos V5 beta 2 release on
Pyramid's DC/OSx platform (SVR4-based) for the last few days. Some of the
problem I've found so far are:
- putting
#define HasStringH YES
in the "system".cf file has no effect since there's nothing in any of the
code to actually check for this (eventually I see that this gest handed down
to the "USE_STRING_H" symbol -- nowhere in the code is this checked for).
I had a lot of problems figuring out how to put this capability in to the
code itself. I finally gave up and just changed all the "<strings.h>"
lines to "<string.h>" by hand.
- our system (and I suspect a lot of other "almost POSIX" systems) use
"size_t" as a return type from strlen() and as the parameter for malloc().
It might be nice to make this a configuration option.
- we also had to do some kinda gross hacks to make the dbm stuff work (to pull
in the ucb libs where dbm resides), but that's more like an SVR4 problem than
yours.
- I've gotten several "macro redefined: rindex" messages during the build.
Apparently one of them comes from "general.h" in the isode tree. I'm
not sure where the other one comes from.
I'm still working on getting it built.
Thanks
-m----------- Mark "Hoek" Frost (mfrost@pyramid.com)
---mmm--------- R&D Lab Manager Group
-----mmmmm------- Pyramid Technology Corporation
-------mmmmmmm----- 3860 North First Street
---------mmmmmmmmm--- San Jose, California 95134 (408) 428-8163
