[comp.mail.sendmail] Say goodbye to sendmail.cf files!

barnett@grymoire.crd.ge.com (Bruce Barnett) (02/25/91)

I have posted a preliminary version of Ease 3.0 to alt.sources.

Ease is a high level language for sendmail.cf files. 

First the Bad News:
	I am not the original author of Ease. I made extensive
modifications to both Ease and Cfc, which is asking for trouble.  I also
haven't tested it on any operating system other than SunOS and Vax
Ultrix. There may be some portability problems and memory leaks.
The authors have given me permission to become the keeper of the source.

Now the Good News:
	It supports Sun, DEC, and HP versions of sendmail. It supports
99.9% of the IDA sendmail. You can convert your sendmail.cf file into an
ease file in minutes. The translation of Ease back into sendmail is
close to perfect. I must have made 150 modifications and enhancements
to the previously released version.

There are some known problems:

	1. The code doesn't work when compiled with the -O flag in SunOS.
	2. Don't type "make clean" - 
	3. Some people have reported problems with MIPS machines.

I appreciate any patches and bug reports. With your help, I hope to
make this package an indespensible tool.
--
Bruce G. Barnett	barnett@crd.ge.com	uunet!crdgw1!barnett

kjj@varese.UUCP (Kevin Johnson) (02/27/91)

In article <BARNETT.91Feb24231516@grymoire.crd.ge.com> barnett@crdgw1.ge.com writes:
>I have posted a preliminary version of Ease 3.0 to alt.sources.
>[MEAT DELETED]
>There are some known problems:
>
>	1. The code doesn't work when compiled with the -O flag in SunOS.
>	2. Don't type "make clean" - 
>	3. Some people have reported problems with MIPS machines.
>
>I appreciate any patches and bug reports. With your help, I hope to
>make this package an indespensible tool.

I've been trying my hand at porting it to SYSV...
et seems to memory faulting...
I'm trying to localize the source of the error - not having much luck yet...
If anyone can lend some insight into the problem - it would be appreciated.

#include <standard_disclaimer>
'"Arf", she said.'
	- Mr. Frank Zappa
.-----------------------------------------------------------------------------.
| Kevin Johnson                     kjj@phx.mcd.mot.com | ...!mcdphx!QIS1!kjj |
| QIS System Administrator  Motorola MCG                                      |

bent@lccinc.UUCP (Ben Taylor) (02/28/91)

kjj@varese.UUCP (Kevin Johnson) writes:

>In article <BARNETT.91Feb24231516@grymoire.crd.ge.com> barnett@crdgw1.ge.com writes:
>>I appreciate any patches and bug reports. With your help, I hope to
>>make this package an indespensible tool.

>I've been trying my hand at porting it to SYSV...
>et seems to memory faulting...
>I'm trying to localize the source of the error - not having much luck yet...
>If anyone can lend some insight into the problem - it would be appreciated.

First, a couple of problems looking for <strings.h>.  This is in C files 
as well as HARD-Coded into the makefile dependencies.  WHY, WHY, WHY are
the standard include files in the dependency list in the makefile?
Define SYSV in the makefiles, and remove the /usr/include/strings.h from
the dependency lines in the makefiles.

Whereever <strings.h> is used in c files,
#ifdef SYSV
#include <string.h>
#else
#include <strings.h>
#endif


Second. index/rindex are generally bsd calls.  Whereever index/rindex
show up add:

#ifdef SYSV
#define index strchr
#define rindex strrchr
#endif

For src/fixstrings.h.  Try the following to handle the second level
of indirection this is causing.

#ifdef SYSV
#define strchr Xindex
#define strrchr Xrindex
#else
#define index Xindex
#define rindex Xrindex
#endif

I quickly hacked this out to get it working under UHC SYSVR4 with about
an hours worth of work.  I will be working on getting it running under
Xenix next.  Hope this helps.

>| Kevin Johnson                     kjj@phx.mcd.mot.com | ...!mcdphx!QIS1!kjj |
>| QIS System Administrator  Motorola MCG                                      |

Ben Taylor
Systems Administrator
LCC Incorporated
uunet!lccinc!bent