I originally added a flavor because I though that was the standard way
of doing things. However, I completely agree that libssh2 is small and
self-contained enough to build in by default. Plus, it would be nice for
OpenBSD's port of Zabbix to "just work" with respect to SSH checks.

Patch below to include libssh by default:

Index: Makefile
===================================================================
RCS file: /cvs/ports/net/zabbix/Makefile,v
retrieving revision 1.127
diff -u -p -r1.127 Makefile
--- Makefile    8 Jan 2017 19:24:53 -0000       1.127
+++ Makefile    13 Oct 2017 16:47:27 -0000
@@ -46,12 +46,13 @@ RUN_DEPENDS-main =
 LIB_DEPENDS-server =   ${LIB_DEPENDS} \
                        textproc/libxml \
                        net/net-snmp \
-                       textproc/iksemel
+                       textproc/iksemel \
+                       security/libssh2
 RUN_DEPENDS-server =   net/fping
 # we want the _zabbix user&group as well as README-main
 RUN_DEPENDS-server +=  net/zabbix,-main>=${VERSION}
 WANTLIB-server =       ${WANTLIB} crypto ssl z netsnmp iksemel pthread \
-                       xml2
+                       xml2 ssh2

 CONFIGURE_STYLE =      gnu
 CONFIGURE_ARGS =       --enable-server \
@@ -62,7 +63,8 @@ CONFIGURE_ARGS =      --enable-server \
                        --with-libxml2 \
                        --with-net-snmp \
                        --with-ldap \
-                       --with-jabber="${LOCALBASE}"
+                       --with-jabber="${LOCALBASE}" \
+                       --with-ssh2
 CONFIGURE_ENV =                CPPFLAGS="-I${LOCALBASE}/include" \
                        LDFLAGS="-L${LOCALBASE}/lib -lexecinfo" \
                        ac_cv_search___res_query=no


───────────────────────────────
William Leuschner
wel2...@rit.edu

On 10/13/2017 4:23 AM, Stuart Henderson wrote:
> On 2017/10/13 07:11, Antoine Jacoutot wrote:
>> Is it worth adding yet another FLAVOR to this port? Can't we enable it by
>> default?
> 
> +1, libssh2 is self-contained and small.
> 

Reply via email to