Noted while following up on misc@ post from Pavel Pocheptsov. Bump all the
subpackages for safety though in theory only -cgi should be affected.
While there tweak pre-configure to allow 'make configure' to be re-run
after interruption.
ok ajacoutot@
- move MESSAGE to README and flesh it out a bit
- fix libwrap detection in autoconf
- install the avahi service definition to the right place again; this got
broken when moving to using SYSCONFDIR overrides / BASESYSCONFDIR
- Move nut-scanner to a separate package, not everybody wants to
scan their nuts and it pulls in various deps not wanted for the
main package.
- Add missing dep on avahi (for the scanner only) and install
static service configuration files.
ok ajacoutot
- thanks ajacoutot@ for looking over the scripts
- note that there has been some major work on apcsmart; in the event
of problems the old driver is still available as apcsmart-old
devices, including APC 5G, various Eaton, HP, and Tripp Lite models,
and a number of USB HID devices)
- build the snmp package by default
- use correct user/group names in MESSAGE
* base the default maxconn on available filehandles.
* change the error message to better indicate that it's an error
not a warning.
ok jasper@
while there, remove an out-of-date comment in MESSAGE-main.
- upsimage.cgi used upsimage_cgi_LDFLAGS to tweak for -all-static,
and not *_LDADD as upsstats.cgi and upsset.cgi did. as a result
upsimage.cgi was linked dynamically, therefore broken inside chroot
- add USE_GMAKE, so that it builds on landisk
- make lib-depends-check happy
- add @bin marker for bin/snmp-ups
- bump all PKGNAMEs
ok sthen@
(now called agiler-old)
upses such as the below can now work again:
{0x0665, 0x5161, &agiler_old_subdriver}, /* Belkin F6C1200-UNV */
{0x06da, 0x0003, &agiler_old_subdriver}, /* Mustek Powermust */
{0x0f03, 0x0001, &agiler_old_subdriver}, /* Unitek Alpha 1200Sx *
ok sthen@
preserve my changes to the fentonups driver, allowing the detection logic
to be bypassed and a model to be forced - for UPSes that do implement the
protocol except the identification command, which seems to be somewhat common
the Xanto S3000R thing here behaves like that, and support for it is added
ok mbalmer
the FULLPKGPATH, thus providing changes to packing-lists which shouldn't
happen, and making update more difficult.
Accordingly, bump all pkgnames with PSEUDO_FLAVORS, and provide an
update @pkgpath for the bug for most of them (left out the ones with 3
or 4 pseudo flavors for space constraints...)
