antonlyap/nut-debian
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
9675a8feb9
nut-debian/ChangeLog
arnaud.quette@free.fr 26fb71b504 Imported Upstream version 2.4.3
2010-03-26 00:20:59 +01:00

1092 lines
42 KiB
Plaintext
Raw Blame History

2.4.3
2010-02-23 Arnaud Quette <arnaud.quette@free.fr>
* [r2366] ChangeLog, configure.in, NEWS, UPGRADING: Final update for
2.4.3 release
2010-02-20 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2365] m4/nut_check_ipv6.m4: Fix broken logic to enable IPv6
support
2.4.2
2010-02-19 Arnaud Quette <arnaud.quette@free.fr>
* [r2363] ChangeLog, configure.in: Final update for 2.4.2 release
2010-02-19 Arnaud Quette <arnaud.quette@free.fr>
* [r2362] NEWS, data/driver.list: Add HP R1500 G2 to compatibility
list for bcmxcp
(reported by Stephan Schupfer)
2010-02-19 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2361] clients/upsmon.h: Fixed typo
* [r2360] clients/upsmon.c, clients/upsmon.h: You can't compare
strings with '!=' in C, so the logic to determine if the stock
message was overridden was flawed (causing problems with
(de)allocation of memory)
2010-02-19 Arnaud Quette <arnaud.quette@free.fr>
* [r2357] conf/nut.conf.sample, conf/upsmon.conf.sample, docs/FAQ,
man/upsmon.conf.5: Remove remaining references to ACL /
allowfrom.
2010-02-19 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2355] configure.in: Replace __func__ on pre-C99 compilers with
__FUNCTION__ (may be available on older GNU C compilers) or
__LINE__ (which is always available)
2010-02-16 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2353] drivers/main.h: Static initialization of flexible array
members is non-portable (GNU C extension), so don't use a
flexible array member here
* [r2352] configure.in: Check if the C compiler supports flexible
array members and/or variable-length arrays
* [r2350] drivers/upscode2.c: Don't use signed constants in a
bitfield
2010-02-15 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2349] drivers/main.h: Fix missing semi-colon
* [r2348] drivers/main.h: Attempt to fix compiler warning on
Solaris (although I have my doubts about the portability of array
declarations within structures)
* [r2347] drivers/snmp-ups.c, drivers/snmp-ups.h: Fix for buggy
Net-SNMP config should be before including headers, since our
autoconf generated header is the first to be included (so the
conflict is caused by subsequently including the Net-SNMP
headers)
* [r2346] drivers/apcsmart.c, drivers/metasys.c, drivers/serial.c:
Fix unreachable code warnings
* [r2345] drivers/upscode2.c: An (int) is only guaranteed to be at
least 16-bit while we need 32-bit here (so it's better to use a
fixed width integer)
2010-02-14 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2344] configure.in, docs/configure.txt, drivers/powerman-pdu.c,
m4/nut_check_libneon.m4, m4/nut_check_libnetsnmp.m4,
m4/nut_check_libpowerman.m4, m4/nut_check_libusb.m4: Allow to
override CFLAGS and LDFLAGS detected by the *.m4 macros in case
pkg-config is not available
* [r2343] docs/configure.txt, m4/nut_check_libssl.m4: Allow cflags
and ldflags overrides for openssl (may be useful/needed if
pkg-config is not available)
2010-02-13 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2342] drivers/liebertgxt2.c: Use (unsigned char) instead of
(char) in command definitions (to prevent overflow warnings
definitions)
* [r2341] drivers/mge-shut.h: The usage code must be at least a
32-bit integer (and 'int' is only guaranteed to be at least
16-bit).
* [r2340] m4/nut_check_libnetsnmp.m4: Check if Net-SNMP library is
usable
2010-02-12 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2339] docs/cables/apc.txt: APC cable diagrams as reported on
the mailinglist
[http://lists.alioth.debian.org/pipermail/nut-upsuser/2005-August/000118.html]
2010-02-11 Charles Lepple <clepple@gmail.com>
* [r2338] ChangeLog: Generate ChangeLog entries from r2337 through
r2331.
2010-02-11 Arnaud Quette <arnaud.quette@free.fr>
* [r2337] NEWS: Complete news information for the 2.4.2 release.
2010-02-11 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2336] (numerous files) No functional changes (properties only)
* [r2335] AUTHORS: No functional changes (properties only)
2010-02-10 Arnaud Quette <arnaud.quette@free.fr>
* [r2334] ChangeLog: Trimmed 2.4.0 entries.
2010-02-10 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2333] m4/nut_check_ipv6.m4: Check for presence of structures
required for IPv6 support in the header files that are defined by
POSIX (don't rely on the netdb.h header to include them)
2010-02-10 Charles Lepple <clepple@gmail.com>
* [r2332] ChangeLog: Generate ChangeLog entries from SVN r2331 back
through r2117.
2010-02-10 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2331] drivers/genericups.c: Fix typo in previous commit
* [r2330] drivers/genericups.c: Don't hang up on last close
(initial state of HUPCL is implementation dependent)
[http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=327072]
2010-02-09 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2329] drivers/main.c: Some drivers set
'driver.version.internal' in upsdrv_initinfo() so we should set
the generic message before running that
* [r2327] man/upssched.conf.5: Fix wrong reference
[http://lists.alioth.debian.org/pipermail/nut-upsdev/2010-February/004532.html]
* [r2326] server/netuser.c: Check if user is allowed to login
before accepting connection
[http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=544390]
2010-02-08 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2324] UPGRADING: The tcp-wrappers support is optional in
addition to the LISTEN directive (not the other way around)
2010-02-08 Arnaud Quette <arnaud.quette@free.fr>
* [r2323] UPGRADING: Add a double explicit note about the ACL
mechanism removal in 2.4.0.
(reported on Debian by Anthony DeRobertis, Bug #526811)
2010-02-08 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2322] data/driver.list: Added lots of Ippon devices that were
mentioned on the mailinglists (serial devices will undoubtly
work, USB support still experimental)
2010-02-07 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2321] UPGRADING, drivers/blazer_usb.c, man/blazer.8: Use
'ippon' subdriver for devices with USB id 06da:0003 (old
'phoenix' subdriver still available)
* [r2320] drivers/raritan-mib.h, drivers/snmp-ups.c: Fix wrong name
* [r2319] drivers/raritan-mib.h: Add missing file in previous
commit
* [r2318] drivers, drivers/Makefile.am, drivers/apc-mib.c,
drivers/apc-mib.h, drivers/apccmib.h, drivers/baytech-mib.c,
drivers/baytech-mib.h, drivers/baytechmib.h,
drivers/compaq-mib.c, drivers/compaq-mib.h,
drivers/cpqpowermib.h, drivers/eaton-aphel-mib.h,
drivers/eaton-mib.c, drivers/eaton-mib.h, drivers/ietf-mib.c,
drivers/ietf-mib.h, drivers/ietfmib.h, drivers/mge-mib.c,
drivers/mge-mib.h, drivers/mgemib.h, drivers/netvision-mib.c,
drivers/netvision-mib.h, drivers/netvisionmib.h,
drivers/powerware-mib.c, drivers/powerware-mib.h,
drivers/pwmib.h, drivers/raritan-mib.h,
drivers/raritan-pdu-mib.c, drivers/raritan-pdu-mib.h,
drivers/snmp-ups.c, drivers/snmp-ups.h: Extract subdriver info
from main driver body into separate modules. This prevents
namespace conflicts and ensures that changes in one subdriver
don't lead to surprise changes in others.
It also allows to use a single mapping file to be used in several
subdrivers by specifying a different entry point (which doesn't
work if everything is in one module).
2010-02-07 Charles Lepple <clepple@gmail.com>
* [r2317] drivers/apcsmart.c, drivers/apcsmart.h: apcsmart: Add
input.transfer.reason
Contributed by Michael Haardt:
http://article.gmane.org/gmane.comp.monitoring.nut.devel/4498
2010-02-05 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2316] drivers/cpqpowermib.h: Could this be a Powerware OEM
device?
2010-02-04 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2315] drivers/cpqpowermib.h: Based on patch from Philip Ward
[http://lists.alioth.debian.org/pipermail/nut-upsdev/2010-February/004509.html]
2010-02-01 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2305] clients/upsimage.c, common/strerror.c, drivers/bcmxcp.c,
drivers/main-hal.c, drivers/powerman-pdu.c: Prefer snprintf over
sprintf to prevent overflowing buffers (also removes trailing
whitespace)
* [r2303] drivers/Makefile.am, drivers/cpqpowermib.h,
drivers/snmp-ups.c, man/snmp-ups.8:
2010-01-31 Charles Lepple <clepple@gmail.com>
* [r2297] drivers/tripplite_usb.c: Revert trunk commit r2294, which
was meant for a branch.
* [r2296] drivers/clone-outlet.c: clone-outlet.c: remove executable
bit
* [r2294] drivers/tripplite_usb.c: tripplite_usb: initial changes
to test protocol 3004
2010-01-30 Arnaud Quette <arnaud.quette@free.fr>
* [r2286] tools/Makefile.am, tools/device-recorder.sh: Create a
script to record device running sequence and dump it in a .seq
format.
The .seq file can then be used by dummy-ups to replay the
sequence.
2010-01-28 Charles Lepple <clepple@gmail.com>
* [r2282] m4/nut_report_feature.m4: Another attempt to fix the
feature report on OS X.
2010-01-27 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2281] m4/nut_report_feature.m4: Attempt to get rid of spurious
'-n' on MacOSX-10.5 BuildBot
* [r2280] m4/nut_check_libhal.m4: Prefer to use 'test' over '[ ]'
(on many shells, this first a built-in function while the second
is a symbolic link to 'test')
2010-01-27 Charles Lepple <clepple@gmail.com>
* [r2279] m4/nut_check_libhal.m4: Remove fallback HAL information
path, and remove extra parentheses.
2010-01-25 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2278] drivers/megatec.c: Don't check reliability of connection
at driver startup. This confuses users and really serves no
purpose for normal use. In case the connection is unreliable, it
should ultimately be the decision of the user whether or not it
is acceptable (not the driver).
* [r2277] drivers/apc-hid.c, drivers/belkin-hid.c,
drivers/cps-hid.c, drivers/explore-hid.c, drivers/liebert-hid.c,
drivers/mge-hid.c, drivers/powercom-hid.c,
drivers/tripplite-hid.c, include/common.h: The "config.h" header
should be the first included, since it contains the _GNU_SOURCE
declaration (enables Posix extensions to C)
* [r2276] configure.in: Add AC_USE_SYSTEM_EXTENSIONS to use
features of Posix that are extensions to C (requires Autoconf
2.60 or better)
* [r2275] common/common.c: Fix compiler warning on MacOSX-10.5
* [r2274] common/common.c: Fix compiler warning on FreeBSD
* [r2273] configure.in:
* [r2272] configure.in, drivers/mge-hid.c: Check for presence of
strptime() before using it
2010-01-24 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2271] configure.in, drivers/mge-hid.c: The previous patch was a
resounding failure (reverting back to setting _XOPEN_SOURCE in
one specific source file instead)
* [r2270] configure.in: Request SUSv2 compliance for all sources
(experimental)
* [r2269] drivers/mge-hid.c: _XOPEN_SOURCE seems to break on
FreeBSD
* [r2268] drivers/mge-xml.c: Fix compiler warning
2010-01-24 Charles Lepple <clepple@gmail.com>
* [r2267] drivers/mge-hid.c: Define _XOPEN_SOURCE=500 for
strptime() prototype in glibc2
2010-01-24 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2266] drivers/mge-hid.c: Removing stuff that was added to fix
strptime on Debian Etch (incompatible with FreeBSD)
2010-01-23 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2265] drivers/mge-hid.c: Fourth attempt of fixing strptime()
declaration missing on Debian
* [r2264] drivers/mge-hid.c: Third attempt of fixing strptime()
declaration missing on Debian
* [r2263] drivers/mge-hid.c: Second attempt of fixing strptime()
declaration missing on Debian
* [r2262] drivers/apc-hid.c, drivers/belkin-hid.c,
drivers/cps-hid.c, drivers/explore-hid.c, drivers/liebert-hid.c,
drivers/mge-hid.c, drivers/powercom-hid.c,
drivers/tripplite-hid.c, scripts/subdriver/path-to-subdriver.sh:
Remove redundant #include's (are being dealt with by main.h)
* [r2261] data/cmdvartab: Add descriptions for
'ups.timer.(start|reboot|shutdown)' variables
* [r2260] drivers/mge-hid.c: Need to define _XOPEN_SOURCE for
strptime() prototype in glibc2
2010-01-22 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2257] drivers/mge-shut.c, drivers/mge-shut.h: Change naming of
ups.delay.* variables to ups.timer.*, since writing these values
immediately starts the countdown (and not when a shutdown command
is issued)
* [r2256] drivers/powercom-hid.c: Fallback to stored value of
'ups.delay.(start|shutdown)' only if no parameter is given (to
fix 'load.on' command)
2010-01-21 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2255] drivers/powercom-hid.c: Fix beeper.status and add
load.on.delay command (to allow starting up with delay when UPS
is shutdown)
2010-01-19 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2254] data/driver.list, drivers, drivers/Makefile.am,
drivers/liebertgxt2.c, man/Makefile.am, man/liebertgxt2.8: Based
on the patch from the below message (needs to be tested)
http://lists.alioth.debian.org/pipermail/nut-upsdev/2010-January/004391.html
* [r2253] drivers/powercom-hid.c: Add
'beeper.(enable|disable|mute)' instcmd's
2010-01-17 Charles Lepple <clepple@gmail.com>
* [r2250] drivers/tripplite_usb.c: tripplite_usb: Detect protocol
before using tl_model
Reported by Chase Wallis.
2010-01-17 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2248] man/blazer.8: Add 'ippon' subdriver to man page
* [r2247] drivers/blazer_usb.c: ETIME is not defined on FreeBSD, so
make this test conditional (also fixes wrong sign)
2010-01-13 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2244] drivers/blazer_usb.c: Close device handle after
usb_reset() is called and reconnect
* [r2243] drivers/blazer_usb.c: Attempt to clear stalled devices by
sending them an usb_reset() command
2010-01-13 Charles Lepple <clepple@gmail.com>
* [r2242] drivers/tripplite_usb.c: tripplite_usb: don't reset
watchdog on SMARTPRO models
Patch by Chase Wallis:
http://article.gmane.org/gmane.comp.monitoring.nut.devel/4376
2010-01-12 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2241] drivers/blazer_usb.c: Fix bug in ippon subdriver
2010-01-07 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2240] drivers/blazer_usb.c: Fix minor bug in debug code
* [r2239] drivers/blazer_usb.c: Add experimental 'ippon' subdriver
to communicate with UPS devices of the same name
2010-01-06 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2238] drivers/blazer_usb.c: Clear potential stall condition
while flushing interrupt pipe
* [r2237] drivers/libusb.c, drivers/usbhid-ups.c: Reconnect if
usb_clear_halt() is not implemented (older libusb on *BSD
systems) so that we don't get stuck with a stalled interrupt
endpoint (correction of 2235)
* [r2236] drivers/usbhid-ups.c: EPIPE should not result in
reconnecting since we're reading from a control endpoint here,
not an interrupt endpoint (where this would be an appropriate
action after attempting to fix the stall condition through
usb_clear_halt() first)
* [r2235] drivers/usbhid-ups.c: Reconnect if usb_clear_halt() is
not implemented (older libusb on *BSD systems) so that we don't
get stuck with a stalled interrupt endpoint
2010-01-05 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2234] drivers/libusb.c: Don't ignore persistent "Broken pipe"
on stalled interrupt endpoint (only ignore them on control
endpoints)
* [r2233] drivers/libusb.c: Clear stall condition on interrupt
endpoint if needed (since we no longer reconnect, this may be
needed)
* [r2232] drivers/libusb.c, drivers/usbhid-ups.c: Change libusb.c
error handling. Most important (functional) change is that
"Broken pipe" is no longer considered a reason to reconnect
(fixes MGE/Eaton Ellipse series UPS stability problems).
2010-01-04 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2231] drivers/libusb.c: Fix previous patch
* [r2230] drivers/libusb.c: Log message to syslog when
usb_get/set_* operation fails (except timeouts) at LOG_DEBUG to
make sure we have something to work with for debugging frequent
disconnects
2010-01-02 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2224] m4/nut_check_libneon.m4: Fix output when library is not
installed
* [r2223] m4/nut_check_libhal.m4: Don't wast effort trying to
detect HAL stuff if we already know it is not available
* [r2222] m4/nut_check_libgd.m4: Remove braces around test
conditions
* [r2221] m4/nut_check_libpowerman.m4, m4/nut_check_libssl.m4:
Remove braces around test conditions
* [r2220] m4/nut_check_libusb.m4: Fix one "==" that was left behind
(and slightly improve message when pkg-config is not available)
2010-01-01 Charles Lepple <clepple@gmail.com>
* [r2219] m4/nut_check_libusb.m4: Remove extra parentheses from
libusb test.
Also change '==' to '=' for string equality test.
2009-12-31 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2218] m4/nut_check_libusb.m4: Add pkg-config as method to
retrieve CFLAGS and LDFLAGS (but fall back to libusb-config)
2009-12-30 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2217] drivers/libusb.c: Fix previous patch (failed to remove
closing #endif statement)
* [r2216] drivers/libusb.c, drivers/usbhid-ups.c: Enable the use of
the interrupt pipe on Solaris (for drivers that require this),
but disable it on the usbhid-ups driver (for which it is
optional).
* [r2215] drivers/powercom-hid.c: Honor 'offdelay' configuration
parameter and allow changing the 'ups.delay.shutdown' variable at
runtime
2009-12-28 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2214] data/driver.list: Add new PowerCOM devices with HID PDC
interface to list of supported models
* [r2213] drivers/powercom-hid.c: The PowerCOM subdriver has now
reached a state where it is useable. Commands have been tested on
a BNT-500AP.
* [r2212] drivers/usbhid-ups.c: The composite commands
'shutdown.return' and 'shutdown.stayoff' require 'load.on.delay'
and 'load.off.delay' (not 'ups.delay.start' and
'ups.delay.shutdown').
2009-12-27 Charles Lepple <clepple@gmail.com>
* [r2203] scripts/hal/Makefile.am: Hard-code source file in
scripts/hal
2009-12-26 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2202] m4/nut_check_libhal.m4: Revert previous patch that breaks
on OpenSUSE (and add HAL_CALLOUTS_PATH instead of relying on the
default)
* [r2201] m4/nut_check_libhal.m4: Use ${datarootdir} instead of
/usr/share or /usr/local/share (we check for the presence of a
subdirectory, so the chance of a false positive is low)
* [r2200] m4/nut_check_libneon.m4: Only check for optional
functions if we know library is available
* [r2199] m4/nut_check_libgd.m4: Simplify the --with-gd-includes
and --with-gd-libs processing
2009-12-25 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2198] configure.in: FreeBSD requires both <sys/types.h> and
<libutil.h> so explicitly use only these two headers to check for
availability of uu_lock
* [r2197] m4/nut_check_libusb.m4: Check for presence of 'usb.h'
(not 'libusb.h', which is what we use internally)
2009-12-24 Charles Lepple <clepple@gmail.com>
* [r2196] m4/nut_check_libhal.m4: Test for FDI files in /usr/local
as well as /usr
This should fix HAL-enabled builds on FreeBSD.
2009-12-24 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2195] m4/nut_check_libhal.m4: If we can't detect the
HAL_FDI_PATH automatically, set it to /dev/null and emit warning
* [r2194] m4/nut_check_libusb.m4: Fix missing default -lusb library
and improve comments during detection
* [r2193] m4/nut_check_libusb.m4: Work around missing libusb-config
on FreeBSD 8 by falling back to defaults for CFLAGS and LDFLAGS
* [r2192] configure.in: Try to workaround some FreeBSD bugs
* [r2191] drivers/powercom-hid.c: Updated HID to NUT mappings (work
in progress and still quite broken, do not run on production
systems!)
2009-12-22 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2190] drivers/libhid.c: Add debug information for input reports
that are not in the report descriptor
2009-12-21 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2189] drivers/libhid.c: Use hexadecimal report numbers in debug
messages (to be consistent throughout the code)
2009-12-20 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2188] m4/nut_check_libwrap.m4: Small improvement to better
mimic the behavior of AC_SEARCH_LIBS (by checking if we need to
add -lwrap before adding it to LIBS)
2009-12-19 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2187] m4/ax_create_stdint_h.m4: Update location of Autoconf
Archive in header (so that we can check for modifications)
* [r2186] m4/nut_check_libgd.m4: AC_SEARCH_LIBS puts the result in
LIB, so we need to add that to the LIBGD_LDFLAGS (otherwise, this
will be part of the global LIBS that are used)
2009-12-18 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2185] m4/nut_check_libgd.m4: Apparently, gdlib-config sometimes
fails to mention '-lgd' when asked for the LDFLAGS, so we need to
test once again if we need to include this library.
* [r2184] common/snprintf.c, configure.in: Instead of trying to
figure out if 'long long' is supported through AC_TRY_RUN
(obsolete), use AC_TYPE_LONG_LONG_INT (and AC_TYPE_LONG_DOUBLE).
Since this is only used by common/snprintf.c, only do this if
snprintf isn't a build-in function (which means probably never).
* [r2183] configure.in: Don't run preprocessor tests on
AC_CHECK_HEADERS (so CPPFLAGS is no longer needed). Replace shell
scripting by autoconf macros.
* [r2182] m4/nut_check_libgd.m4, m4/nut_check_libnetsnmp.m4,
m4/nut_check_libwrap.m4: Don't run preprocessor tests on
AC_CHECK_HEADERS (so CPPFLAGS is no longer needed).
* [r2181] m4/nut_check_libhal.m4: Don't run preprocessor tests on
AC_CHECK_HEADERS (so CPPFLAGS is no longer needed). Reformatted
macro (no functional changes).
* [r2180] m4/nut_check_libwrap.m4: Prefer AC_CHECK_HEADERS over
AC_CHECK_HEADER
* [r2179] m4/nut_check_libgd.m4: Set CPPFLAGS in case autoconf only
runs the preprocessor for AC_CHECK_HEADERS (which doesn't use
CFLAGS)
2009-12-17 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2178] m4/nut_check_libnetsnmp.m4: Set CPPFLAGS in case autoconf
only runs the preprocessor (which doesn't use CFLAGS)
* [r2177] m4/nut_check_libneon.m4: Cleanup autoconf macro
* [r2176] clients/upsimage.c: Remove redundant test (this is
already handled by autoconf)
* [r2175] Makefile.am:
* [r2174] configure.in: Keep defaults and checks together
* [r2173] drivers/bcmxcp.c: Automatic variables must have a length
that can be computed at compile time rather than runtime. GCC
will happily accept this C++ construction here, but this is not
C.
Follow the same principle as used elsewhere in the code, by hard
coding this to PW_ANSWER_MAX_SIZE, the maximum size of replies we
expect.
2009-12-16 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2172] m4/nut_check_libgd.m4: Check for all GD specific header
files while we're at it
* [r2171] clients/upsimage.c: The HAVE_GD_H should not be used
here, it is better to check for HAVE_LIBGD since that one is only
set if both the header file and the gd library are found (and not
just the first).
* [r2170] m4/nut_check_libnetsnmp.m4: Cleanup autoconf macro
* [r2169] m4/nut_check_libgd.m4: Cleanup GD autoconf stuff (also
fixes double -lgd linker flag)
* [r2168] configure.in, m4/nut_check_ipv6.m4,
m4/nut_check_libssl.m4, m4/nut_check_libwrap.m4: Cleanup some
autoconf macros
2009-12-15 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2167] m4/nut_check_libnetsnmp.m4: Use AC_CHECK_HEADER to check
for required headers.
* [r2166] m4/nut_check_libwrap.m4: Fix autoconf magic for Solaris.
Apparently, allow/deny_severity are not defined as weak symbols,
so autoconf fails with undefined symbols when checking for the
library through AC_SEARCH_LIBS.
2009-12-15 Charles Lepple <clepple@gmail.com>
* [r2165] include/common.h: Fix __attribute__ in common.h
Patch provided by Tim Rice <tim@multitalents.net>
http://lists.alioth.debian.org/pipermail/nut-upsdev/2009-December/004288.html
2009-12-10 Arnaud Quette <arnaud.quette@free.fr>
* [r2163] drivers/mge-utalk.c: complete r2151 with the
implementation details that were removed from the new mge-utalk
man page.
2009-12-09 Arnaud Quette <arnaud.quette@free.fr>
* [r2161] drivers/mge-hid.c: complete alarms and status handling
for Eaton / MGE / Dell.
2009-12-04 Arnaud Quette <arnaud.quette@free.fr>
* [r2153] m4/nut_check_libssl.m4: fix a wrong pkg-config call
parameter, that resulted in using pkg-config's own version
instead of OpenSSL's one.
2009-12-02 Arnaud Quette <arnaud.quette@free.fr>
* [r2148] drivers/libhid.c: skip reports 254/255 for Eaton / MGE /
Dell due to special handling needs
2009-11-30 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2145] man/upsd.8, man/upsd.conf.5, man/upsd.users.5: Add
upgrading notes for versions earlier than nut-2.4.0
* [r2143] drivers/usbhid-ups.c, drivers/usbhid-ups.h: Use contents
of interrupt report. In case these are broken, set 'pollonly'
flag in ups.conf or use_interrupt_pipe = FALSE in subdriver (when
device matches!)
2009-11-26 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2141] man/usbhid-ups.8: Document the 'pollonly' flag
2009-11-25 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2140] drivers/microdowell.c: Fix compiler warning (dead code,
no functional change)
* [r2139] drivers/usbhid-ups.c: Fix compiler warning
* [r2138] drivers/usbhid-ups.c: Add some USB error codes observed
in the field for possible future handling (and the comments from
the headerfiles where these are defined)
2009-11-24 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2133] drivers/mge-hid.c: We already know if the value to the
reverse conversion functions should be a time or date value, so
there is no need to guess it.
2009-11-23 Arnaud Quette <arnaud.quette@free.fr>
* [r2128] drivers/mge-hid.c: - fix mge_time_date_conversion_nuf()
to actually use the input value
- force ignoring the DST offset for the output value
2009-11-19 Arnaud Quette <arnaud.quette@free.fr>
* [r2122] data/driver.list, drivers/libshut.c, man/mge-shut.8: Add
serial / SHUT support for the new Dell UPS range
- list all models in driver.list using newmge-shut,
- add "battery.runtime.elapsed" support,
- update mge-shut manual page and add an alias for newmge-shut.
2009-11-19 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2121] drivers/mge-hid.c: Prefer to use non-negated logic when
possible (also minimizes changes)
2009-11-19 Arnaud Quette <arnaud.quette@free.fr>
* [r2119] data/driver.list, docs/new-names.txt, drivers/mge-hid.c,
man/usbhid-ups.8, scripts/dkp/95-devkit-power-hid.rules,
scripts/hal/ups-nut-device.fdi.in,
scripts/hotplug/libhid.usermap, scripts/udev/nut-usbups.rules.in:
Add USB support for the new Dell UPS range
- list all models in driver.list,
- add "battery.runtime.elapsed" in mge-hid.c and NUT namespace,
- adapt ups.model formating to Dell rules,
- update USB helper files,
- update usbhid-ups manual page.
2009-11-18 Arjen de Korte <adkorte-guest@alioth.debian.org>
* [r2118] drivers/netxml-ups.c: Re-subscribe to NMC when it doesn't
send us alarm messages anymore
2009-11-18 Arnaud Quette <arnaud.quette@free.fr>
* [r2117] drivers/mge-xml.c: minor typo fix on comment
Mon Nov 16 01:57:23 UTC 2009 / Charles Lepple <clepple@gmail.com>
- drivers/tripplite_usb.c: fix copy-n-paste error in load.on instcmd.
Fri Oct 9 11:26:49 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- m4/nut_check_libssl.m4,nut_check_libusb.m4: show version information during
configuration
Thu Oct 8 19:38:03 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- m4/nut_check_libneon.m4,nut_check_libusb.m4: changes in autoconf magic to
detect optional functions (without making assumptions on where they are
located)
- configure.in: update for the above
- drivers/libusb.c,netxml-ups.c,richcomm_usb.c: update for the above
Wed Oct 7 17:53:42 UTC 2009 / Arnaud Quette <arnaud.quette@free.fr>
* drivers/dummy-ups.c: dummy-ups general improvements
- dummy mode now loop on reading the .dev file, instead of reading it once.
Interacting through this method is also possible, by changing the .dev file
content on the fly,
- dummy mode now support actions (TIMER only ATM) in the .dev file. This
allows to have an automated scripting sequence (see data/evolution.dev and
dummy-ups manpage for examples),
- repeater mode now check the upsd connection, and reconnect if needed,
(reported by Gabor Kiss)
- cleanup the inline TODO list,
- bump the driver version to 0.10.
* man/dummy-ups.8: update according to the above changes.
* data/epdu-managed.dev: new dummy-ups definition file.
Wed Oct 7 14:05:35 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- drivers/snmp-ups.c: remove suspicious code for shutdown.* commands
Wed Oct 7 13:43:15 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- drivers/apc-hid.c: add support for Back-UPS RS series test.battery.*
commands (experimental)
Tue Oct 6 19:39:22 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- m4/nut_check_libssl.m4: use pkg-config for compiler and linker flags
- m4/nut_check_libusb.m4: minor change
Tue Oct 6 19:16:08 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- MAINTAINERS: Remove entry for Peter Selinger (by request)
Tue Oct 6 12:13:01 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- drivers/mge-shut.c: actually apply the low battery level if provided (either
through "-x lowbatt" or ups.conf). Also change the driver reported name from
"MGE UPS SYSTEMS" to "Eaton".
(reported by Daniel O'Connor)
Mon Oct 5 20:02:42 UTC 2009 / Arnaud Quette <arnaud.quette@free.fr>
- drivers/nut_usb.c: remove the calls to usb_set_configuration() and
usb_set_altinterface(), which were not needed and caused troubles
with some Eaton / Powerware models. Also add some older HP T500 / T750 which
seem to be supported,
- data/driver.list: add entries for the above.
Mon Sep 21 08:30:18 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- drivers/tripplite-hid.c: add ECO550UPS to list of supported devices.
- data/driver.list: add an entry for the above.
Sun Sep 20 19:06:34 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- drivers/tripplite-hid.c: fix broken matcher function that would
unconditionally reject possibly supported devices, even when the
productid option was supplied.
Tue Sep 15 09:06:32 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- data/driver.list: add an entry for Eaton E Series DX UPS (1-20 kVA) with
mge-utalk.
Tue Sep 15 08:16:32 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- data/driver.list: add an entry for Eaton E Series NV UPS (400-2000 VA) with
megatec_usb (reported by Vitor Choi Feitosa).
Tue Sep 8 18:57:51 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- m4/nut_check_libnetsnmp.m4: check for presence of header files to see
if development files are installed
Tue Sep 8 09:39:24 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- drivers/ivtscd.c: complete general information (device.{mfr,model,type}),
- man/ivtscd.8, man/Makefile.am: add the manpage for ivtscd driver,
- data/driver.list: add an entry for IVT SCD-series devices.
Fri Sep 4 08:36:21 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- scripts/perl/Nut.pm: embed the NUT Perl client module,
(patch from Gabor Kiss)
- COPYING, scripts/README: update information for Perl module inclusion.
Fri Sep 4 08:12:52 UTC 2009 / Arnaud Quette <arnaud.quette@free.fr>
- drivers/apc-hid.c: fix input.transfer.reason support by using the standard
BOOL lookup mechanism (thanks Arjen), and bump the subdriver version to 0.94.
Thu Sep 3 11:49:12 UTC 2009 / Arnaud Quette <arnaud.quette@free.fr>
- drivers/apc-hid.c: add input.transfer.reason and input.sensitivity, along
with some more reverse engineered information.
(reported by Markus Wildi)
Wed Sep 2 19:52:21 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- drivers/ietfmib.h,drivers/mgemib.h: Indexes start at 1 (not 0).
Tue Aug 25 20:40:28 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- drivers/usbhid-ups.[ch]: remove delayed matching which wasn't really
useful, since it requires claiming the device, breaking any existing
bindings in the process (should also fix Alioth bug #311869)
Tue Aug 18 01:22:14 UTC 2009 / Charles Lepple <clepple@gmail.com>
- drivers/bestfortress.c, man/bestfortress.8, data/driver.list: Resurrect the
bestfortress driver (submitted by Stuart D. Gathman)
Sat Aug 15 17:42:49 UTC 2009 / Charles Lepple <clepple@gmail.com>
- drivers/cps-hid.c, data/driver.list: Add CyberPower OR2200 to usbhid-ups
(reported by James Erickson)
Wed Aug 12 13:42:21 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- data/drivers.list: also list Inform Informer Compact 1000VA with blazer_ser
(reported by Mutlu Tunç)
Mon Aug 10 12:18:42 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- drivers/mge-xml.c: add support for newer Eaton ePDU monitored
- man/netxml-ups.8: add a reference to Eaton ePDU monitored
- data/drivers.list: update the Eaton ePDU monitored to include netxml-ups
Sun Aug 9 15:43:21 UTC 2009 / Arnaud Quette <arnaud.quette@free.fr>
- data/drivers.list: add Inform Informer Compact 1000VA
(reported by Mutlu Tunç)
Fri Jul 31 03:07:27 UTC 2009 / Charles Lepple <clepple@gmail.com>
- drivers/liebert.c: Hard-code the baud rate to 9600 to prevent problems with
noise.
Wed Jul 29 01:37:06 UTC 2009 / Charles Lepple <clepple@gmail.com>
- data/driver.list: Add "CP 1500C" to compatibility list for usbhid-ups
(reported by Svein Skogen)
Mon Jul 27 22:00:00 UTC 2009 / Arnaud Quette <arnaud.quette@free.fr>
- tools/nut-usbinfo.pl, scripts/dkp/95-devkit-power-hid.rules: remove
the 0x prefix from the VendorIDs
Mon Jul 27 11:21:21 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- tools/nut-usbinfo.pl, scripts/dkp/95-devkit-power-hid.rules: generate
the DeviceKit-power rule file, so that it can be maintained up to
date from the NUT source.
Mon Jul 20 19:20:35 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- Renamed 'virtual' to 'clone' driver, to resolve a conflict with the
Postfix virtual domain mail delivery agent man pages
Sat Jul 11 15:58:16 UTC 2009 / Charles Lepple <clepple@gmail.com>
* Makefile.am: Add "." to "find" command
Sat Jul 11 14:25:00 UTC 2009 / Charles Lepple <clepple@gmail.com>
* drivers/nut_usb.c: Add usb_strerror() to USB error messages.
Tue Jul 7 11:06:32 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
* device collection completion, for the transition period
- drivers/main.c: remap device.{mfr,model,serial} from the ups.*
equivalent, and preset device.type to ups so that it only has to be
overriden if necessary.
- drivers/skel.c: some update for the device collection
- docs/new-names.txt: add the device collection description
Mon Jul 6 09:28:10 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- drivers/tripplite.c: fix the reading with some SmartUPS models
(patch from Robert Waldie, Opengear)
Tue Jun 30 13:28:42 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- drivers/baytechmib.h: add outlet.{current,voltage} and fix some
comments (from Scott Burns, Opengear)
Wed May 27 20:35:41 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- clients/upsmon.c: multilevel debug information (similar to other NUT
binaries)
Mon May 25 19:53:02 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
* Enable timestamp on debug messages by default
- common/common.c: prepend elapsed number of seconds and microseconds
since start of program for upsdebug() messages,
- drivers/upsdrvctl.c: inform people that adding debug flags to upsdrvctl
is not the way to debug a driver,
- revert remainder of previous patch
Mon May 18 11:48:31 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
* Enable timestamp on output messages (format "%H:%M:%S: msg")
- common/common.c, include/common.h: modify vupslog() to allow optional
appending of a timestamp on debug output,
- clients/upsmon.c, drivers/main.c, drivers/upsdrvctl.c, server/upsd.c:
enable timestamping using the "-T" option,
- man/upsmon.8, man/nutupsdrv.8, man/upsdrvctl.8, man/upsd.8: document the
timestamp option.
Thu May 14 19:19:52 UTC 2009 / Arnaud Quette <aquette.dev@gmail.com>
- drivers/baytechmib.h: add native NUT support for Baytech SNMP PDUs (RPCs),
- drivers/snmp-ups.c: enable Baytech MIB,
- drivers/Makefile.am: add the new MIB file,
- data/driver.list: add the supported PDUs from BayTech,
- man/snmp-ups.8: document the new MIB and devices.
(patch from Scott Burns, Opengear)
Thu May 14 21:06:11 UTC 2009 / Arnaud Quette <aquette.dev@gmail.com>
- drivers/raritan-mib.h: use the outlets template mechanism and remove all
the OIDs defines for a better readability
- drivers/snmp-ups.c: enable Raritan MIB back.
Thu May 14 19:19:52 UTC 2009 / Arnaud Quette <aquette.dev@gmail.com>
- drivers/eaton-aphel-mib.h: use the outlets template mechanism.
Thu May 14 14:44:22 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- docs/configure.txt: a bit of completion and standardisation.
Thu May 14 14:23:41 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- drivers/snmp-ups.[ch]: add a template mechanism for outlets declaration, and
the needed helpers. Also disable temporarily the Raritan MIB as it conflicts
with Eaton (Aphel) .
Wed May 13 13:54:32 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- man/Makefile.am, man/nut.conf.5: add a manpage for nut.conf
(patch from Debian BTS, #528222)
Tue May 05 13:26:21 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- drivers/apcsmart.h: add support for older SmartUPS 600 responding to other
D codes (D7, D8 and D9)
(patch from Konstantin 'Kastus' Shchuka)
Tue May 05 02:01:13 UTC 2009 / Charles Lepple <clepple@gmail.com>
- drivers/tripplite_usb.c, man/tripplite_usb.8, data/driver.list: update list
of units tested with tripplite_usb
Mon May 04 08:11:01 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- drivers/apcsmart.[ch]: add support for older SmartUPS, which do not respond
to the ^Z, a or b commands (used to inquire the supported data)
(patch from Thomas Juerges)
Thu Apr 30 21:20:42 UTC 2009 / Arnaud Quette <aquette.dev@gmail.com>
- data/driver.list: add Apollo 1000A and 1000F compatibility with genericups
upstype=4 (reported by mvochin)
Thu Apr 23 12:46:12 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- man/snmp-ups.8: remove the EXPERIMENTAL section to the profit of a LIMITATION
section.
Fri Apr 17 13:26:21 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- data/driver.list: add UPSonic DS-800 compatibility with megatec_usb
(reported by Nick Jenkins)
Sun Mar 29 20:59:09 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
- drivers/netxml.c: Either do a quick poll or a complete poll (not both).
Sun Mar 29 20:21:19 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
* clients/upsimage.c, clients/upsstats.c:
- Add ups.temperature, ambient.temperature and ambient.humidity (sorry folks,
Celsius only for now).
- The graph for the battery.charge now honors battery.charge.low for the red zone.
- The number of minor dashes for the battery.charge is reduced.
- Value on the scales is now right aligned. Display of negative values is corrected.
* conf/upsstats-single.html.sample:
- If ambient.(temperature|humidity) is supported, an 'Ambient' column is added to
display the bars for them.
Thu Mar 26 20:04:53 UTC 2009 / Arjen de Korte <arjen@de-korte.org>
* drivers/netxml.[ch], drivers/mge-xml.c:
- Allow multiple entry points for the initial XML page that is loaded
- Use different XML pages for quick and full status updates (the latter
is only polled once every 10 times)
Mon Mar 23 17:34:10 UTC 2009 / Kjell Claesson <kjell.claesson@epost.tidanet.se>
- data/driver.list, add Eaton Powerware 9130 compatibility with bcmxcp and usbhid-ups.
Sun Mar 22 01:13:18 UTC 2009 / Alexander I. Gordeev <lasaine@lvk.cs.msu.su>
- drivers/blazer_usb.c: read data from device until timeout.
Mon Mar 16 22:10:30 UTC 2009 / Kjell Claesson <kjell.claesson@epost.tidanet.se>
- drivers/bcmxcp.[ch]: Extend alarm map to fix debug crash on PW9130. Driver version 0.22.
Wed Mar 04 19:37:12 UTC 2009 / Arnaud Quette <aquette.dev@gmail.com>
- drivers/mge-utalk.c: fix enable_ups_comm()
(patch from Hans-Werner Paulsen)
Sun Mar 01 19:51:54 UTC 2009 / Arnaud Quette <aquette.dev@gmail.com>
- data/driver.list, drivers/tripplite-hid.c: add HP T750 INTL
(reported by KP Kirchdoerfer)
Fri Feb 27 09:33:27 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- data/driver.list: add Mustek PowerMust 424 / 636 / 848 (USB) compatibility
with blazer_usb (reported by Vitaly Polyakov)
Fri Feb 27 09:07:05 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- drivers/apcsmart.[ch]: add APC Matrix 5000, vintage 12/00
(patch from Jarett Stevens)
Tue Feb 17 22:00:01 UTC 2009 / Arnaud Quette <aquette@debian.org>
- drivers/Makefile.am: Fix snmp-ups overlinking with lcrypto.
(Debian Lintian report)
2.4.1
Mon Feb 16 17:41:32 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- configure.in: 2.4.1
Mon Feb 16 16:13:47 UTC 2009 / David Goncalves <david@lestat.st>
- scripts/python/app/NUT-Monitor: Changed version from 1.0 to 1.1. Changed
the way NUT-Monitor deals with 'ups.status' var to better handle composite
status (OL + TRIM, OL + CHRG, ...)
Thu Feb 12 13:23:41 UTC 2009 / Arnaud Quette <ArnaudQuette@Eaton.com>
- scripts/udev/Makefile.am, scripts/hal/Makefile.am, scripts/hotplug/Makefile.am,
Makefile.am: fix the autotools issue with the clean/distclean target removing
the USB helper files.
Wed Feb 11 10:43:10 UTC 2009 / Kjell Claesson <kjell.claesson@epost.tidanet.se>
- data/driver.list: Updated list with ups supported by microdowell driver.
Thu Feb 05 22:16:10 UTC 2009 / Kjell Claesson <kjell.claesson@epost.tidanet.se>
- drivers/microdowell.c: Cleaned some of the code. Fixed reading of ups.conf variables.
Wed Feb 04 15:01:20 UTC 2009 / Kjell Claesson <kjell.claesson@epost.tidanet.se>
- Added microdowell.c/h into drivers/. add man/microdowell.8. Adjusted
drivers/Makefile.am and man/makefile.am. Adjusted the code for 2.4.0
in the microdowell.c.
*** File trimmed here 28 January 2009 ***
For entries before this point, start with version 2.4.1 and work back.
Reference in New Issue View Git Blame Copy Permalink