1 <!DOCTYPE book PUBLIC
"-//OASIS//DTD DocBook V4.1//EN" >
3 <title>Monitoring Plugins Development Guidelines
</title>
9 <orgname>Monitoring Plugins Development Team
</orgname>
14 <pubdate>2013</pubdate>
15 <title>Monitoring Plugins Development Guidelines
</title>
18 <year>2000 -
2024</year>
19 <holder>Monitoring Plugins Development Team
</holder>
25 <preface id=
"preface"><title>Preface
</title>
26 <para>The purpose of this guidelines is to provide a reference for
27 the plugin developers and encourage the standardization of the
28 different kind of plugins: C, shell, perl, python, etc.
</para>
30 <para>Monitoring Plugins Development Guidelines Copyright (C)
2000-
2024
31 (Monitoring Plugins Team)
</para>
33 <para>Permission is granted to make and distribute verbatim
34 copies of this manual provided the copyright notice and this
35 permission notice are preserved on all copies.
</para>
37 <para>The plugins themselves are copyrighted by their respective
42 <section id=
"DevRequirements"><title>Development platform requirements
</title>
44 Monitoring Plugins are developed to the GNU standard, so any OS which is supported by GNU
45 should run the plugins. While the requirements for compiling the Monitoring Plugins release
46 are very basic, developing from the Git repository requires additional software to be
47 installed. These are the minimum levels of software required:
57 To compile from Git, after you have cloned the repository, run:
67 <section id=
"PlugOutput"><title>Plugin Output for Nagios
</title>
69 <para>You should always print something to STDOUT that tells if the
70 service is working or why it is failing. Try to keep the output short -
71 probably less that
80 characters. Remember that you ideally would like
72 the entire output to appear in a pager message, which will get chopped
73 off after a certain length.
</para>
75 <para>As Nagios does not capture stderr output, you should only output to
76 STDOUT and not print to STDERR.
</para>
78 <section><title>Print only one line of text
</title>
79 <para>Starting with version
3, Nagios will process plugins' multiline
80 output, which should be formatted as:
</para>
82 SERVICE STATUS: First line of output | First part of performance data
83 Any number of subsequent lines of output, but note that buffers
84 may have a limited size | Second part of performance data, which
85 may have continuation lines, too
87 <para>Note, however, that the default configs still do not include
88 the output's continuation lines into the notifications sent when
89 Nagios notifies contacts about potential problems.
90 Thus, keep your output short and to the point.
</para>
92 <para>Output should be in the format:
</para>
94 SERVICE STATUS: Information text
96 <para>However, note that this is not a requirement of the API, so you cannot depend on this
97 being an accurate reflection of the status of the service - the status should always
98 be determined by the return code.
</para>
101 <section><title>Verbose output
</title>
102 <para>Use the -v flag for verbose output. You should allow multiple
103 -v options for additional verbosity, up to a maximum of
3. The standard
104 type of output should be:
</para>
106 <table id=
"verboselevels"><title>Verbose output levels
</title>
110 <entry><para>Verbosity level
</para></entry>
111 <entry><para>Type of output
</para></entry>
116 <entry align=
"center"><para>0</para></entry>
117 <entry><para>Single line, minimal output. Summary
</para></entry>
120 <entry align=
"center"><para>1</para></entry>
121 <entry><para>Single line, additional information (eg list processes that fail)
</para></entry>
124 <entry align=
"center"><para>2</para></entry>
125 <entry><para>Multi line, configuration debug output (eg ps command used)
</para></entry>
128 <entry align=
"center"><para>3</para></entry>
129 <entry><para>Lots of detail for plugin problem diagnosis
</para></entry>
136 <section><title>Screen Output
</title>
137 <para>The plugin should print the diagnostic and just the
138 usage part of the help message. A well written plugin would
139 then have --help as a way to get the verbose help.
</para>
141 <para>Code and output should try to respect the
80x25 size of a
142 crt (remember when fixing stuff in the server room!)
</para>
145 <section><title>Plugin Return Codes
</title>
146 <para>The return codes below are based on the POSIX spec of returning
147 a positive value. Netsaint prior to v0.0
.7 supported non-POSIX
148 compliant return code of
"-1" for unknown. Nagios supports POSIX return
149 codes by default.
</para>
151 <para>Note: Some plugins will on occasion print on STDOUT that an error
152 occurred and error code is
138 or
255 or some such number. These
153 are usually caused by plugins using system commands and having not
154 enough checks to catch unexpected output. Developers should include a
155 default catch-all for system command output that returns an UNKNOWN
158 <table id=
"ReturnCodes"><title>Plugin Return Codes
</title>
162 <entry><para>Numeric Value
</para></entry>
163 <entry><para>Service Status
</para></entry>
164 <entry><para>Status Description
</para></entry>
169 <entry align=
"center"><para>0</para></entry>
170 <entry valign=
"middle"><para>OK
</para></entry>
171 <entry><para>The plugin was able to check the service and it
172 appeared to be functioning properly
</para></entry>
175 <entry align=
"center"><para>1</para></entry>
176 <entry valign=
"middle"><para>Warning
</para></entry>
177 <entry><para>The plugin was able to check the service, but it
178 appeared to be above some
"warning" threshold or did not appear
179 to be working properly
</para></entry>
182 <entry align=
"center"><para>2</para></entry>
183 <entry valign=
"middle"><para>Critical
</para></entry>
184 <entry><para>The plugin detected that either the service was not
185 running or it was above some
"critical" threshold
</para></entry>
188 <entry align=
"center"><para>3</para></entry>
189 <entry valign=
"middle"><para>Unknown
</para></entry>
190 <entry><para>Invalid command line arguments were supplied to the
191 plugin or low-level failures internal to the plugin (such as unable to fork,
192 or open a tcp socket) that prevent it from performing the specified
193 operation. Higher-level errors (such as name resolution errors,
194 socket timeouts, etc) are outside of the control of plugins and should
195 generally NOT be reported as UNKNOWN states.
196 </para><para>The --help or --version output should also result in Unknown state.
206 <section id=
"thresholdformat"><title>Threshold and ranges
</title>
207 <para>A range is defined as a start and end point (inclusive) on a numeric scale (possibly
208 negative or positive infinity).
210 <para>A threshold is a range with an alert level (either warning or critical). Use the
211 set_thresholds(thresholds *, char *, char *) function to set the thresholds.
213 <para>The theory is that the plugin will do some sort of check which returns
214 back a numerical value, or metric, which is then compared to the warning and
215 critical thresholds. Use the get_status(double, thresholds *) function to
216 compare the value against the thresholds.
</para>
217 <para>This is the generalised format for ranges:
</para>
225 <listitem><para>start
≤ end
</para>
227 <listitem><para>start and
":" is not required if start=
0</para>
229 <listitem><para>if range is of format
"start:" and end is not specified,
230 assume end is infinity
</para>
232 <listitem><para>to specify negative infinity, use
"~"</para>
234 <listitem><para>alert is raised if metric is outside start and end range
235 (inclusive of endpoints)
</para>
237 <listitem><para>if range starts with
"@", then alert if inside this range
238 (inclusive of endpoints)
</para>
242 <para>Note: Not all plugins are coded to expect ranges in this format yet.
243 There will be some work in providing multiple metrics.
</para>
245 <table id=
"ExampleRanges"><title>Example ranges
</title>
249 <entry><para>Range definition
</para></entry>
250 <entry><para>Generate an alert if x...
</para></entry>
256 <entry>< 0 or
> 10, (outside the range of {
0 ..
10})
</entry>
260 <entry>< 10, (outside {
10 ..
∞})
</entry>
264 <entry>> 10, (outside the range of {-
∞ ..
10})
</entry>
268 <entry>< 10 or
> 20, (outside the range of {
10 ..
20})
</entry>
271 <entry>@
10:
20</entry>
272 <entry>≥ 10 and
≤ 20, (inside the range of {
10 ..
20})
</entry>
277 <table id=
"CommandLineExamples"><title>Command line examples
</title>
281 <entry><para>Command line
</para></entry>
282 <entry><para>Meaning
</para></entry>
287 <entry>check_stuff -w10 -c20
</entry>
288 <entry>Critical if
"stuff" is over
20, else warn if over
10 (will be critical if
"stuff" is less than
0)
</entry>
291 <entry>check_stuff -w~:
10 -c~:
20</entry>
292 <entry>Same as above. Negative
"stuff" is OK
</entry>
295 <entry>check_stuff -w10: -c20
</entry>
296 <entry>Critical if
"stuff" is over
20, else warn if
"stuff" is below
10 (will be critical if
"stuff" is less than
0)
</entry>
299 <entry>check_stuff -c1:
</entry>
300 <entry>Critical if
"stuff" is less than
1</entry>
303 <entry>check_stuff -w~:
0 -c10
</entry>
304 <entry>Critical if
"stuff" is above
10; Warn if
"stuff" is above zero (will be critical if
"stuff" is less than
0)
</entry>
307 <entry>check_stuff -c5:
6</entry>
308 <entry>Critical if
"stuff" is less than
5 or more than
6</entry>
311 <entry>check_stuff -c@
10:
20</entry>
312 <entry>OK if stuff is less than
10 or higher than
20, otherwise critical
</entry>
319 <section><title>Performance data
</title>
320 <para>Nagios
3 and newer will concatenate the parts following a
"|" in a) the first
321 line output by the plugin, and b) in the second to last line, into a string it
322 passes to whatever performance data processing it has configured. (Note that it
323 currently does not insert additional whitespace between both, so the plugin needs
324 to provide some to prevent the last pair of a) and the first of b) getting run
325 together.) Please refer to the Nagios documentation for information on how to
326 configure such processing. However, it is the responsibility of the plugin writer
327 to ensure the performance data is in a
"Nagios Plugins" format.
328 This is the expected format:
</para>
331 'label'=value[UOM];[warn];[crit];[min];[max]
336 <listitem><para>space separated list of label/value pairs
</para>
338 <listitem><para>label can contain any characters except the equals sign or single quote (')
</para>
340 <listitem><para>the single quotes for the label are optional. Required if
341 spaces are in the label
</para>
343 <listitem><para>label length is arbitrary, but ideally the first
19 characters
344 are unique (due to a limitation in RRD). Be aware of a limitation in the
345 amount of data that NRPE returns to Nagios
</para>
347 <listitem><para>to specify a quote character, use two single quotes
</para>
349 <listitem><para>warn, crit, min or max may be null (for example, if the threshold is
350 not defined or min and max do not apply). Trailing unfilled semicolons can be
353 <listitem><para>min and max are not required if UOM=%
</para>
355 <listitem><para>value, min and max in class [-
0-
9.]. Must all be the
356 same UOM. value may be a literal
"U" instead, this would indicate that the
357 actual value couldn't be determined
</para>
359 <listitem><para>warn and crit are in the range format (see
360 <xref linkend=
"thresholdformat">). Must be the same UOM
</para>
362 <listitem><para>UOM (unit of measurement) is one of:
</para>
364 <listitem><para>no unit specified - assume a number (int or float)
365 of things (eg, users, processes, load averages)
</para>
367 <listitem><para>s - seconds (also us, ms)
</para></listitem>
368 <listitem><para>% - percentage
</para></listitem>
369 <listitem><para>B - bytes (also KB, MB, TB)
</para></listitem>
370 <listitem><para>c - a continuous counter (such as bytes
371 transmitted on an interface)
</para></listitem>
376 <para>It is up to third party programs to convert the Monitoring Plugins
377 performance data into graphs.
</para>
381 <section id=
"SysCmdAuxFiles"><title>System Commands and Auxiliary Files
</title>
383 <section><title>Don't execute system commands without specifying their
385 <para>Don't use exec(), popen(), etc. to execute external
386 commands without explicitly using the full path of the external
389 <para>Doing otherwise makes the plugin vulnerable to hijacking
390 by a trojan horse earlier in the search path. See the main
391 plugin distribution for examples on how this is done.
</para>
394 <section><title>Use spopen() if external commands must be executed
</title>
396 <para>If you have to execute external commands from within your
397 plugin and you're writing it in C, use the spopen() function
398 that Karl DeBisschop has written.
</para>
400 <para>The code for spopen() and spclose() is included with the
401 core plugin distribution.
</para>
404 <section><title>Don't make temp files unless absolutely required
</title>
406 <para>If temp files are needed, make sure that the plugin will
407 fail cleanly if the file can't be written (e.g., too few file
408 handles, out of disk space, incorrect permissions, etc.) and
409 delete the temp file when processing is complete.
</para>
412 <section><title>Don't be tricked into following symlinks
</title>
414 <para>If your plugin opens any files, take steps to ensure that
415 you are not following a symlink to another location on the
419 <section><title>Validate all input
</title>
421 <para>use routines in utils.c or utils.pm and write more as needed
</para>
429 <section id=
"PerlPlugin"><title>Perl Plugins
</title>
431 <para>Perl plugins are coded a little more defensively than other
432 plugins because of embedded Perl. When configured as such, embedded
433 Perl Nagios (ePN) requires stricter use of the some of Perl's features.
434 This section outlines some of the steps needed to use ePN
439 <listitem><para> Do not use BEGIN and END blocks since they will be called
440 only once (when Nagios starts and shuts down) with Embedded Perl (ePN). In
441 particular, do not use BEGIN blocks to initialize variables.
</para>
444 <listitem><para>To use utils.pm, you need to provide a full path to the
445 module in order for it to work.
</para>
449 use lib
"/usr/local/nagios/libexec";
454 <listitem><para>Perl scripts should be called with
"-w"</para>
457 <listitem><para>All Perl plugins must compile cleanly under
"use strict" - i.e. at
458 least explicitly package names as in
"$main::x" or predeclare every
462 <para>Explicitly initialize each variable in use. Otherwise with
463 caching enabled, the plugin will not be recompiled each time, and
464 therefore Perl will not reinitialize all the variables. All old
465 variable values will still be in effect.
</para>
468 <listitem><para>Do not use
>DATA
< handles (these simply do not compile under ePN).
</para>
471 <listitem><para>Do not use global variables in named subroutines. This is bad practise anyway, but with ePN the
472 compiler will report an error
"<global_var> will not stay shared ..". Values used by
473 subroutines should be passed in the argument list.
</para>
476 <listitem><para>If writing to a file (perhaps recording
477 performance data) explicitly close close it. The plugin never
478 calls
<emphasis role=
"strong">exit
</emphasis>; that is caught by
479 p1.pl, so output streams are never closed.
</para>
482 <listitem><para>As in
<xref linkend=
"runtime"> all plugins need
483 to monitor their runtime, specially if they are using network
484 resources. Use of the
<emphasis>alarm
</emphasis> is recommended
485 noting that some Perl modules (eg LWP) manage timers, so that an alarm
486 set by a plugin using such a module is overwritten by the module.
487 (workarounds are cunning (TM) or using the module timer)
488 Plugins may import a default time out ($TIMEOUT) from utils.pm.
492 <listitem><para>Perl plugins should import %ERRORS from utils.pm
493 and then
"exit $ERRORS{'OK'}" rather than
"exit 0"
501 <section id=
"runtime"><title>Runtime Timeouts
</title>
503 <para>Plugins have a very limited runtime - typically
10 sec.
504 As a result, it is very important for plugins to maintain internal
505 code to exit if runtime exceeds a threshold.
</para>
507 <para>All plugins should timeout gracefully, not just networking
508 plugins. For instance, df may lock if you have automounted
509 drives and your network fails - but on first glance, who'd think
510 df could lock up like that. Plus, it should just be more error
511 resistant to be able to time out rather than consume
514 <section><title>Use DEFAULT_SOCKET_TIMEOUT
</title>
516 <para>All network plugins should use DEFAULT_SOCKET_TIMEOUT to timeout
</para>
521 <section><title>Add alarms to network plugins
</title>
523 <para>If you write a plugin which communicates with another
524 networked host, you should make sure to set an alarm() in your
525 code that prevents the plugin from hanging due to abnormal
526 socket closures, etc. Nagios takes steps to protect itself
527 against unruly plugins that timeout, but any plugins you create
528 should be well behaved on their own.
</para>
536 <section id=
"PlugOptions"><title>Plugin Options
</title>
538 <para>A well written plugin should have --help as a way to get
539 verbose help. Code and output should try to respect the
80x25 size of a
540 crt (remember when fixing stuff in the server room!)
</para>
542 <section><title>Option Processing
</title>
544 <para>For plugins written in C, we recommend the C standard
545 getopt library for short options. Getopt_long is always available.
548 <para>For plugins written in Perl, we recommend Getopt::Long module.
</para>
550 <para>Positional arguments are strongly discouraged.
</para>
552 <para>There are a few reserved options that should not be used
553 for other purposes:
</para>
556 -V version (--version)
558 -t timeout (--timeout)
559 -w warning threshold (--warning)
560 -c critical threshold (--critical)
561 -H hostname (--hostname)
562 -v verbose (--verbose)
565 <para>In addition to the reserved options above, some other standard options are:
</para>
568 -C SNMP community (--community)
569 -a authentication password (--authentication)
570 -l login name (--logname)
571 -p port or password (--port or --passwd/--password)monitors operational
572 -u url or username (--url or --username)
575 <para>Look at check_pgsql and check_procs to see how I currently
576 think this can work. Standard options are:
</para>
579 <para>The option -V or --version should be present in all
580 plugins. For C plugins it should result in a call to print_revision, a
581 function in utils.c which takes two character arguments, the
582 command name and the plugin revision.
</para>
584 <para>The -? option, or any other unparsable set of options,
585 should print out a short usage statement. Character width should
586 be
80 and less and no more that
23 lines should be printed (it
587 should display cleanly on a dumb terminal in a server
590 <para>The option -h or --help should be present in all plugins.
591 In C plugins, it should result in a call to print_help (or
592 equivalent). The function print_help should call print_revision,
593 then print_usage, then should provide detailed
594 help. Help text should fit on an
80-character width display, but
595 may run as many lines as needed.
</para>
597 <para>The option -v or --verbose should be present in all plugins.
598 The user should be allowed to specify -v multiple times to increase
599 the verbosity level, as described in
<xref linkend=
"verboselevels">.
</para>
601 <para>The exit code for version information or help should be UNKNOWN
606 <title>Plugins with more than one type of threshold, or with
607 threshold ranges
</title>
609 <para>Old style was to do things like -ct for critical time and
610 -cv for critical value. That goes out the window with POSIX
611 getopt. The allowable alternatives are:
</para>
615 <para>long options like -critical-time (or -ct and -cv, I
620 <para>repeated options like `check_load -w
10 -w
6 -w
4 -c
621 16 -c
10 -c
10`
</para>
625 <para>for brevity, the above can be expressed as `check_load
626 -w
10,
6,
4 -c
16,
10,
10`
</para>
630 <para>ranges are expressed with colons as in `check_procs -C
631 httpd -w
1:
20 -c
1:
30` which will warn above
20 instances,
632 and critical at
0 and above
30</para>
636 <para>lists are expressed with commas, so Jacob's check_nmap
637 uses constructs like '-p
1000,
1010,
1050:
1060,
2000'
</para>
641 <para>If possible when writing lists, use tokens to make the
642 list easy to remember and non-order dependent - so
643 check_disk uses '-c
10000,
10%' so that it is clear which is
644 the percentage and which is the KB values (note that due to
645 my own lack of foresight, that used to be '-c
10000:
10%' but
646 such constructs should all be changed for consistency,
647 though providing reverse compatibility is fairly
653 <para>As always, comments are welcome - making this consistent
654 without a host of long options was quite a hassle, and I would
655 suspect that there are flaws in this strategy.
660 <section id=
"Testcases"><title>Test cases
</title>
662 Tests are the best way of knowing if the plugins work as expected. Please
663 create and update test cases where possible.
667 To run a test, from the top level directory, run
"make test". This will run
668 all the current tests and report an overall success rate.
671 <section><title>Test cases for plugins
</title>
672 <para>These use perl's Test::More. To do a one time test, run
"cd plugins && perl t/check_disk.t".
675 <para>There will sometimes be failures seen in this output which are known failures that
676 need to be fixed. As long as the return code is
0, it will be reported as
"test pass".
677 (If you have a fix so that the specific test passes, that will be gratefully received!)
681 If you want a summary test, run:
"cd plugins && prove t/check_disk.t".
682 This runs the test in a summary format.
686 For a good and amusing tutorial on using Test::More, see this
687 <ulink url=
"http://search.cpan.org/dist/Test-Simple/lib/Test/Tutorial.pod">
693 <section><title>Testing the C library functions
</title>
695 We use
<ulink url=
"http://jc.ngo.org.uk/trac-bin/trac.cgi/wiki/LibTap">the libtap library
</ulink>, which gives
697 (Test Anything Protocol) output. This is used by the FreeBSD team for their regression testing.
701 To run tests using the libtap library, download the latest tar ball and extract.
702 There is a problem with tap-
1.01 where
703 <ulink url=
"http://jc.ngo.org.uk/trac-bin/trac.cgi/ticket/25">pthread support doesn't appear to work
</ulink>
704 properly on non-FreeBSD systems. Install with '
CPPFLAGS=
"-UHAVE_LIBPTHREAD" ./configure && make && make check && make install'.
708 When you run Monitoring Plugins' configure, it will look for the tap library and will automatically
709 setup the tests. Run
"make test" to run all the tests.
714 <section id=
"CodingGuidelines"><title>Coding guidelines
</title>
715 <para>See
<ulink url=
"http://www.gnu.org/prep/standards_toc.html">GNU
716 Coding standards
</ulink> for general guidelines.
717 Some parts of the guidelines (namely and explicitly the coding style)
718 are not seen as best practice by us.
</para>
719 <section><title>C coding
</title>
721 <para>You should use the type
"bool" and its values
722 "true" and
"false" instead of the
"int" type for booleans.
726 <section><title>Crediting sources
</title>
727 <para>If you have copied a routine from another source, make sure the license
728 from your source allows this. Add a comment referencing the ACKNOWLEDGEMENTS
729 file, where you can put more detail about the source.
</para>
730 <para>For contributed code, do not add any named credits in the source code
731 - contributors should be added into the THANKS.in file instead.
735 <section><title>Commit Messages
</title>
736 <para>If the change is due to a contribution, please quote the contributor's name
737 and, if applicable, add the GitHub Issue Tracker number. Don't forget to
738 update the THANKS.in file.
</para>
739 <para>If you have a change that is useful for noting in the next release, please
740 update the NEWS file.
</para>
741 <para>All commits will be written to a ChangeLog at release time.
746 <section id=
"SubmittingChanges"><title>Submission of new plugins and patches
</title>
748 <section id=
"Patches"><title>Patches
</title>
749 <para>If you have a bug patch, please supply a unified or context diff against the
750 version you are using. For new features, please supply a diff against
751 the Git
"master" branch.
</para>
753 <para>Patches should be submitted via
754 <ulink url=
"https://github.com/monitoring-plugins/monitoring-plugins/pulls">GitHub's
755 Pull requests for Monitoring Plugins
</ulink>
758 <para>Submission of a patch implies that the submmitter acknowledges that they
759 are the author of the code (or have permission from the author to release the code)
760 and agree that the code can be released under the GPL. The copyright for the changes will
761 then revert to the Monitoring Plugins Development Team - this is required so that any copyright
762 infringements can be investigated quickly without contacting a huge list of copyright holders.
763 Credit will always be given for any patches through a THANKS file in the distribution.
</para>
767 <section id=
"Contributedplugins"><title>Contributed plugins
</title>
768 <para>Plugins that have been contributed to the project and
769 distributed with the Monitoring Plugins files are held in the contrib/ directory and are not installed
770 by default. These plugins are not officially supported by the team.
771 The current policy is that these plugins should be owned and maintained by the original
772 contributor, preferably hosted on
<ulink url=
"https://exchange.icinga.com/">Icinga Exchange
</ulink>.
774 <para>If patches or bugs are raised to an contributed plugin, we will start communications with the
775 original contributor, but seek to remove the plugin from our distribution.
777 <para>The aim is to distribute only code that the Monitoring Plugins team are responsible for.
781 <section id=
"Newplugins"><title>New plugins
</title>
782 <para>If you would like others to use your plugins, please add it to
783 the official
3rd party plugin repository,
784 <ulink url=
"https://exchange.icinga.com/">Icinga Exchange
</ulink>.
787 <para>We are not accepting requests for inclusion of plugins into
788 our distribution at the moment, but when we do, these are the minimum
794 <para>Include copyright and license information in all files. Copyright must be solely
795 granted to the Monitoring Plugins Development Team
</para>
798 <para>The standard command options are supported (--help, --version,
799 --timeout, --warning, --critical)
</para>
802 <para>It is determined to be not redundant (for instance, we would not
803 add a new version of check_disk just because someone had provide
804 a plugin that had perf checking - we would incorporate the features
805 into an existing plugin)
</para>
808 <para>One of the developers has had the time to audit the code and declare
809 it ready for core
</para>
812 <para>It should also follow code format guidelines, and use functions from
813 utils (perl or c or sh) rather than using its own
</para>
816 <para>Includes patches to configure.in if required (via the EXTRAS list if
817 it will only work on some platforms)
</para>
820 <para>If possible, please submit a test harness. Documentation on sample
821 tests coming soon
</para>