This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Update Time-HiRes to CPAN version 1.9742
[perl5.git] / dist / Time-HiRes / Changes
CommitLineData
df003a36
JH
1Revision history for the Perl extension Time::HiRes.
2
5cd155b0
CBW
31.9742 [2017-04-16]
4 - prefer 3-argument open: blead 1ae6ead9
5 - fix dist/Time-HiRes/t/*.t that assumed '.' in @INC: blead 465db51d
6 - fix cases where 'do file' should be 'do ./file'.: blead 8b69401c
7
12389a22
CBW
81.9741 [2016-11-20]
9 - C++11 compatibility: blead a914236c
10 - El Capitan compatibility: blead 45bbc013
11 - use CLOCK_REALTIME for clock_nanosleep scan: blead 56ed632c
12 - include file consistency in scans: blead 605b4dbe
13 - use clockid_t consistently: blead 2979fdf3
14 - use hv_fetchs(): blead 07e4dd7a
15 - scan for clockid_t (needed for macos Sierra): blead 69d1f2c2
16 - darwin lacks clockid_t: blead d1e7b5aa
17 [rt.cpan.org #129789]
18 - provide template for tempfile: blead 91dca83e
19 - explicit cast for clockid_t in C++11: blead 1c566547
20 - upgrade to a newer ppport (3.35)
1c566547 21
3a57bd4d
JK
221.9740 [2016-09-25]
23 - the ext3/ext2 filesystems do not have subsecond resolution,
24 therefore skip the t/utime.t test
25 [rt.cpan.org #116127]
26
c4a535af
SH
271.9739 [2016-06-28]
28 - the upcoming macOS 10.12 (Sierra, the operating system formerly
29 known as OS X, or Darwin) has implemented the clock_gettime()
30 and clock_getres() interfaces but not the clock_nanosleep()
31
321.9738 [2016-06-26]
33 - fix the dates in the Changes, it's not July yet
34 [rt.cpan.org #115553]
35 - in cygwin, drop the utime.t expected timestamp resolution
36 - documented for utime that the actual timestamp resolution depends
37 on the operating system and the filesystem, just like for stat
38 - looks like in gnukfreebsd 10.1 neither futimens or utimensat work
39 even though they exist, skipping utime tests there for now
40 - demodernize the Test::More use in utime.t (no subtest) to
41 better support older Perl installations
42 - for all the tests, drop the Test::More 0.82 requirement,
43 the only "modern" thing used is the note routine which
44 can be replaced with print("# ...\n") (or printf),
45 with this works back at least with Test::More 0.54 (perl 5.8.7)
46
471.9737 [2016-06-22]
48 - utime in HiRes.xs had dXSARGS which was unnecessary, leading
49 to two markstack pops, panic, and two-minute wait for the watchdog
50 (thanks to Dave Mitchell for the eagle-eye)
51 - even though 5.10.x should have the Perl_ck_warner, many CPAN
52 smokes proved otherwise, so requiring now at least 5.12 for
53 Perl_ck_warner use in IS_SAFE_PATHNAME()
54
551.9736 [2016-06-22]
56 - the IS_SAFE_PATHNAME() emulation was too modern for Perl 5.10
57 - make it so that only one value is set for -DTIME_HIRES_STAT
58 even on systems that support many options
59 - the 1.9734 (only released in blead, and as part of the development
60 release 5.25.2) fix for Darwin's clock_gettime was wrong under
61 threaded builds, leading into mutex panics in clock.t
62 - the d_hires_utime determination was made more precise
63 - the utime.t was failing in platforms where only one of
64 futimens/utimensat was present (neither or both was okay)
65
661.9735 [2016-06-20]
67 - Time::HiRes should override `utime` to allow setting hires
68 (futimens and utimensat support)
69 [rt.perl.org #114809]
70 - the utime patch used IS_SAFE_PATHNAME() which isn't available in
71 too old Perls, so emulate (in case the Devel::PPPort is too old)
72
731.9734 [2016-06-17]
74 - fix Darwins with clock_gettime: blead 2d41a263
75 [rt.perl.org #128427]
76
f3dd6bae 771.9733 [2016-04-23]
b011e696
JH
78 - C90 declaration-after-statement error with darwin threads: blead de1003b4
79 [rt.cpan.org #113856]
f3dd6bae 80
02fb1e6b
JH
811.9732 [2016-03-13]
82 - MUTEX_LOCK, not PERL_MUTEX_LOCK: blead e5b02b5d
83 - also hrt_ualarm_itimer() is unused: 1cb6cce3
84 - the mutex needs init: 2d639e20
85 - version bump to 1.9732: 730d7fdc
86
e7185dd0 871.9731 [2016-03-13]
febe9c7a
JH
88 - mark unused variable as such: blead a914236c
89 - OS X emulation mutex accidentally unused: da7a6455
90 - remove hrt_ualarm: 6da77c36
91 - pod error fixes: 919ca095
92 - nanosleep and clock_nanosleep cleanups: c8ea02b8..e3ff671b
919ca095
JH
93 - static funcs in HiRes.xs: 4e922c26
94 - Remove unused variable: 52ffb1b5
95
1625f1d3
JH
961.9730 [2016-02-17]
97 - TIME_HIRES_DONT_RUN_PROBES=1 to build the probes but not run them
98 [rt.cpan.org #111391]
99
bcd2c0ef
JH
1001.9729 [2016-02-13]
101 - new constants
102 [rt.cpan.org #111170]
103 - clock_getres(), clock_gettime(), clock_nanosleep() emulation for OS X
104 [rt.cpan.org #78566]
105
df003a36
JH
1061.9728 [2015-11-13]
107 - No changes since 1.9727_03
108
1091.9727_03 [2015-11-06]
110 - use PERL_UNUSED_ARG() in the not-here-croak branches
111 - Makefile.PL/metadata improvements
112
1131.9727_02 [2015-08-29]
114 - correct documentation to note that Time::NVtime is NV, not double
115 [perl #123879]
116 - avoid a warning about clock_t on OS X / clang 3.6 - 3.6
117 [rt.cpan.org #102718]
118
1191.9727_01 [2015-08-17]
120 - Remove usage of 5.005 thread APIs in favor of 5.6 and up thread APIs
121 [rt.cpan.org #102799].
122 - Silence a C++ compiler warning [rt.cpan.org #101497].
123 - respect $Config{run} for cross-compiling [rt.cpan.org #92607]
124
1251.9726 [2013-08-17]
126 - Correct s/us splitting of usleep(1000000) [rt.cpan.org #78266].
127 - Avoid integer overflow in itimer-based alarm() with large
128 argument [rt.cpan.org #87160].
129 - Define PERL_NO_INLINE_FUNCTIONS during probe compilations, to
130 avoid false negatives caused by not linking with the perl core.
131 - Be more careful about context in stat().
132 - Install into "site" directories by default on perl 5.11+
133 [rt.cpan.org #79797].
134 - Fix a couple of doc typos [rt.cpan.org #85365].
135 - Fix function name in a doc example [rt.cpan.org #86318].
136 - Provide lstat() that yields high-res timestamps, alongside
137 the existing high-res stat() [rt.cpan.org #78732].
138
1391.9725 [2012-02-01]
140 - Correct stack discipline in stat(), which was screwing up list
141 operations in expressions containing calls to it [rt.cpan.org
142 #72926].
143 - Add missing OUTPUT sections to the XS code [rt.cpan.org #70930].
144 - Skip itimer tests on GNU/Hurd, which has the API but lacks
145 the implementation [rt.cpan.org #72754].
146 - Fix a doubled word in the documentation [rt.cpan.org #72763].
147
1481.9724 [2011-06-09]
149 - Correct XS parameter list, and therefore prototype, for
150 unimplemented-on-this-platform version of clock_nanosleep()
151 [rt.cpan.org #68700].
152 - Declare package variables with "our" rather than "use vars".
153 - Corresponding to "our" usage, check for minimum Perl version
154 5.006.
155 - Declare module dependencies.
156
1571.9723 [2011-06-07]
158 - Remove $ENV{PERL_CORE} logic from test suite, which is no
159 longer desired in the core.
160 - Convert test suite to use Test::More.
161 - Factor out watchdog code from test suite.
162 - In test suite, be consistent about using fully-qualified form
163 of function names.
164 - Divide test suite into feature-specific scripts.
165 - Make ualarm timing test less vulnerable to delay-induced false
166 failure, from Dave Mitchell.
167
1681.9722 [2011-05-18]
169 - Update for changes in build process in the core, patches
170 from BinGOs [rt.cpan.org #58858] and Craig Berry [rt.cpan.org
171 #63363].
172 - Fix broken linkage on Windows with gcc 3.4 seen with ActivePerl,
173 report from Christian Walde [rt.cpan.org #61648], fix derived
174 from Vincent Pit.
175 - Jump through hoops to avoid compiler warnings.
176
1771.9721 [2010-03-17]
178 - Address [rt.cpan.org #54196] alarm and ularm return values are bogus,
179 additional fix from Gisle Aas
180 - Address [rt.cpan.org #55665] "Bad plan" on Windows,
181 report and fix from Jan Dubois
182
1831.9720 [2010-02-14]
184 - Address [rt.cpan.org #54196] alarm and ularm return values are bogus,
185 report and fix from Nicholas Clark
186
1871.9719 [2009-01-04]
188 - As with QNX, Haiku has the API of interval timers but not
189 the implementation (bleadperl change #34630), hence skip
190 the tests, via David Mitchell.
191
1921.9718 [2008-12-31]
193 - .xs code cleanup from Albert Dvornik
194 - in the #39 and #40 do not do us I did, mixing alarm() and
195 sleep(). Now instead spin until enough time has passed.
196
1971.9717 [2008-12-30]
198 - Skip the tests added in 1.9716 (#39, #40) if there's no subsecond
199 alarm capability, like with the older subsecond alarm tests
200
2011.9716 [2008-12-26]
202 - Change documentation to agree with reality: there are
203 no interval timers in Win32.
204 - Address [rt.cpan.org #35899] (problem in subsecond sleeps),
205 add two tests to guard against this problem
206 - Address [rt.cpan.org #36600] 'Division by zero' failure in test suite
207 - Address [rt.cpan.org #37340] [PATCH] Address timer process in test
208 - Address [rt.cpan.org#40311 ] bad implementation of hrt_usleep
209 with TIME_HIRES_NANOSLEEP
210
2111.9715 [2008-04-08]
212 - Silly me: Makefile.PL does need to accept arguments other than mine.
213 Some testing frameworks obviously do this.
214 - Add retrying for tests 34..37, which are the most commonly
215 failing tests. If this helps, consider extending the retry
216 framework to all the tests. [Inspired by Slaven Rezic,
217 [rt.cpan.org #34711] Occasional failures of test 35 or 36 (FreeBSD)]
218
2191.9714 [2008-04-07]
220 - Under Perl 5.6.* NVgf needs to be "g", reported by Zefram,
221 it seems that ppport.h 3.13 gets this wrong.
222 - remove the check in Makefile.PL for 5.7.2, shouldn't be
223 (a) necessary (b) relevant
224 - add logic to Makefile.PL to skip configure/write Makefile
225 step if the "xdefine" file already exists, indicating that
226 the configure step has already been done, one can still
227 force (re)configure by "perl Makefile.PL configure",
228 or of course by "make clean && perl Makefile.PL".
229
2301.9713 [2008-04-04]
231 - for alarm() and ualarm() [Perl] prefer setitimer() [C]
232 instead of ualarm() [C] since ualarm() [C] cannot portably
233 (and standards-compliantly) be used for more than 999_999
234 microseconds (rt.cpan.org #34655)
235 - it seems that HP-UX has started (at least in 11.31 ia64)
236 #defining the CLOCK_REALTIME et alia (instead of having
237 them just as enums)
b88265bf 238 - document all the diagnostics
df003a36
JH
239
2401.9712 [2008-02-09]
241 - move the sub tick in the test file back to where it used to be
242 - in the "consider upgrading" message recommend at least Perl 5.8.8
243 and make the message to appear only for 5.8.0 since 5.8.1 and
244 later have the problem fixed
245 - VOS tweak for Makefile (core perl change #33259)
246 - since the test #17 seems to fail often, relax its limits a bit
247
2481.9711 [2007-11-29]
249 - lost VMS test skippage from Craig Berry
250 - reformat the test code a little
251
2521.9710 [2007-11-29]
253 - I got the sense of the QNX test the wrong way in an attempt
254 to generalize it for future
255
2561.9709 [2007-11-28]
257 - casting fixes from Robin Barker for g++ and 64bitint
258 - in QNX skip the itimer tests because though the API
259 is there, the implementation isn't, from Matt Kraai
260 - raise the dead man timer to 180 seconds for really
261 slow/busy systems
262 - elaborate the UTF-8 locale warning from Makefile.PL
263
2641.9708 [2007-10-05]
265 - [rt.cpan.org #29747]: Build failure with perl 5.005_05
266 Fixed by regenerating the ppport.h using Devel::PPPort 3.13.
267
2681.9707 [2007-02-27]
269 - t/HiRes.t failed in Perl 5.6.2,
270 "action is not of type POSIX::SigAction at t/HiRes.t line 318",
271 reported and fixed by Anton Berezin, the reason was faulty
272 use of sigaction() when restoring the old action.
273
2741.9706 [2007-02-25]
275 - with bleadperl in VMS the HiRes.t overrun the maximum number
276 of deferred signals because the libc SIGALRM was not strong
277 enough to interrupt select(), and select() got restarted every
278 time, solution is to use POSIX::SigAction if available.
279 A fix from Craig Berry (not 100% there, but helps).
280 - allow for more measuring noise for ualarm() tests 35..37
281
2821.9705 [2007-02-06]
283 - nanosleep() and clock_nanosleep() detection and use were
284 quite broken; in Linux -lrt needed; fixes from Zefram
285 - [internal] slightly cleaner building of $DEFINE in Makefile.PL,
286 should avoid double/conflicting -D flags
287
2881.9704 [2007-01-01]
289 - allow 10% of slop in test #14 (testing difference between
290 CORE::time() and Time::HiRes::time()), there seem to be often
291 transient failures from Perl smoke builds on this test
292 - small pod tweaks
293
2941.9703 [2006-12-08]
295 - use int main(int argc, char **argv) consistently in Makefile.PL,
296 should help with
297 [rt.cpan.org #23868] nanosleep not detected under Mac OS 10.3.9 starting with Time::HiRes 1.96
298 - if someone still has the locale-broken Perl 5.8.0,
299 suggest that they upgrade their Perl
300
3011.9702 [2006-12-06]
302 - restore the -DATLEASTFIVEOHOHFIVE, Win32 needed it still
303
3041.9701 [2006-12-04]
305 - upgrade to ppport.h 3.10_02
306 - remove the -DATLEASTFIVEOHOHFIVE
307 - use the ppport.h PL_ppaddr, PL_statcache, PL_laststatval
308 - use the ppport.h aTHXR for calling Perl stat()
309 - switch into four-digit version since 2.0 is coming up
310 awfully fast but not feeling like a major rewrite
311
3121.97 [2006-11-30]
313 - 1.95 broke building in Win32 (since pp_stat is not exported),
314 figured out how to call an op directly in 5.005 (use Perl_ppaddr
315 instead of PL_ppaddr)
316 - backport to Perl 5.004_05 (requires using statcache
317 and laststatval instead of PL_statcache and PL_laststatval)
318 (also checked to work in 5.005_04, 5.6.1, and 5.8.8 with threads)
319
3201.96 [2006-11-30]
321 - 1.95 broke builds for threaded Perls, rt.cpan.org tickets:
322 [rt.cpan.org #23694] Time::HiRes fails tests on Solaris and Perl 5.6.1
323 [rt.cpan.org #23712] Time-HiRes 1.95 Fails make on AIX 5.2 with Perl 5.8.8
324 [rt.cpan.org #23730] Time::HiRes 1.95 fails make on MacOS X 10.3.9/perl 5.8.8
325 - use main() prototype consistently in Makefile.PL
326
3271.95 [2006-11-29]
328 - integrate core change #29180: Silence VC++ compiler warnings
329 from Steve Hay
330 - do not use PL_ppaddr in stat() because that is not available
331 in Perl 5.005_04
332 - regenerate fallback/*.inc for older Perls without
333 ExtUtils::Constant because of d_hires_stat, resolves
334 [rt.cpan.org #23694] Time::HiRes fails tests on Solaris and Perl 5.6.1
335 - Make Makefile.PL more defensive against false PERL_CORE
336
3371.94 [2006-10-16]
338 - file timestamps oddities seen: the atime and mtime
339 can be out of sync (modify first and read second can leave
340 atime < mtime) and mtime can be subsecond while atime is not.
341 So make the test more forgiving.
342
3431.93 [2006-10-15]
344 - the ualarm() tests (34-37) assumed that ualarm(N)
345 could never alarm in less than N seconds, widened
346 the acceptable relative range to 0.9..1.5. Addresses
347 [rt.cpan.org #22090] and [rt.cpan.org #22091].
348
349 - skip the stat() tests in cygwin and win32, because
350 if run on FAT the timestamp granularity is only 2 seconds.
351 Any good way to detect (cygwin or win32) whether we are
352 being run on NTFS or anywhere with better timestamps?
353 Addresses [rt.cpan.org #22089] and [rt.cpan.org #22098].
354
3551.92 [2006-10-13]
356 - scan for subsecond resolution timestamps in struct stat,
357 some known possibilities:
358
359 (1) struct timespec st_atimespec;
360 st_atimespec.tv_nsec;
361 (2) time_t st_atime;
362 long st_atimensec;
363 (3) time_t st_atime;
364 int st_atime_n;
365 (4) timestruc_t st_atim;
366 st_atim.tv_nsec
367 (5) time_t st_atime;
368 int st_uatime;
369
370 If something like this is found, one can do
371
372 use Time::HiRes;
373 my @stat = Time::HiRes::stat();
374
375 or even override the standard stat():
b88265bf 376
df003a36
JH
377 use Time::HiRes qw(stat);
378
379 to get the stat() timestamps
380
381 my ($atime, $mtime, $ctime) = @stat[8, 9, 10];
382
383 with subsecond resolution (assuming both the operating
384 system and the filesystem support that kind of thing).
385
386 Contributions for more systems (especially non-UNIX,
387 e.g. but not limited to: Win32, VMS, OS/2) gladly accepted.
388 (also more UNIX variants welcome: HP-UX? IRIX?)
389
390 Thanks to H.Merijn Brand, John Peacock, and Craig
391 Berry for brave beta testing.
392
3931.91 [2006-09-29]
394 - ualarm() in SuSE 10.1 was overflowing after ~4.2 seconds,
395 possibly due to a glibc bug/feature (suspected overflow at
396 2**32 microseconds?), workaround by using the setitimer()
397 implementation of ualarm() if either useconds or
398 interval > 999_999 (this case seems to vary between systems:
399 are useconds more than 999_999 for ualarm() defined or not)
400 Added more ualarm() tests to catch various overflow points,
401 hopefully no problems in various platforms.
402 (The problem report by Mark Seger and Jon Paul Sullivan of HP.)
403
4041.90 [2006-08-22]
405 - tweak still needed for Const64(), from Jerry Hedden
406 - get a freshly generated ppport.h
407 - update Copyright years
408
4091.89 [2006-08-22]
410 - Const64() already appends an 'LL' (or i64), so provide LL and i64
411 forms for the IV_1E[679] (effects Win32 and Cygwin), reported by
412 Jerry Hedden.
413 - the Changes entry for 1.88 talked about [IN]V_1[679],
414 missing the 'E'.
415
4161.88 [2006-08-21]
417 - clean up the g++ warnings in HiRes.xs, all of them
418 about mixing integer and floating point, introduce
419 constants IV_1E[679] and NV_1E[679]
420
4211.87 [2006-02-13]
422 - [rt.cpan.org #17442] 'make test' frequently fails under
423 Cygwin Perl v5.8.8, reported and patched by J. R. Hedden
424 (two race condition bugs in the END block in the case the
425 main process dies before the timer process, unearthed
426 by a bug in Cygwin ualarm)
427
4281.86 [2005-12-17]
429 - HiRes.t:s/ok 32/ok 33/, from Dominic Dunlop
430 - tighten up the clock() test marginally by requiring non-negative
431 - clock_nanosleep() and clock() doc tweaks
432
4331.85 [2005-12-16]
434 - the interface to clock_nanosleep() is more natural
435 when it is like (hires) time() (instead of like nanosleep),
436 and the .xs implementation of clock_nanosleep() in 1.84
437 was broken anyway
438 - the semantics of clock() are not quite so silly as I thought,
439 but still somewhat odd, documented as such
440 - additional enhancements to the clock() documentation
441 - add test for clock_nanosleep() (I cannot test this
442 since none of my systems have the function)
443 - add test for clock()
444
4451.84 [2005-12-16]
446 - add clock() which returns the processor time in
447 (floating point) seconds since an arbitrary era
448 - add clock_nanosleep() which suspends the current
449 thread until either absolute time or for relative time
450 - [rt.cpan.org #16486] printf missing value in HiRes.t
451 - add constants CLOCKS_PER_SEC, CLOCK_SOFTTIME, TIMER_ABSTIME
452 - tiny typo fixes
453
4541.83 [2005-11-19]
455 - has_symbol() was wrong since e.g. ITIMER_VIRTUAL is exported
456 via @EXPORT_OK even when it is not available. This is heinous.
457 @EXPORT_OK should be determined at Makefile.PL time.
458 - be more lenient is testing clock_gettime(): allow more slop,
459 and retry up to three times, sleeping a random nap between
460 the retries
461 - human months are one-based (noticed by Anton Berezin)
462
4631.82 [2005-10-06]
464 - CLOCK_REALTIME is an enum value (of the clockid_t enum)
465 in HP-UX (and might be so elsewhere, too), debugged by
466 H. Merijn Brand
467 - include const-c.inc as late as possible (from Randy Kobes,
468 [rt.cpan.org #15552] to avoid undefined usleep() on Win32
469
4701.81 [2005-11-05]
471 - try to be more robust and consistent in the detection of
472 CLOCK_REALTIME and ITIMER_VIRTUAL in HiRes.t: the proper
473 way is
474
475 sub has_symbol {
476 my $symbol = shift;
477 eval 'import Time::HiRes qw($symbol)';
478 return 0 unless $@ eq '';
479 return exists ${"Time::HiRes::$symbol"};
480 }
481
482 and then use
483
484 &FOO_BAR
485
486 in the test. All these moves are needed because
487
488 1) one cannot directly do eval 'Time::HiRes::FOO_BAR'
489 because FOO_BAR might have a true value of zero
490 (or in the general case an empty string or even undef)
491
492 2) In case FOO_BAR is not available in this platform,
493 &FOO_BAR avoids the bareword warning
494
495 - wait more (1.5 seconds instead of 0.1) for the CLOCK_REALTIME test
496 but expect the 'customary' slop of 0.20 instead of 0.25
497 - fixed inside a comment HAS_POLL -> TIME_HIRES_NANOSLEEP
498 - at the end of HiRest.t tell how close we were to termination
499
5001.80 [2005-11-04]
501 - Gisle noticed a mistake (using HAS_NANOSLEEP) in 1.79
502
5031.79 [2005-11-03]
504 - try nanosleep for emulating usleep -- may help in some weird
505 embedded realtime places which have nanosleep but neither usleep
506 nor select nor poll (doesn't have to be weird embedded realtime
507 place, though -- in many places usleep is nanosleep anyway)
508 - try poll for emulating usleep -- this may help some obscure/old
509 SVR4 places that have neither usleep nor select
510 - a redundant test guard in HiRes.t
511
5121.78 [2005-11-03]
513 - ITIMER_VIRTUAL detection in HiRes.t had problems (that we cannot
514 in the general case fail already at 'use' phase is suboptimal)
515 - fixes to the documentation of clock_gettime() and clock_getres()
516
5171.77 [2005-11-03]
518 - add support for the POSIX clock_gettime() and clock_getres(),
519 if available, either as library calls or as syscalls
520 - be more defensive about missing functionality: break out
521 early (during 'use') if no e.g. clock_getres() is available,
522 and protect our back by trapping those cases also in HiRes.xs
523 - the test added in 1.76 could cause an endless loop e.g. in Solaris,
524 due to mixing of sleep() and alarm() (bad programmer, no cookie!)
525
5261.76 [2005-10-22]
527 - testing for nanosleep had wrong logic which caused nanosleep
528 to become undefined for e.g. Mac OS X
529 - added a test for a core dump that was introduced by Perl 5.8.0
530 safe signals and was fixed for the time of 5.8.1 (one report of
531 the core dump was [perl #20920]), the test skipped pre-5.8.1.
532 - *cough* s/unanosleep/nanosleep/g; *cough*
533
5341.75 [2005-10-18]
535 - installation patch from Gisle Aas: in Perls 5.8.x and later
536 use MakeMaker INSTALLDIRS value of 'perl' instead of 'site'.
537
5381.74 [2005-09-19]
539 - [cpan #14608] Solaris 8 perl 5.005_03 File::Spec module does not have method rel2abs
540 (the workaround is not to use rel2abs, should not be necessary)
541 - [cpan #14642] U2time wrongly exported on the C API
542 (patch supplied by the reporter, SALVA@cpan.org)
543 - add release dates to Changes
544
5451.73 [2005-08-16]
546 - Time::HiRes::nanosleep support for Solaris [PATCH]
547 (POSIX::uname() not available if building with core perl,
548 from Gisle Aas, via perl5-porters, perl change #25295)
549
5501.72 [2005-07-01]
551 - going back to the 1.68 loader setup (using DynaLoader)
552 since too many weird things starting breaking
553 - fix a typo in José Auguste-Etienne's name
554
5551.71 [2005-06-28]
556 - a thinko in the nanosleep() detection
557 - move more changes stuff from the README to Changes
558 - add -w to the Makefile.PL
559
5601.70 [2005-06-26]
561 - oops in 1.69 about @ISA (not affecting anything but silly)
562 - add copyright 2005 to HiRes.pm
563 - add copyright and license to HiRes.xs
564 - add copyrights 2003, 2004, 2005 to README
565
5661.69 [2005-06-25]
567 - actually run a test for nanosleep
568 (if there is no $Config{d_nanosleep}) since e.g. in AIX 4.2
569 it seems that one can link in nanosleep() but then calling
570 it fails instantly and sets errno to ENOSYS (Not implemented).
571 This may be fixable in the AIX case by figuring out the right
572 (realtime POSIX?) libs and whatnot, but in the general case
573 running a real test case is better. (Of course, this change
574 will no doubt run into portability problems because of the
575 execution step...) Note that because of hysterical raisins
576 most Perls do NOT have $Config{d_nanosleep} (scanning for
577 it by Configure would in many platforms require linking in
578 things like -lrt, which would in many platforms be a bad idea
579 for Perl itself).
580 (from José Auguste-Etienne)
581 - support XSLoader also since it's much faster
582 (from Alexey Tourbin)
583 - add SEE ALSO (BSD::Resource and Time::TAI64)
584
5851.68 [2005-05-14]
586 - somehow 1.67 had a lot of doubled lines (a major cut-and-paste
587 error suspected), but miraculously it still worked since the
588 doubling took place below the __END__ token
589 - undef Pause() before defining it to avoid redefinition warnings
590 during compilation in case perl.h had already defined Pause()
591 (part of perl change #24271)
592 - minor doc tweaks
593
5941.67 [2005-05-04]
595 - (internal) don't ignore the return value of gettimeofday()
596 - (external) return undef or an empty if the C gettimeofday() fails
597 (affects Time::HiRes gettimeofday() and the hires time())
598
5991.66 [2004-12-19]
600 - add nanosleep()
601 - fix the 'hierachy' typo in Makefile.PL [rt.cpan.org #8492]
602 - should now build in Solaris [rt.cpan.org #7165] (since 1.64)
603 - should now build in Cygwin [rt.cpan.org #7535] (since 1.64)
604 - close also [rt.cpan.org #5933] "Time::HiRes::time does not
605 pick up time adjustments like ntp" since ever reproducing it
b88265bf 606 (and therefore verifying a possible fix) in the same environment
df003a36
JH
607 has become rather unlikely
608
6091.65 [2004-09-18]
610 - one should not mix u?alarm and sleep (the tests modified
611 by 1.65, #12 and #13, hung in Solaris), now we just busy
612 loop executing an empty block
613 - in the documentation underline the unspecificity of mixing
614 sleeps and alarms
615 - small spelling fixes
616
6171.64 [2004-09-16]
618 - regenerate ppport.h with Devel::PPPort 3.03,
619 now the MY_CXT_CLONE is defined in ppport.h,
620 we no more need to do that.
621
622 - the test #12 would often hang in sigsuspend() (at least that's
623 where Mac OS X' ktrace shows it hanging). With the sleep()s
624 changed to sleep(1)s, the tests still pass but no hang after
625 a few hundred repeats.
626
6271.63 [2004-09-01]
628 - Win32 and any ithread build: ppport.h didn't define
629 MY_CXT_CLONE, which seems to be a Time-HiRes-ism.
630
6311.62 [2004-08-31]
632 - Skip testing if under PERL_CORE and Time::HiRes has not
633 been Configured (from Marcus Holland-Moritz, core change
634 #23246)
635 - Use ppport.h generated by Devel::PPPort 3.01,
636 allowing cutting away our own portability code.
637 - Don't use $ENV{PERL_CORE} for < 5.6.0.
638 - Don't use "for my $i" for <= 5.003.
639 - Don't use Pause() for <= 5.003.
640 - Can't use newSVpvf for <= 5.003.
641 (most of the changes from Marcus)
642
6431.61 [2004-08-21]
644 - Win32: reset reading from the performance counters every
645 five minutes to better track wall clock time (thanks to
646 PC timers being often quite bad), should help long-running
647 programs.
648
6491.60 [2004-08-15]
650 - Win32: Patch from Steve Hay
651 [PATCH] Re: [perl #30755] [Win32] Different results from Time::HiRes::gettimeofdayunder the debugger
652 to [perl #30755] reported by Nigel Sandever
653
654 - Cygwin: Use the Win32 recalibration code also in Cygwin if the
655 <w32api/windows.h> APIs are available. Cygwin testing by
656 Yitzchak Scott-Thoennes.
657
658 - Solaris: use -lposix4 to get nanosleep for Solaris 2.6,
659 after that keep using -lrt, patch from Alan Burlison,
660 bug reported in [cpan #7165]
661
6621.59 [2004-04-08]
663 - Change the Win32 recalibration limit to 0.5 seconds and tweak
664 the documentation to blather less about the gory details of the
665 Win32 implementation and more about the complications in general
666 of meddling with the system clock.
667
6681.58 [2004-04-08]
669 - Document the 1.57 change better.
670
6711.57 [2004-07-04]
672 - Win32/Cygwin/MinGW: if the performance counter drifts by more
673 than two seconds from the system clock (due to ntp adjustments,
674 for example), recalibrate our internal counter: from Jan Dubois,
675 based on [cpan #5933] by Jerry D. Hedden.
676
6771.56 [2004-29-02]
678 - Give a clearer message if the tests timeout (perl change #22253)
679 - Don't use /tmp or its moral equivalents (perl bug #15036,
680 perl change #22258)
681
6821.55 [2004-01-14]
683 - Windows: mingw32 patch from Mike Pomraning (use Perl's Const64()
684 instead of VC-specific i64 suffix)
685
6861.54 [2003-12-31]
687 - Solaris: like Tru64 (dec_osf) also Solaris need -lrt for nanosleep
688
6891.53 [2003-12-30]
690 - Windows: higher resolution time() by using the Windows
691 performance counter API, from Jan Dubois and Anton Shcherbinin.
692 The exact new higher resolution depends on the hardware,
693 but it should be quite a bit better than using the basic
694 Windows timers.
695
6961.52 [2003-10-28]
697 - In AIX (v?) with perl 5.6.1 the HiRes.t can hang after
698 the subtest 18. No known analysis nor fix, but added
699 an alarm (that requires fork() and alarm()) to the test.
700
7011.51 [2003-09-22]
702 - doc tweaks from mjd (perl change #20456)
703 - NCR MP-RAS hints file added (svr4.pl) (perl change #21249)
704
7051.50 [2003-08-02]
706 - add a message (for non-core builds) to Makefile.PL about
707 the LC_ALL=C workaround
708 - &Time::HiRes::d_nanosleep was broken (perl change #20131)
709 - the nanosleep() probe was broken (perl change #20061)
710 - use existence instead of definedness for feature probes
711 (perl change #20043)
712 - MPE/iX tweak (perl change #20042)
713 - do not use HAS_NANOSLEEP (perl change #19898)
714
7151.49 [2003-06-23]
716 - UVuf for non-IVSIZE platforms (from Keiichiro Nagano)
717 - OS/2 can always mix subsecond sleeps with signals
718 (part of perl change #19789)
719
7201.48 [2003-06-04]
721 - workaround for buggy gcc 2.95.3 in openbsd/sparc64
722 (perl change #19592)
723
7241.47 [2003-05-03]
725 - do not use -lrt in Linux (from March Lehmann, perl change #19449)
726 - unnecessary (nanosleep is in libc anyway)
727 - harmful (-lrt slows down execution)
728 - incompatible (with many distributions' pthreads)
729
7301.46 [2003-04-25]
731 - do not create files in blib directories under core
732 (perl change #19160, from rgs)
733 - detypo s/VTLARM/VTARLM/ (perl change #19328, from mjd)
734
7351.45 [2003-04-01]
736 - guarantee that $xdefine in HiRes.t is always defined
737 (perl change #19109, from IlyaZ)
738 - a cleaner way to detect PERL_CORE (perl change #19111,
739 from IlyaZ)
740
7411.44 [2003-03-30]
742 - add hints/irix.pl to turn off overly POSIX flags that
743 cause hide struct timespec to be hidden (and compilation
744 to fail) (bleadperl change #19085)
745 - documentation tweaks
746
7471.43 [2003-03-11]
748 - add c:/temp to the list of temp directories to probe
749 so that cygwin (and win*?) builds are happy. This was
750 needed at least in my cygwin 1.3.20/w2k setup.
751
7521.42 [2003-01-07]
753 - modernize the constants code (from Nicholas Clark)
754
7551.41 [2003-01-03]
756 - At some point the ability to figure our the correct incdir
757 for EXTERN.h (either a core perl build, or an installed perl)
758 had broken (which lead into all test compiles failing with
759 a core perl build, but thanks to the robustness of Makefile.PL
760 nothing of this was visible). The brokenness seemed to be
761 caused by $ENV{PERL_CORE} not being on for core builds?
762 Now stole a trick from the Encode that sets $ENV{PERL_CORE}
763 right, and both styles of build should work again.
764
7651.40 [2003-01-03]
766 - Nicholas Clark noticed that the my_catdir() emulation function
767 was broken (which means that we didn't really work for Perls
768 5.002 and 5.003)
769 - inspired by fixing the above made the whole Makefile.PL -w
770 and strict clean
771 - tightened up the Makefile.PL output, less whitespace
772
7731.39 [2003-10-20]
774 - fix from Craig Berry for better building in VMS with PERL_CORE
775
7761.38 [2003-10-13]
777 - no functional changes
778 - move lib/Time/HiRes.pm as Hires.pm
779 - libraries scanning was slightly broken (always scanned
780 for a library even when $Config{libs} already had it)
781
7821.37 [2003-09-23]
783 - Ray Zimmerman ran into a race condition in Mac OS X.
784 A 0.01-second alarm fired before the test expected.
785 The test first slept indefinitely (blocking for signals)
786 and only after that tested for the signal having been sent.
787 Since the signal had already been sent, the test #12 never
788 completed. The solution: test first, then block.
789 - default to being silent on all probing attempts, set the
790 environment variable VERBOSE to a true value to see the
791 details (the probing command and the possible errors)
792
7931.36 [2003-09-12]
794 - do not clear MAN3PODS in Makefile.PL (Radoslaw Zielinski)
795 - INSTALLDIRS => 'perl' missing which means that Time::HiRes
796 cannot be upgraded from CPAN to override the 5.8.0 version
797 (Guido A. Ostkamp)
798 - Time::HiRes 1.35 could not be dropped as-is to bleadperl
799 because the include directories did not adjust themselves
800 if $ENV{PERL_CORE} (Hugo van der Sanden)
801 - add documentation about the restart of select() under alarm()
802
8031.35 [2003-08-24]
804 - small documentation tweaks
805
806
8071.34 [2003-08-22]
808 - better VMS operation (Craig Berry)
809
8101.33 [2003-08-20]
811 - our time machine is accelerating: now works with Perl 5.004_01
812 (tried with 5.003_07 and 5.002 but I get segmentation faults
813 from running the Makefile.PL with those in Tru64 4.0D)
814
8151.32 [2003-08-20]
816 - backward compatibility (pre-5.6.0) tweaks:
817 - no XSLoader in 5.00503, use DynaLoader instead
818 - no SvPV_nolen, either
819 - no PerlProc_pause(), either
820 - now tested with 5.00404 and 5.00503
821 - Makefile.PL requires 5.00404 (no more 5.002)
822 - use nanosleep instead of usleep, if it is available (Wilson Snyder)
823 (this means that one can mix subsecond sleeps with alarms)
824 - because of nanosleep we probe for -lrt and -lposix4
825 - the existence of getitimer/nanosleep/setitimer/ualarm/usleep
826 is available by exportable constants Time::HiRes::d_func
827 (since older Perl do not have them in %Config, and even
828 5.8.0 does not probe for nanosleep)
829
8301.31 [2003-08-19]
831 - backward compatibility (pre-5.6.1) tweaks:
832 - define NV if no NVTYPE
833 - define IVdf if needed (note: the Devel::PPPort
834 in 5.8.0 does not try hard hard enough since
835 the IVSIZE might not be defined)
836 - define NVgf if needed
837 - grab the typemap from 5.8.0 for the NV stuff
838
839 1.31 and 1.32 add more backward compatibility (now all the way
840 back to Perl 5.00404), and using nanosleep() (if available) for
841 subsecond sleeps.
842
8431.30 [2003-08-16]
844
845 - release 1.29_02 as 1.30
846
847 1.30 adds all the changes made during the Perl 5.6->5.7->5.8
848 development cycle. Most notably portability across platforms has been
849 enhanced, and the interval timers (setitimer, getitimer) have been
850 added. Note that the version of Time::HiRes that is included in Perl
851 5.8.0 calls itself 1.20_00, but it is equivalent to this Time::HiRes
852 version. Note also that in 1.30 Wegscheid turns over the maintenance
853 to Jarkko Hietaniemi.
854
8551.29_02 [2003-08-16]
856
857 - fix a silly unclosed comment typo in HiRes.xs
858 - document and export REALTIME_REALPROF (Solaris)
859
8601.29_01 [2003-08-16]
861
862 - only getitimer(ITIMER_REAL) available in Cygwin and Win32
863 (need to patch this also in Perl 5.[89])
864 - remove CVS revision log from HiRes.xs
865
8661.29_00 [2003-08-14]
867
868 The following numbered patches refer to the Perl 5.7 changes,
869 you can browse them at http://public.activestate.com/cgi-bin/perlbrowse
870
871 - 17558: Add #!./perl to the .t
872 - 17201: linux + usemorebits fix, from Rafael Garcia-Suarez
873 - 16198: political correctness, from Simon Cozens
874 - 15857: doc tweaks, from Jarkko Hietaniemi
875 - 15593: optimization in .xs, from Paul Green
876 - 14892: pod fixes, from Robin Barker
877 - 14100: VOS fixes, from Paul Green
878 - 13422: XS segfault, from Marc Lehmann
879 - 13378: whether select() gets restarted on signals, depends
880 - 13354: timing constraints, again, from Andy Dougherty
881 - 13278: can't do subsecond alarms with ualarm;
882 break out early if alarms do not seem to be working
883 - 13266: test relaxation (cygwin gets lower hires
884 times than lores ones)
885 - 12846: protect against high load, from Jarkko Hietaniemi
886 - 12837: HiRes.t VMS tweak, from Craig A. Berry
887 - 12797: HiRes.t VMS tweak, from Charles Lane
888 - 12769: HiRes.t VMS tweak, from Craig A. Berry
889 - 12744: gcc vs MS 64-bit constant syntax, from Nick Ing-Simmons
890 - 12722: VMS ualarm for VMS without ualarm, from Charles Lane
891 - 12692: alarm() ain't gonna work if ualarm() ain't,
892 from Gurusamy Sarathy
893 - 12680: minor VMS tweak, from Charles Lane
894 - 12617: don't try to print ints as IVs, from Jarkko Hietaniemi
895 - 12609: croak on negative time, from Jarkko Hietaniemi
896 - 12595: Cygwin rounds up for time(), from Jarkko Hietaniemi
b88265bf 897 - 12594: MacOS Classic timeofday, from Chris Nandor
df003a36
JH
898 - 12473: allow for more than one second for sleep() and usleep()
899 - 12458: test tuning, relax timing constraints,
900 from Jarkko Hietaniemi
901 - 12449: make sleep() and usleep() to return the number
902 of seconds and microseconds actually slept (analogously
903 with the builtin sleep()), also make usleep() croak if
904 asked for more than 1_000_000 useconds, from Jarkko Hietaniemi
905 - 12366: Time::HiRes for VMS pre-7.0, from Charles Lane
906 - 12199: do not use ftime on Win32, from Gurusamy Sarathy
907 - 12196: use ftime() on Win32, from Artur Bergman
908 - 12184: fix Time::HiRes gettimeofday() on Win32, from Gurusamy Sarathy
909 - 12105: use GetSystemTime() on Win32, from Artur Bergman
910 - 12060: explain the 1e9 seconds problem, from Jarkko Hietaniemi
911 - 11901: UNICOS sloppy division, from Jarkko Hietaniemi
912 - 11797: problem in HiRes.t, from John P. Linderman
913 - 11414: prototype from Time::HiRes::sleep(), from Abhijit Menon-Sen
914 - 11409: Time::HiRes qw(sleep) failed, from Abhijit Menon-Sen
b88265bf 915 - 11270: dynix/ptx 4.5.2 hints fix, from Peter Prymmer
df003a36
JH
916 - 11032: VAX VMS s/div/lib\$ediv/ fix, from Peter Prymmer
917 - 11011: VAX VMS s/qdiv/div/ fix, from Peter Prymmer
918 - 10953: SCO OpenServer 5.0.5 requires an explicit -lc for usleep(),
919 from Jonathan Stowe
920 - 10942: MPE/IX test tweaks, from Mark Bixby
b88265bf 921 - 10784: unnecessary pod2man calls, from Andy Dougherty
df003a36
JH
922 - 10354: ext/ + -Wall, from Doug MacEachern
923 - 10320: fix the BOOT section to call myU2time correctly
924 - 10317: correct casting for AIX< from H. Merijn Brand
925 - 10119: document that the core time() may be rounding, not truncating
926 - 10118: test fix, from John Peacock
927 - 9988: long =item, from Robin Barker
928 - 9714: correct test output
929 - 9708: test also the scalar aspect of getitimer()
930 - 9705: Add interval timers (setitimer, getitimer)
931 - 9692: do not require at least 5.005 using XS
b88265bf 932
df003a36
JH
933 The following changes were made on top of the changes
934 made for Time::HiRes during the Perl 5.7 development
b88265bf 935 cycle that culminated in the release of Perl 5.8.0.
df003a36
JH
936
937 - add "require 5.005" to the Makefile.PL
938 - remove the REVISION section (CVS log) from HiRes.pm
939 - add jhi's copyright alongside Douglas'
940 - move HiRes.pm to lib/Time/
941 - move HiRes.t to t/
942 - modify HiRes.t to use $ENV{PERL_CORE}
943 - modify the original Time::HiRes version 1.20 Makefile.PL
944 to work both with Perl 5.8.0 and the new code with pre-5.8.0
945 Perls (tried with 5.6.1)
946 - tiny tweaks and updates in README and TODO
947 - bump the VERSION to 1.29
948
9491.20 Wed Feb 24 21:30 1999
b88265bf 950 - make our usleep and ualarm substitutes into hrt_usleep
df003a36
JH
951 and hrt_ualarm. This helps static links of Perl with other
952 packages that also have usleep, etc. From
953 Ilya Zakharevich <ilya@math.ohio-state.edu>
954 - add C API stuff. From Joshua Pritikin
955 <joshua.pritikin@db.com>
956 - VMS Makefile.PL fun. From pvhp@forte.com (Peter Prymmer)
957 - hopefully correct "-lc" fix for SCO.
958 - add PPD stuff
959
960 1.20 adds a platform neutral set of C accessible routines if you are
961 running 5.005+. All other changes are packaging changes and build
962 fixes(?) for statically linked Perl, SCO, and VMS.
963
9641.19 Tue Sep 29 22:30 1998
965 - put VMS gettimeofday() in. Patch is from Sebastian Bazley
966 <seb@stian.demon.co.uk>
967 - change GIMME_V to GIMME to help people with older versions of
968 Perl.
969 - fix Win32 version of gettimeofday(). It didn't affect anything,
970 but it confuses people reading the code when the return value
971 is backwards (0 is success).
972 - fix Makefile.PL (more) so that detection of gettimeofday is
973 more correct.
974
975 1.19 has better VMS support.
976
9771.18 Mon Jul 6 22:40 1998
978 - add usleep() for Win32.
979 - fix Makefile.PL to fix reported HP/UX feature where unresolved
980 externals still cause an executable to be generated (though no
981 x bit set). Thanks to David Kozinn for report and explanation.
982 Problems with the fix are mine :)
983
984 1.18 has limited Win32 support (no ualarm). Added usleep for Win32.
985 Probably buggy. I'm sure I'll hear.
986
9871.17 Wed Jul 1 20:10 1998
988 - fix setitimer calls so microseconds is not more than 1000000.
989 Hp/UX 9 doesn't like that. Provided by Roland B Robert, PhD.
990 - make Win32. We only get gettimeofday (the select hack doesn't
991 seem to work on my Win95 system).
b88265bf 992 - fix test 4 on 01test.t. add test to see if time() and
df003a36
JH
993 Time::HiRes::time() are close.
994
9951.16 Wed Nov 12 21:05 1997
996 - add missing EXTEND in new gettimeofday scalar code.
997
998 1.16+ should be closer to building out of the box on Linux. Thanks
999 to Gisle Aas for patches, and the ualarm equivalent using setitimer.
1000
1001 If your underlying operating system doesn't implement ualarm(), then
1002 a fake using setitimer() will be made. If the OS is missing usleep(),
1003 a fake one using select() will be made. If a fake can't be made for
1004 either ualarm() or usleep(), then the corresponding Perl function will
1005 not be available. If the OS is missing gettimeofday(), you will get
1006 unresolved externals, either at link- or run-time.
1007
1008 This is an improvement; the package used to not even build if
1009 you were missing any of these bits. Roderick Schertler
1010
1011 <roderick@argon.org> did all the conditional compilation stuff,
1012 look at HiRes.pm and the test suites; it's good educational reading.
1013
10141.15 Mon Nov 10 21:30 1997
1015 - HiRes.pm: update pod. Provided by Gisle Aas.
1016 - HiRes.xs: if gettimeofday() called in scalar context, do
1017 something more useful than before. Provided by Gisle Aas.
1018 - README: tell of xsubpp '-nolinenumber' woes. thanks to
1019 Edward Henigin <ed@texas.net> for pointing out the problem.
1020
10211.14 Wed Nov 5 9:40 1997
1022 - Makefile.PL: look for setitimer
1023 - HiRes.xs: if missing ualarm, but we have setitimer, make up
1024 our own setitimer. These were provided by Gisle Aas.
1025
10261.13 Tue Nov 4 23:30 1997
1027 - Makefile.PL: fix autodetect mechanism to do try linking in addition
1028 to just compiling; should fix Linux build problem. Fix was provided
1029 by Gisle Aas.
1030
10311.12 Sun Oct 12 12:00:00 1997
1032 - Makefile.PL: set XSOPT to '-nolinenumbers' to work around xsubpp bug;
1033 you may need to comment this back out if you have an older xsubpp.
1034 - HiRes.xs: set PROTOTYPES: DISABLE
1035
10361.11 Fri Sep 05 16:00:00 1997
1037 - Makefile.PL:
1038 Had some line commented out that shouldn't have been (testing
1039 remnants)
1040 - README:
1041 Previous version was corrupted.
1042
10431.10 Thu May 22 20:20:00 1997
1044 - HiRes.xs, HiRes.pm, t/*:
b88265bf 1045 - only compile what we have OS support for (or can
df003a36 1046 fake with select())
b88265bf 1047 - only test what we compiled
df003a36 1048 - gross improvement to the test suite
b88265bf 1049 - fix EXPORT_FAIL.
df003a36
JH
1050 This work was all done by Roderick Schertler
1051 <roderick@argon.org>. If you run Linux or
b88265bf
JH
1052 one of the other ualarm-less platforms, and you like this
1053 module, let Roderick know; without him, it still wouldn't
df003a36
JH
1054 be working on those boxes...
1055 - Makefile.PL: figure out what routines the OS has and
b88265bf 1056 only build what we need. These bits were written by Jarkko
df003a36
JH
1057 Hietaniemi <jhi@iki.fi>. Again, gratitude is due...
1058
10591.02 Mon Dec 30 08:00:00 1996
1060 - HiRes.pm: update documentation to say what to do when missing
1061 ualarm() and friends.
1062 - README: update to warn that ualarm() and friends need to exist
1063
10641.01 Fri Oct 17 08:00:00 1996
1065 - Makefile.PL: make XSPROTOARGS => '-noprototyopes'
b88265bf 1066 - HiRes.pm: put blank line between __END__ and =head1 so that
df003a36
JH
1067 pod2man works.
1068
10691.00 Tue Sep 03 13:00:00 1996
1070 - original version; created by h2xs 1.16