-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathChanges
6006 lines (5867 loc) · 290 KB
/
Changes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1.84 2025-??-??
- Explicit rsync protocol for default rsync URL
- Convert APC (deprecated) FTP to general purpose FTP sync:
- Make FTP Host and FTP Port configurable
- Do not sync symlinks (was breaking permissions)
- Promote again FTP sync as available sync
- Remove xt/sysinfo/sysinfo_macos.t that was moved to System::Info
1.83 2024-12-07
- "Hello World!" from tib
- Maintenance release: cleaning, moving some files in distribution/repository
1.82 2024-04-28T13:51:53+01:00 (70842fe => Abe Timmerman)
- (Abe Timmerman, Sun, 28 Apr 2024 13:51:53 +0100) Autocommit for
distribution Test-Smoke 1.82 (minor)
1.81_03 2024-04-28T10:03:23+01:00 (cdac604 => Abe Timmerman)
- (Abe Timmerman, Sun, 28 Apr 2024 10:03:23 +0100) Autocommit for
distribution Test-Smoke 1.81_03 (same)
1.81_02 2024-04-27T17:34:20+01:00 (5d04dd4 => Abe Timmerman)
- (Abe Timmerman, Sat, 27 Apr 2024 17:34:20 +0100) Autocommit for
distribution Test-Smoke 1.81_02 (same)
- (Abe Timmerman, Sun, 28 Apr 2024 10:00:50 +0100) Allow ':' in paths (for
Win32)
1.81_01 2024-04-27T16:44:46+01:00 (cb4aff5 => Abe Timmerman)
- (Abe Timmerman, Sat, 27 Apr 2024 16:44:46 +0100) Autocommit for
distribution Test-Smoke 1.81_01 (same)
- (Abe Timmerman, Sat, 27 Apr 2024 17:31:58 +0100) Add JSON as test
dependency, work around perl bug
1.81 2023-10-27T09:21:36+02:00 (c932d7e => Abe Timmerman)
- (Abe Timmerman, Fri, 27 Oct 2023 09:21:36 +0200) Autocommit for
distribution Test-Smoke 1.81 (minor)
- (Abe Timmerman, Wed, 25 Oct 2023 19:54:55 +0200) Add support for
Cpanel::JSON::XS
- We should also support Cpanel::JSON::XS, dunno if speed is a thing
- here...
- Nicked 'JSON' from JSON::MaybeXS
- (Abe Timmerman, Fri, 27 Oct 2023 09:24:40 +0200) Open new development
branch
- (Abe Timmerman, Mon, 30 Oct 2023 09:44:25 +0100) Merge branch
'feature/cpanel-json-xs' into development
- (Abe Timmerman, Wed, 25 Oct 2023 20:33:16 +0200) Add a Queue for reports
to repost
- Post reports to the API-funcion (https://$host/api/report)
- Archiving *must* be used (adir set)
- The qfile holds the commit-id's for each report to resend for the
- archive
- This is WorkInProgress:
- bin/tshandle_queue.pl
- Test::Smoke::App::HandleQueue
- (Abe Timmerman, Thu, 26 Oct 2023 17:21:58 +0200) Add the queue handler
mechanism
- (Abe Timmerman, Fri, 27 Oct 2023 17:16:36 +0200) FIX: MSWin32 cmd uses
'REM' for comments
- (Abe Timmerman, Sat, 28 Oct 2023 18:15:22 +0200) FIX, commas between
arguments...
- (Abe Timmerman, Sat, 28 Oct 2023 18:16:12 +0200) Add a question about
the new qfile.
- After we have determined that both archiving and posting to Perl5
- CoreSmokeDB is configured, we can ask about qfile.
- (Abe Timmerman, Fri, 27 Oct 2023 10:12:47 +0200) Make Test::Smoke
installable as regular Perl module
- In the past, most users of the software wanted the software installed in
- a separate directory, but we never created an easy way to install as a
- regular Perl module. This change makes it posible to do so:
- $ SMOKE_INST_SITE_LIB=1 perl Makefile.PL
- or
- $ perl Makefile.PL --site-lib
- Installing as a regular Perl module makes the scripts and documentation
- more readily available.
- (Abe Timmerman, Mon, 30 Oct 2023 10:12:52 +0100) All .pm files should
have `our $VERSION = '...';`
- Added the version in some files, changed the `use vars...` also to `our`
- in other files.
- (Abe Timmerman, Wed, 22 Nov 2023 14:01:07 +0100) FIX: t/poster-post.t
- the test was looking for "darwin", which is fine on my mac, but goes
- very wrong on all other platforms. Look for `$^O`, that's what we use as
- input in the test.
- (Abe Timmerman, Mon, 8 Jan 2024 15:51:28 +0100) RT#151070: README not
clear enough (and outdated)
- We have switched to `tsconfigsmoke.pl` since 1.80, so the documentation
- should reflect that.
- (H.Merijn Brand - Tux, Sat, 27 Apr 2024 12:46:15 +0200) Skip unused
lines in legenda for perlio_only report
- (Abe Timmerman, Sat, 27 Apr 2024 16:16:54 +0100) Merge branch
'rename-readme' of https://github.com/afresh1/Test-Smoke into
afresh1-rename-readme
- (Abe Timmerman, Sat, 27 Apr 2024 16:35:20 +0100) Merge remote-tracking
branch 'github/no-stdio'
- (Abe Timmerman, Sat, 27 Apr 2024 16:36:17 +0100) Merge branch
'development'
1.80_06 2023-10-24T18:03:34+02:00 (f194172 => Abe Timmerman)
- (Abe Timmerman, Tue, 24 Oct 2023 18:03:34 +0200) Autocommit for
distribution Test-Smoke 1.80_06 (test)
- (Abe Timmerman, Wed, 25 Oct 2023 15:50:22 +0200) Add a separate AUTHORS
file
- mostly: git shortlog --summary --email
1.80_05 2023-10-21T17:06:05+02:00 (4c6b101 => Abe Timmerman)
- (Abe Timmerman, Sat, 21 Oct 2023 17:06:05 +0200) Autocommit for
distribution Test-Smoke 1.80_05 (test)
- (Abe Timmerman, Mon, 23 Oct 2023 23:27:43 +0200) Restore the default
'smokecurrent' for config-prefix
- (Abe Timmerman, Tue, 24 Oct 2023 12:22:27 +0200) Default build configs
should be OS dependend
- We have these example/default build configurations files in the
- distribution, so we should use them during configsmoke.
- (Abe Timmerman, Tue, 24 Oct 2023 12:25:47 +0200) FIX some POD.
- (Abe Timmerman, Tue, 24 Oct 2023 13:16:31 +0200) Remove configsmoke.pl,
add configsmoke.pod and README2.md
- We have gathered all the information from the old `configsmoke.pl` and
- put it in a new `configsmoke.pod` file, this file is converted to
- README2.md
- (Abe Timmerman, Tue, 24 Oct 2023 17:58:41 +0200) Make sure
'configsmoke.pod' is installed on make install
- We refer to that documentation, so it should be available after
- installation.
- Also regenerate the README2.md
1.80_04 2023-10-19T13:53:17+02:00 (120827f => Abe Timmerman)
- (Abe Timmerman, Thu, 19 Oct 2023 13:53:17 +0200) Autocommit for
distribution Test-Smoke 1.80_04 (test)
- (Abe Timmerman, Sat, 21 Oct 2023 14:34:34 +0200) Change test a bit: more
diagnostics
- (Abe Timmerman, Sat, 21 Oct 2023 14:35:43 +0200) Remove extra print of
answer in prompt_yn()
- Although this was the correct(ed) answer for a Yes/No question, it was
- rather irritating to see the "Got ..." twice and it didn't contribute to
- clarity.
- (Abe Timmerman, Sat, 21 Oct 2023 14:38:39 +0200) Improve
MSWin32-schtasks cron handling
- We now look if we can find an existing entry in the schedule for this
- task. If it exists we warn about it and add " /F" to the command to
- force changes to the schedule.
- Also the add2cron-option gets improved defaults handling.
1.80_03 2023-10-17T12:08:02+02:00 (b6eaf1f => Abe Timmerman)
- (Abe Timmerman, Tue, 17 Oct 2023 12:08:02 +0200) Autocommit for
distribution Test-Smoke 1.80_03 (test)
- (Abe Timmerman, Thu, 19 Oct 2023 11:23:39 +0200) Change the syncer_git
test a bit
- Use a bare repo for upstream and edit stuff in a working clone
- Add some diagnostics
- Smoothen the set-up flow a bit
- (Abe Timmerman, Thu, 19 Oct 2023 11:25:48 +0200) FIX: Some MSWin32
scheduler and smoke-script stuff
- Change the name of the Task for MS schtasks (include prefix)
- Add a note about querying MS schtasks in the .cmd
- Make the .sh and .cmd version equal wrt. PATH
- keep the PATH that was set during tsconfigsmoke.pl
- (Abe Timmerman, Thu, 19 Oct 2023 13:38:17 +0200) FIX: ConfigSmoke stuff
- The MSWin32 schtasks wasn't used correctly
- Use of the wrong variable in write_as_<shell|bat>()
1.80_02 2023-10-15T11:47:28+02:00 (4448a59 => Abe Timmerman)
- (Abe Timmerman, Sun, 15 Oct 2023 11:47:28 +0200) Autocommit for
distribution Test-Smoke 1.80_02 (test)
- (Abe Timmerman, Tue, 17 Oct 2023 10:00:23 +0200) Add t/syncer_git.t that
got forgotten
- (Abe Timmerman, Tue, 17 Oct 2023 10:01:44 +0200) FIX: iffy rounding test
by adding more digits
- (Abe Timmerman, Tue, 17 Oct 2023 10:03:31 +0200) Update the
example-config for MSWin32
- We now have 64-bit and loads of other stuff.
- Test::Smoke::Util::Configure_win32() was updated to understand more
- switches, this is reflected here.
- (Abe Timmerman, Tue, 17 Oct 2023 11:47:27 +0200) There seem to be more
ways to set a proxy
- After more failing tests with a proxy, it turns out `ALL_PROXY` is also
- looked at by HTTP::Tiny (and in both upper- and lower-case).
1.80_01 2023-10-13T22:12:09+02:00 (7ea15ed => Abe Timmerman)
- (Abe Timmerman, Fri, 13 Oct 2023 22:12:09 +0200) Autocommit for
distribution Test-Smoke 1.80_01 (test)
- (Abe Timmerman, Sat, 14 Oct 2023 13:53:36 +0200) Make sure we always use
(double) quotes(")
- Commands and arguments that contain spaces or are allready delimited,
- should always go out to the "shell" with quotes(") (not apostrophes(')).
- (Abe Timmerman, Sat, 14 Oct 2023 19:57:41 +0200) FIX: GitHubIssue #78
- The `sprintf()` rounds-up all numbers with a fraction-bit >= 0.5, so it
- can round-up to 60 again. We now check that and correct.
- (Abe Timmerman, Sat, 14 Oct 2023 20:01:55 +0200) Merge branch
'metacpan-20230214' of https://github.com/jkeenan/Test-Smoke into
jkeenan-metacpan-20230214
1.80 2023-05-01T08:56:06+02:00 (3b74d8a => Abe Timmerman)
- (Abe Timmerman, Mon, 1 May 2023 08:56:06 +0200) Autocommit for
distribution Test::Smoke 1.80 (minor)
- (Abe Timmerman, Tue, 9 May 2023 13:54:24 +0200) Split the 'building' of
the commandline from run()
- We didn't properly quote the command or arguments - a quoted command was
- re-quoted - this undos the quoting on MSWin32 and makes commands with
- spaces break.
- We have now split off the building of the command line from the run
- method in order to test it properly.
- (Abe Timmerman, Tue, 9 May 2023 13:59:54 +0200) Use quote-meta for paths
- On MSWin32 the backslash (\) path-separator is interpreted as the escape
- char, this messes up the regex for matching them. We'll quote-meta the
- paths for the test.
- (Abe Timmerman, Tue, 9 May 2023 14:03:05 +0200) Remove HTTPS?_PROXY
while running the test
- A test-failure showing a 403 from a squid-proxy makes me think we need
- to temporarily remove the ENVironment variables HTTP_PROXY and
- HTTPS_PROXY if set.
- (Abe Timmerman, Tue, 9 May 2023 14:07:33 +0200) Use tempdir() for new
target dirs
- We shouldn't use the 't/' directory for targeting our destinations, a
- tempdir() is nicer and most likely always writable.
- Also add a test for the UNCify_path() function.
- (Abe Timmerman, Tue, 27 Jun 2023 16:28:30 +0200) Create option to skip
PERLIO=stdio
- On some platforms - e.g. HP-UX - stdio will not ever work again, we need
- a way to run the test suite without the stdio option. This will create
- problems for the reporter, but we will have to solve them in stage 2 of
- this change.
- (Abe Timmerman, Fri, 30 Jun 2023 14:35:44 +0200) Fix: Reporter and
Smoker
- Smoker: multi-locale
- Reporter: no-stdio
- (Abe Timmerman, Tue, 10 Oct 2023 14:27:52 +0200) FIX: some MSWin32 stuff
- mistakes in MSWin32 (ConfigSmoke)
- Syncer::Git -> MSWin32 needs an extra option for clone:
- `--config core.autocrlf input`
- (Abe Timmerman, Thu, 12 Oct 2023 18:22:54 +0200) Add -Accflags='' and
-Aldflags='' for MSWin32
- take into account that textfiles are now checked out without CRLF in
- git
- -Accflags='' to add compiler flags: sets BUILDOPTEXTRA
- -Aldflags='' to add linker flags: sets LINK_FLAGSEXTRA
- (Abe Timmerman, Fri, 13 Oct 2023 19:46:11 +0200) FIX: --config
core.autocrlf
- This needs an '=' inbetween the name and value for `git clone`
- (Abe Timmerman, Fri, 13 Oct 2023 19:49:46 +0200) Update
Configure_win32()
- Add -Dprefix as alias for -DINST_TOP which does the same
- Reimplement -Accflags to set BUILDOPTEXTRA
- Add -Aldflags=... and make it set PRIV_LINK_FLAGS
- (Abe Timmerman, Fri, 13 Oct 2023 19:57:03 +0200) FIX: we don't need
quotes in result of whereis()
- Commands with a space in their path/name should no longer have
- delimiting quotes, this is taken care of in Test::Smokke::Util::Execute
- (Abe Timmerman, Fri, 13 Oct 2023 20:10:45 +0200) Merge branch
'fix/mswin32-configure' into development
- (Abe Timmerman, Fri, 13 Oct 2023 21:47:17 +0200) FIX: POD, remove =back
without =over
- (Abe Timmerman, Fri, 13 Oct 2023 21:48:13 +0200) We're starting a new
beta cycle
- (Abe Timmerman, Fri, 13 Oct 2023 21:48:56 +0200) FIX: tests for
-Accflags
- Now that we have reimplemented -Accflags to set BUILDOPTEXTRA
- (introduced in 2005) we need to fix the test-makefiles and tests
- (Abe Timmerman, Fri, 13 Oct 2023 22:04:12 +0200) Merge branch
'feature/no-stdio' into development
- (Abe Timmerman, Fri, 13 Oct 2023 22:09:09 +0200) Exclude "perlio_only"
from pod-coverage in xt/03-pod-coverage.t
1.79_10 2023-04-30T13:30:06+02:00 (df7aa5b => Abe Timmerman)
- (Abe Timmerman, Sun, 30 Apr 2023 13:30:06 +0200) Autocommit for
distribution Test::Smoke 1.79_10 (test)
1.79_09 2023-04-29T15:14:46+02:00 (59570a0 => Abe Timmerman)
- (Abe Timmerman, Sat, 29 Apr 2023 15:14:46 +0200) Autocommit for
distribution Test::Smoke 1.79_09 (same)
- (Abe Timmerman, Sun, 30 Apr 2023 13:18:23 +0200) Fix some
ExtUtils::MakeMaker problems
- PERL_MM_OPT='INSTALL_BASE=...' interferes with our use of PREFIX, we
- now take that into account (and honour INSTALL_BASE)
- Set a minimum version for ExtUtils::MakeMaker to 6.55_03
- (BUILD_REQUIRES)
1.79_08 2023-04-27T10:14:41+02:00 (9889db4 => Abe Timmerman)
- (Abe Timmerman, Thu, 27 Apr 2023 10:14:41 +0200) Change the default
poster module to HTTP::Tiny
- The default for config is HTTP::Tiny, but the "sane default" was
- LWP::UserAgent, this is now also HTTP::Tiny.
- (Abe Timmerman, Sat, 29 Apr 2023 11:43:52 +0200) tsrepostjsn.pl: add
option to re-post a specific file
- We now have a new option '--jsonreport' that lets one post a specific
- file to the reports server.
- Also fixed a bug for the case where no '--configfile' was specified and
- the 'configfile_error' attribute didn't exist.
- (Abe Timmerman, Sat, 29 Apr 2023 14:22:38 +0200) BUGFIX:
https://github.com/abeltje/Test-Smoke/issues/74
- Older 'curl' versions (before 7.68.0) needed the '--globoff' switch to
- make the IPv6 syntax for [::1] work. We now check the version of curl to
- see if we need to add the switch.
- See also commit: f4b90b293a2e163f49b55fa2996a6368115e08b9
- (Abe Timmerman, Sat, 29 Apr 2023 15:10:33 +0200) Add 'curlargs' to
exceptions
- New app-options should be excluded from the pod-coverage test, they are
- not documented in POD.
1.79_07 2023-02-14T20:04:05+00:00 (3bfc222 => James E Keenan)
- (James E Keenan, Tue, 14 Feb 2023 20:04:05 +0000) Replace
search.cpan.org URLs with metacpan.org URLs
- ... since search.cpan.org has been gone for several years now.
- (James E Keenan, Tue, 14 Feb 2023 20:17:43 +0000) FAQ: Correct spelling
error
1.79_06 2022-08-30T13:37:30+02:00 (e5ff85e => Abe Timmerman)
- (Abe Timmerman, Tue, 30 Aug 2022 13:37:30 +0200) Autocommit for
distribution Test::Smoke 1.79_06 (test)
- (Abe Timmerman, Thu, 1 Sep 2022 08:43:49 +0200) FIX: typo
- Duseithreads (e before a)
- (Bram, Tue, 30 Aug 2022 22:13:13 +0200) Fix
`Test::Smoke::Util::set_local_patch`
- The `set_local_patch` function got 'broken' when the 'patchlevel.h'
- in the perl5 repo got detabified (committed on 17 Jan 2021).
- It no longer worked because the code was using: `/\t,NULL/`
- That worked fine when the tabs were used for indentation but it
- got silently broken when the tabs were replaces with spaces.
- (Bram, Tue, 30 Aug 2022 23:02:54 +0200) Update 'patchlevel.h' file used
in the test
- This was using an old patchlevel.h file, update it to a more
- recent version.
- Also add xt/perl/git_version.h, this is used by xt/perl/patchlevel.h
- and is needed for the compiling tests. (Without tests in xt/ fail.)
- (Bram, Tue, 30 Aug 2022 23:08:10 +0200) Patch the patchlevel.h copy
- Other test still expect the old format of patchlevel.h which
- contains a patchnum... Even the number is important for the
- tests...
- So re-add the patchnum and remove 'uncommitted-changes' (the tests
- can't deal with it..)
- Another option would be to clean up the other tests but that is
- a bigger task...
- (Abe Timmerman, Thu, 1 Sep 2022 09:10:29 +0200) carp() when setting
local patch fails
- This code was silently broken for about 20 months because
- patchlevel.h was detabified.
- To avoid the same problem in the future: carp() when adding
- the local patch to patchlevel.h fails we should keep an eye on the
- logfiles.
- (Abe Timmerman, Thu, 1 Sep 2022 13:40:47 +0200) Merge branch
'bram-perl-bram/windows-git-test-fixes'
- (Abe Timmerman, Fri, 2 Sep 2022 11:53:18 +0200) Add USE_MINGW_ANSI_STDIO
as default for Configure_win32()
- https://github.com/Perl/perl5/issues/20214
- I guess this is a case of "Doctor, it hurts when I do this...". We'll
- just switch it on (it won't hurt MSVC and benefits gcc) and provide and
- escape with `-UUSE_MINGW_ANSI_STDIO`. We should aim to build the best
- possible Perl on Windows that we can imo.
- (Abe Timmerman, Sun, 4 Sep 2022 20:48:14 +0200) New helper programme:
tsrepostjsn.pl
- Sometimes the reportserver isn't there or won't accept a report, now,
- one can repost that report with this tool.
- It can be interactive and will look in the archive directory (adir) for
- the jsn files one wants to resend.
- (Abe Timmerman, Mon, 5 Sep 2022 11:05:47 +0200) Also install the new
tsrepostjsn.pl script
- I forgot to put it in the EXE_FILES key for WriteMakefile().
1.79_05 2022-08-30T11:43:29+02:00 (e117112 => Abe Timmerman)
- (Abe Timmerman, Tue, 30 Aug 2022 11:43:29 +0200) Autocommit for
distribution Test::Smoke 1.79_05 (test)
- (Abe Timmerman, Tue, 30 Aug 2022 13:36:03 +0200) Remove debug code
(Test::LongString)
1.79_04 2022-08-24T14:10:34+02:00 (1d9a306 => Abe Timmerman)
- (Abe Timmerman, Wed, 24 Aug 2022 14:10:34 +0200) Autocommit for
distribution Test::Smoke 1.79_04 (test)
- (Abe Timmerman, Wed, 24 Aug 2022 14:25:39 +0200) Restore another patch
that I removed
- (Abe Timmerman, Tue, 30 Aug 2022 10:07:18 +0200) Drop support for
dmake.exe (MSWin32)
- dmake.exe is no longer supported where gmake.exe is, so we drop the
- support for dmake.
- (Abe Timmerman, Tue, 30 Aug 2022 10:18:05 +0200) FIX: Configure_win32()
was broken!
- With the introduction of gmake on MSWin32 Configure_win32() has been
- broken quite badly (assignments went from `*=` in dmake to `:=` in
- gmake; the regexen didn't look at that).
- I've also introduced more Configure switches to the platform:
- CHANGE:
- Duseithreads is now on by default,
- Uuseithreads to switch it explicitly off
- FIX:
- UWIN64 now works as it should
- ADD:
- Duse64bitint (-UWIN64)
- Duselongdouble (gcc)
- Dusequadmath (gcc)
- Dusesitecustomize
- Udefault_inc_excludes_dot
- (Abe Timmerman, Tue, 30 Aug 2022 11:28:49 +0200) Put the compiler
messages in the .out-file
- https://github.com/abeltje/Test-Smoke/issues/54
- Test::Smoke was relying on the logfile (created from STDOUT and STDERR
- redirecting) for its compilermessages. I now moved them to the .out-file
- which should be more reliable - and adjusted the calls to `grepccmsg`
- and `grepnonfatal`.
1.79_02 2022-08-18T20:47:54+02:00 (dac3861 => Abe Timmerman)
- (Abe Timmerman, Thu, 18 Aug 2022 20:47:54 +0200) Autocommit for
distribution Test::Smoke 1.79_02 (test)
- (Abe Timmerman, Wed, 24 Aug 2022 10:37:18 +0200) FIX: xtest Makefile
target
- 83dafe7e6effe5489 was too gnu-make specific and broke bsd-make
- It looks that this is a portable solution, it's just a developer helper
- (Bram, Thu, 18 Aug 2022 18:20:12 +0200) Remove test repository when
rsync isn't found
- In t/syncer_rsync.t a fixed test directory (t/tsgit) is used.
- At the end of the test it normally removes this directory but when
- some of the tests are skipped then it left this directory in place
- which caused the next test run to have a failure (on `git commit`).
- Alternative fixes:
- use a tmp dir like t/syncer_git.t does
- remove the directory at the start of the test if it exist
- (Bram, Thu, 18 Aug 2022 18:19:54 +0200) Configure git user/emailaddress
for the test repo
- The test create their own git repo and then adds some files in it and
- then it does a `git commit`.
- On Windows the `git commit` failed because it was unable to determine a
valid
- username/email-address. This resulted in extra test failures.
- > After creating a repo set a username and emailaddress in the config of
- that repo.
- (Bram, Thu, 18 Aug 2022 18:18:30 +0200) Abort tests when chdir to repo
dir failed
- The test init their own git repo and then does a `chdir()` to that
- directory.
- If that `chdir()` would fail then abort the test since it would/could
otherwise
- be modifying the Test-Smoke repo.
- (I haven't seen failures with `chdir()`, it's more a safe-guard to
prevent
- against altering the wrong repo)
- (Using `die` instead of `skip` at the request of @abeltje)
- (Bram, Thu, 18 Aug 2022 18:18:23 +0200) Check exit status of every git
command
- During testing (on Windows) several git commands failed but this
- wasn't obvious from the test output.
- > Adapt the test to check the exit status of git after every command.
- (Bram, Thu, 18 Aug 2022 22:25:47 +0200) Lowercase cpu type in smoke
report
- On my Windows laptop System::Info reports the cpu_type as 'AMD64'.
- Looking at the smoke db: other platforms appear to report it as 'amd64'.
- Lowercase it so that the Windows reports also show as 'amd64'.
- (Bram, Thu, 18 Aug 2022 22:39:31 +0200) Add '.mailmap' to ignore list of
MANIFEST check
- The '.mailmap' file is not included in the MANIFEST, it's also
- special cased in perl5.git/t/porting/manifest.t.
- The manifest test of Test::Smoke should also ignore it.
- All reports currently contain:
- MANIFEST messages:
- MANIFEST did not declare '.mailmap'
- which is irrelevant noise.
- (Bram, Fri, 19 Aug 2022 13:37:52 +0200) Filter out the default msg in
`user_skiped_tests`
- `bin/configsmoke.pl` adds a default message when creating the 'skip
tests' file.
- Filter out that message in the report to reduce some 'useless' info.
- For reference: the 'skip tests' file is used by
`...::Smoker::set_skip_tests`
- which does some more filtering (skip comments, ...).
- Exact Filtering that it does:
- $raw =~ m/^\s*#/ and next;
- $raw =~ s/(\S+).*/$1/s;
- if ($raw !~ m/\.t$/ and $raw !~ m/test\.pl$/) {
- next;
- }
- I opted not to do the same filtering and instead just filter the default
- message. If there are comments in the file then they could be useful
- to include in the report (i.e. if they explain why a test file is
skipped)
- (Abe Timmerman, Wed, 24 Aug 2022 12:34:58 +0200) Autocommit for
distribution Test::Smoke 1.79_03 (test)
- (Abe Timmerman, Wed, 24 Aug 2022 14:06:54 +0200) FIX: the test cluttered
Test-Smoke/.git/config
- I don't know what happened, but I guess I cherry-picked the commits in
- the wrong order. Sorry everyone; please check your local git config!
1.79_01 2022-08-17T08:59:35+02:00 (f05cccd => Abe Timmerman)
- (Abe Timmerman, Wed, 17 Aug 2022 08:59:35 +0200) Autocommit for
distribution Test::Smoke 1.79_01 (test)
- (Bram, Mon, 15 Aug 2022 21:07:39 +0200) Include custom hostname in smoke
report
- In commit 5920e4cbcef17d7b00206d52dddd5b8ba33aafea
- Author: abeltje <[email protected]>
- Date: Fri Apr 20 22:38:11 2018 +0200
- Solve https://github.com/abeltje/Test-Smoke/pull/37
- I don't like using ENVironment vars for legitimate settings in the
- software, so I'd rather use the configfile for that.
- Add a question in configsmoke.pl
- The option was added to set a custom hostname in the configfile.
- In that commit the custom hostname is used when sending the report to
- the smokedb (that is: in `Test::Smoke::Reporter::smokedb_data`).
- But the hostname is also used in the preamble of the report and that
- code was left unchanged so the real hostname still leaked in the report.
- Fix this and (also) use the custom hostname in the `preamble` function.
- (+ add a test for it, test file is based on t/report-usernote.t)
- (Note: commit 5920e4c was based on a PR and in the original commit
- the preamble function was also updated. That part of the
- change got lost when converting it to a config option.)
- (Bram, Mon, 15 Aug 2022 21:24:06 +0200) Use `done_testing()` in
t/report-usernote.t
- The test file was added in commit
792da5b7176f3141f5c45e54e16a4fa1d0b34230
- Author: Abe Timmerman <[email protected]>
- Date: Mon Apr 15 00:27:00 2013 +0200
- Implement variable position for the user_note:
- un_position == 'top' => at the start of the report
- un_position == <!top> => at the end of the report
- My guess: `done_testing()` was commented support older perls (with an
- oder versions of Test::More which lacked support for it).
- Today: plenty of other files in the repo are using `done_testing()` so
- there is no reason why t/report-usernote.t shouldn't be using it.
- (Bram, Thu, 18 Aug 2022 16:00:55 +0200) Skip tests with git when `git
init fails`
- When `git init` failed then the test would continue which is
undesirable.
- When running `make test` in the Test-Smoke git repo I observed:
- `git init` failed (because of the '-b' option, unknown in my git
version)
- there was a test failure since it checks the exit code
- it continues running git commands, including `git commit`
- Result: changes were committed to the Test-Smoke git repo and *not* to
the
- test repo that is normally used inside the test.
- Fix this by skipping the git tests when `git init` fails.
- (Bram, Thu, 18 Aug 2022 16:07:57 +0200) Silence default branch name
warning
- On a more recent version of git there is a warning about the
- default branch name. The warning can be silenced by
- using the '-b' option to set a (different) default branch name
- using the '-q' option
- In commit 0f7ceb8c4aa1f2f64a14b86f2fcbef32c7981787 the '-b' option
- was added to `git init` but that breaks compatibility with older git
- versions. These do not know the '-b' option and thus fail.
- To preserve backwards and future compatibility:
- use the '-q' option to silence the warning
- use `git -c init.defaultBranch=master init` to ensure the default branch
- name is master for the test repo.
- (Older git versions use 'master' by default, newer versions currently
- use 'master' but that may change in the future. So git-upstream decides
- to change default name then the '-c' should do the trick to keep it as
- 'master')
- (Abe Timmerman, Thu, 18 Aug 2022 17:21:50 +0200) Merge branch
'bram-perl-bram/git-init'
- https://github.com/abeltje/Test-Smoke/pull/61
- (Abe Timmerman, Thu, 18 Aug 2022 17:32:32 +0200) Merge branch
'bram/custom-hostname' of https://github.com/bram-perl/Test-Smoke into
bram-perl-bram/custom-hostname
- https://github.com/abeltje/Test-Smoke/pull/58
- (Abe Timmerman, Thu, 18 Aug 2022 17:36:54 +0200) Merge branch
'bram-perl-bram/custom-hostname' into bugfixes/by-bram
- https://github.com/abeltje/Test-Smoke/pull/58
- (Abe Timmerman, Thu, 18 Aug 2022 17:38:30 +0200) Merge branch
'bram/usernote-done-testing' of https://github.com/bram-perl/Test-Smoke
into bram-perl-bram/usernote-done-testing
- (Abe Timmerman, Thu, 18 Aug 2022 17:39:06 +0200) Merge branch
'bram-perl-bram/usernote-done-testing' into bugfixes/by-bram
- https://github.com/abeltje/Test-Smoke/pull/59
- (Abe Timmerman, Thu, 18 Aug 2022 18:17:52 +0200) Make sure all
filehandles are AUTOFLUSH during smoke
- Inspired by: https://github.com/abeltje/Test-Smoke/pull/60
- It appears that Windows likes it better if the filehandles are on
- autoflush during the actual smoke.
- (Abe Timmerman, Thu, 18 Aug 2022 18:21:19 +0200) Merge branch
'bugfixes/by-bram'
- Merge work by Bram:
- fix on Tux's fix for the git-defaultBranch (#61)
- autoflush for filehandles used during the actual smoke (not #60)
- fix forgotten `done_testing()` (#59(
- fix leak of actual hostname in the smoke-report (#58)
- (Bram, Thu, 18 Aug 2022 18:18:23 +0200) Check exit status of every git
command
- During testing (on Windows) several git commands failed but this
- wasn't obvious from the test output.
- > Adapt the test to check the exit status of git after every command.
- (Bram, Thu, 18 Aug 2022 19:00:23 +0200) Cache `$daemon->sockhost` in
t/poster-post.t
- On my Windows 10 system 't/poster-post.t' was stuck.
- Debugging showed that this *apparently* happened when it tried to get
- the value of `$daemon->sockhost` for the test with 'HTTP::Tiny'.
- Earlier calls for `$daemon->sockhost` (for the curl test) did succeed
- and did behave as expected...
- Sample output of a run that was stuck:
- # HTTP::Daemon (6.12): IPv6 (::1)
- # Temporary daemon at: http://[::1]:57965/
- ok 1 - An object of class 'Test::Smoke::Poster::LWP_UserAgent' isa
'Test::Smoke::Poster::LWP_UserAgent'
- ok 2 - write_json
- [2022-08-18 16:47:15Z] Posting to http://[::1]:57965/report via
LWP::UserAgent.
- [2022-08-18 16:47:15Z] Reading from (C:\Perl\Test-Smoke\t\testsuite.jsn)
- [2022-08-18 16:47:15Z] Report data: {"sysinfo":"MSWin32"}
- [2022-08-18 16:47:15Z] [CoreSmokeDB] {"id":42}
- ok 3 - Got id (LWP::Useragent: http://[::1]:57965/report)
- ok 4 - An object of class 'Test::Smoke::Poster::Curl' isa
'Test::Smoke::Poster::Curl'
- ok 5 - write_json
- [2022-08-18 16:47:15Z] Posting to "http://[::1]:57965/report" via curl.
- [2022-08-18 16:47:15Z] Reading from (C:\Perl\Test-Smoke\t\testsuite.jsn)
- [2022-08-18 16:47:15Z] Report data: {"sysinfo":"MSWin32"}
- [2022-08-18 16:47:15Z] In pwd(C:/Perl/Test-Smoke) running:
- [2022-08-18 16:47:15Z] qx[C:\Windows\system32\curl.EXE --globoff -A
"Test::Smoke/1.79_01 (Test::Smoke::Poster::Curl)" -d
json=%7B%22sysinfo%22%3A%22MSWin32%22%7D "http://[::1]:57965/report"]
- % Total % Received % Xferd Average Speed Time Time Time
Current
- Dload Upload Total Spent Left Speed
- 100 49 100 9 100 40 522 2322 --:--:-- --:--:-- --:--:--
3062
- [2022-08-18 16:47:15Z] [CoreSmokeDB] {"id":42}
- ok 6 - Got id (curl: http://[::1]:57965/report)
- And then it kept on waiting..
- Caching the value of `$daemon->sockhost` appears to fix it..
- (I've got no idea why tho)
- With this patch applied and re-running the test it continues:
- ok 7 - An object of class 'Test::Smoke::Poster::HTTP_Tiny' isa
'Test::Smoke::Poster::HTTP_Tiny'
- ok 8 - write_json
- [2022-08-18 16:56:15Z] Posting to http://[::1]:57990/report via
HTTP::Tiny.
- [2022-08-18 16:56:15Z] Reading from (C:\Perl\Test-Smoke\t\testsuite.jsn)
- [2022-08-18 16:56:15Z] Report data: {"sysinfo":"MSWin32"}
- [2022-08-18 16:56:15Z] [CoreSmokeDB] {"id":42}
- ok 9 - Got id (HTTP::Tiny: http://[::1]:57990/report
- ok 10 - no warnings
- 1..10
- (Abe Timmerman, Thu, 18 Aug 2022 19:20:38 +0200) Merge branch
'bram/windows-poster_post.t' of https://github.com/bram-perl/Test-Smoke
into bram-perl-bram/windows-poster_post.t
- (Abe Timmerman, Thu, 18 Aug 2022 19:21:05 +0200) Merge branch
'bram-perl-bram/windows-poster_post.t' into bugfixes/by-bram
- https://github.com/abeltje/Test-Smoke/pull/63
- (Abe Timmerman, Thu, 18 Aug 2022 20:24:53 +0200) Add a target to the
makefile to run the xt/ tests
- No convention but on IRC#smoke:
- 17:41:14 Bram | Is there a make target to run the tests in xt/?
- 19:26:26 abeltje | no make target for running tests in xt/; is there a
convention for that?
- 19:29:36 ilmari | dzil has an `xtest` command to run them
- So I named it 'xtest' and used XTEST_FILES to select files to run:
- make xtest TEST_VERBOSE=1 XTEST_FILES=xt/0*.t
- but this also works:
- make xtest
- (Bram, Thu, 18 Aug 2022 18:18:30 +0200) Abort tests when chdir to repo
dir failed
- The test init their own git repo and then does a `chdir()` to that
- directory.
- If that `chdir()` would fail then abort the test since it would/could
otherwise
- be modifying the Test-Smoke repo.
- (I haven't seen failures with `chdir()`, it's more a safe-guard to
prevent
- against altering the wrong repo)
- (Using `die` instead of `skip` at the request of @abeltje)
- (Bram, Thu, 18 Aug 2022 18:19:54 +0200) Configure git user/emailaddress
for the test repo
- The test create their own git repo and then adds some files in it and
- then it does a `git commit`.
- On Windows the `git commit` failed because it was unable to determine a
valid
- username/email-address. This resulted in extra test failures.
- > After creating a repo set a username and emailaddress in the config of
- that repo.
- (Bram, Thu, 18 Aug 2022 18:20:12 +0200) Remove test repository when
rsync isn't found
- In t/syncer_rsync.t a fixed test directory (t/tsgit) is used.
- At the end of the test it normally removes this directory but when
- some of the tests are skipped then it left this directory in place
- which caused the next test run to have a failure (on `git commit`).
- Alternative fixes:
- use a tmp dir like t/syncer_git.t does
- remove the directory at the start of the test if it exist
- (Abe Timmerman, Thu, 18 Aug 2022 20:41:55 +0200) Merge branch
'feature/xtest-make-target'
1.79 2022-03-30T11:19:47+02:00 (6c102ab => Abe Timmerman)
- (Abe Timmerman, Wed, 30 Mar 2022 11:19:47 +0200) Autocommit for
distribution Test::Smoke 1.79 (minor)
- (H.Merijn Brand - Tux, Tue, 16 Aug 2022 08:56:53 +0200) Make git use a
different default branch name to prevent warnings
- ok 1 - Git version 2.37.2
- [2022-08-16 08:48:10+0200] qx[/usr/bin/git init /tmp/B5mqfRwdHq/tsgit]
- hint: Using 'master' as the name for the initial branch. This default
branch name
- hint: is subject to change. To configure the initial branch name to use
in all
- hint: of your new repositories, which will suppress this warning, call:
- hint:
- hint: git config --global init.defaultBranch <name>
- hint:
- hint: Names commonly chosen instead of 'master' are 'main', 'trunk' and
- hint: 'development'. The just-created branch can be renamed via this
command:
- hint:
- hint: git branch -m <name>
1.78 2020-05-20T16:50:59+02:00 (2081d02 => Abe Timmerman)
- (Abe Timmerman, Wed, 20 May 2020 16:50:59 +0200) Autocommit for
distribution Test::Smoke 1.78 (minor)
- (Andrew Hewus Fresh, Sun, 28 Jun 2020 10:42:18 -0700) Rename README so
GitHub will render correctly
- (Abe Timmerman, Mon, 28 Mar 2022 10:51:34 +0200) perl-5.26+ doesn't have
current-dir in @INC
- Fix the fake perl so it can find TestInit.pm again.
- (Abe Timmerman, Sun, 27 Mar 2022 13:27:49 +0200) RT#141857: don't use
the git:// protocol for defaults to GitHub
- GitHub stopped support for unauthenticated ssh/git access, so we switch
- to https.
- (Abe Timmerman, Mon, 28 Mar 2022 11:00:59 +0200) Autocommit for
distribution Test::Smoke 1.78_01 (test)
1.77_04 2020-05-20T12:28:49+02:00 (9c4724d => Abe Timmerman)
- (Abe Timmerman, Wed, 20 May 2020 12:28:49 +0200) Autocommit for
distribution Test::Smoke 1.77_04 (test)
- (Abe Timmerman, Wed, 20 May 2020 16:45:59 +0200) More cleanup of
HTTP::Lite: 28f09320 wasn't enough
- We really dropped support for HTTP::Lite
1.77_03 2020-05-19T23:11:59+02:00 (40678ef => Abe Timmerman)
- (Abe Timmerman, Tue, 19 May 2020 23:11:59 +0200) Autocommit for
distribution Test::Smoke 1.77_03 (test)
- (Abe Timmerman, Wed, 20 May 2020 12:22:50 +0200) More poster-post
problems (HTTP::Tiny)
- IPv6 is not supported in HTTP::Tiny < 0.042. Alas 5.24.[1-4] shipped
- with a DEV-version of the module. We must now use `version->parse()` to
- compare `$HTTP::Tiny::VERSION`
1.77_02 2020-05-19T15:20:21+02:00 (e978fd4 => Abe Timmerman)
- (Abe Timmerman, Tue, 19 May 2020 15:20:21 +0200) Autocommit for
distribution Test::Smoke 1.77_02 (test)
- (Abe Timmerman, Tue, 19 May 2020 15:29:16 +0200) State MIN_PERL_VERSION
in Makefile.PL
- 6 years overdue... sorry Neil
- (Abe Timmerman, Tue, 19 May 2020 16:08:36 +0200) RT#132507: Remove the
version number
- (Abe Timmerman, Tue, 19 May 2020 22:56:34 +0200) More poster-post
problems (curl)
- Older curls do need some help to understand the 'http://[::1]:54321'
- syntax, so quote the argument and use the '--globoff' switch
- (Abe Timmerman, Tue, 19 May 2020 23:00:00 +0200) Merge branch
'cleanup/remove_support_HTTP_Lite'
1.77_01 2020-05-19T11:52:30+02:00 (d959cb4 => Abe Timmerman)
- (Abe Timmerman, Tue, 19 May 2020 11:52:30 +0200) Test::Smoke::LogMixin:
no sprintf without arguments
- The log_* routines use the sprintf() way of creating log-lines. When
- there are no arguments, there is no need to use sprintf() and risk
- warnings about missing arguments or invalid conversion for bare % signs.
- (I ran into this with the output of `git clone`)
- (Abe Timmerman, Tue, 19 May 2020 13:05:13 +0200)
Test::Smoke::Syncer::Git: fix problem with the branchfile
- The first line was read but not chomp()ed, so if one had (mistakenly)
- put multiple lines in the file, the command line for the `git checkout`
- would have been wrong and that branch would not have been checked out.
- I hope that this also resolves part of RT#128950
- (Abe Timmerman, Tue, 19 May 2020 14:43:40 +0200) Prefer JSON::PP over
JSON::XS
- (Abe Timmerman, Tue, 19 May 2020 14:46:53 +0200) Provide better flow and
diagnostics for the Posters
- Untangle the actual POST and json-decode in the `post`-method.
- Also make the clients die from the `_post_data`-method if the request
- didn't return a 2xx.
- (Abe Timmerman, Tue, 19 May 2020 14:44:25 +0200) Provide better
diagnostics on failure with TEST_VERBOSE
- This provides invalid TAP, but compile-errors are worse!
1.77 2020-05-18T18:55:53+02:00 (046d80a => Abe Timmerman)
- (Abe Timmerman, Mon, 18 May 2020 18:55:53 +0200) Remove support for
HTTP::Lite
- We ran into a bug for HTTP::Lite that was reported 6 years ago, that
- will not be fixed anymore. No need to keep support for this module
- around now that HTTP::Tiny is in core (>= 5.14).
1.76_01 2020-05-13T15:51:25+02:00 (c5bf6ca => Abe Timmerman)
- (Abe Timmerman, Wed, 13 May 2020 15:51:25 +0200) Autocommit for
distribution Test::Smoke 1.76_01 (test)
1.76 2020-05-12T20:54:37+02:00 (0801dd6 => Abe Timmerman)
- (Abe Timmerman, Tue, 12 May 2020 20:54:37 +0200) Autocommit for
distribution Test::Smoke 1.76 (minor)
- (H.Merijn Brand, Tue, 23 Apr 2019 15:47:51 +0200) Add support for -Uxx
and -Dxx=y on tssmokeperl.pl
- When a smokescript smokes for two different compilers, it is likely to
use
- something like
- :
- /pro/bin/perl tssmokeperl.pl --nofetch --nopatch \
- c "$CFGNAME" $continue $* -Dcc=gcc > p59.log 2>&1
- :
- :
- /pro/bin/perl tssmokeperl.pl --nofetch --nopatch \
- c "$CFGNAME" $continue $* -Dcc=g++ > p59.log 2>&1
- :
- This change allows tssmokeperl.pl to accept -Dcc=gcc and/or -Ucc
- and make it DWIM. Before this change, these command-line options
- were silently ignored.
- Only existing options can be (un)set
- (H.Merijn Brand - Tux, Thu, 25 Apr 2019 11:55:40 +0200) Moving -Dxx=...
and -Uxx to the correct spot
- (H.Merijn Brand - Tux, Thu, 25 Apr 2019 12:25:25 +0200) Also allow -A
options
- (Abe Timmerman, Wed, 13 May 2020 15:18:21 +0200) Add a Serialiser
- This is a simple serialiser, I still have to hunt the rest of the code
- for uses of Data::Dumper type serialisations and see if I can change
- them.
- (Abe Timmerman, Wed, 13 May 2020 15:23:03 +0200) Add a '--pass_option'
switch to the runsmoke-applet
- This (repeatable) option will let one pass extra options to Configure
- from the 'tsrunsmoke.pl' (and thus 'tssmokeperl.pl') commandline:
- ~/coresmoke/ts/bin/tssmokeperl.pl -c "$CFGNAME" $* \
- -pass-option '-Dusesuperthreads' \
- p '-Uusesuperfiles' > p59.log 2>&1
- This is what Tux put in a commit message for a different solution of
- his own request:
- When a smokescript smokes for two different compilers, it is likely to
- use something like
- /pro/bin/perl tssmokeperl.pl --nofetch --nopatch \
- c "$CFGNAME" $continue $* -Dcc=gcc > p59.log 2>&1
- /pro/bin/perl tssmokeperl.pl --nofetch --nopatch \
- c "$CFGNAME" $continue $* -Dcc=g++ > p59.log 2>&1
- This change allows tssmokeperl.pl to accept -Dcc=gcc and/or -Ucc
- and make it DWIM. Before this change, these command-line options
- were silently ignored.
1.75 2020-03-01T12:41:23+01:00 (995b2f9 => Abe Timmerman)
- (Abe Timmerman, Sun, 1 Mar 2020 12:41:23 +0100) Autocommit for
distribution Test::Smoke 1.75 (same)
- (H.Merijn Brand - Tux, Wed, 6 Nov 2019 14:21:04 +0100) Ignore all
date/time information in dot_patch
- As we are only interested in the branch (first field), the sha and the
- describe (last two fields), just take those and ignore everything in
- between, so we cannot be fooled by (missing) time-zones or date formats
- that do not match our expectation
1.74 2020-02-09T16:36:26+01:00 (f119510 => Abe Timmerman)
- (Abe Timmerman, Sun, 9 Feb 2020 16:36:26 +0100) Autocommit for
distribution Test::Smoke 1.74 (minor)
- (Abe Timmerman, Sun, 1 Mar 2020 12:36:22 +0100) FIX: add dependency
introduced by be0b7a9438bdc9b25176dc8676641aff3bfc6413
- We forgot to add Capture::Tiny to the dependencies in Makefile.PL
1.73 2019-08-13T09:33:02+02:00 (b6ca9bd => Abe Timmerman)
- (Abe Timmerman, Tue, 13 Aug 2019 09:33:02 +0200) Autocommit for
distribution Test::Smoke 1.73 (minor)
- (James E Keenan, Sat, 8 Feb 2020 14:29:57 -0500) Update location of Perl
5 repository
- Note that I did not attempt to update all instances of the old
- repository in bin/configsmoke.pl.
- Needs careful code review.
- For: https://rt.cpan.org/Ticket/Display.html?id=131711
- (Abe Timmerman, Sun, 9 Feb 2020 15:50:07 +0100) Merge branch 'master' of
github.com:abeltje/Test-Smoke
- (Abe Timmerman, Sun, 9 Feb 2020 15:52:00 +0100) Merge branch
'rtc-131711-perl5-repository' of https://github.com/jkeenan/Test-Smoke
into jkeenan-rtc-131711-perl5-repository
- (Abe Timmerman, Sun, 9 Feb 2020 15:52:34 +0100) Merge branch
'jkeenan-rtc-131711-perl5-repository'
1.72_07 2019-08-07T13:10:16+03:00 (2b62635 => Abe Timmerman)
- (Abe Timmerman, Wed, 7 Aug 2019 13:10:16 +0300) Autocommit for
distribution Test::Smoke 1.72_07 (test)
1.72_05 2018-04-22T13:15:31+02:00 (91ae91a => abeltje)
- (abeltje, Sun, 22 Apr 2018 13:15:31 +0200) Autocommit for distribution
Test::Smoke 1.72_05 (test)
- (James E Keenan, Wed, 8 Aug 2018 09:42:01 -0400) Report all failing unit
tests and sequences thereof.
- The 'log' and 'out' files were correctly identifying failed unit tests
and
- sequences thereof, e.g.:
- ../lib/locale.t.............................................FAILED
- 436-437, 441, 444, 458
- However, only the first of these was being captured by
- Test::Smoke::Reporter::_parse() -- called from within
- Test::Smoke::Reporter::read_parse() -- and, hence, only the first was
- available for recording in the 'jsn' file:
- "failures" : [
- {
- "status" : "FAILED",
- "extra" : [
- "436-437"
- ],
- "test" : "../lib/locale.t"
- }
- ],
- And since it is the 'jsn' file that is sent to the CoreSmoke database --
from
- which the web display draws its data -- the web page was only showing
this:
- Test failures:
- ~~ ../lib/locale.t ................................ FAILED 436-437
- Cf. http://perl5.test-smoke.org/report/68034
- This commit improves the regex capturing the numbers of failing unit
tests and
- sequences of such tests. In t/reporter.t one block of tests similar to
- existing blocks is added.
- The commit also adds to .gitignore one file created during testing.
- For: RTC 125932
- (James E Keenan, Sun, 12 Aug 2018 14:12:36 -0400) Correct typo in 4 log
messages.
- For: rt.cpan.org 126045
- (James E Keenan, Tue, 14 Aug 2018 09:22:57 -0400) Correct spelling of
'omitted' in 5 locations.
- (James E Keenan, Wed, 8 Aug 2018 09:42:01 -0400) Report all failing unit
tests and sequences thereof.
- For: https://rt.cpan.org/Ticket/Display.html?id=125932
- In directory logs/smokecurrent/, the '.log' and '.out' files were
correctly
- identifying failed unit tests and sequences thereof, e.g.:
- ../lib/locale.t.............................................FAILED
- 436-437, 441, 444, 458
- However, only the first of these unit test numbers or sequences was
being
- captured by Test::Smoke::Reporter::_parse() -- called from within
- Test::Smoke::Reporter::read_parse(). Hence, only the first was
- available for recording in the '.jsn' file:
- "failures" : [
- {
- "status" : "FAILED",
- "extra" : [
- "436-437"
- ],
- "test" : "../lib/locale.t"
- }
- ],
- And since it is the '.jsn' file that is sent to the CoreSmoke database
-- from
- which the web display draws its data -- the web page was only showing
this:
- Test failures:
- ~~ ../lib/locale.t ................................ FAILED 436-437
- Cf. http://perl5.test-smoke.org/report/68034
- This commit improves the regex capturing the numbers of failing unit
tests and
- sequences of such tests. In t/reporter.t one block of tests similar to
- existing blocks is added.
- The commit also adds to .gitignore one file created during testing.
- For: RTC 125932
- (James E Keenan, Wed, 15 Aug 2018 10:22:16 -0400) Merge branch
'rt-125932-defective-json' of github.com:jkeenan/Test-Smoke into
rt-125932-defective-json
- (Abe Timmerman, Fri, 17 Aug 2018 19:11:14 +0100) Merge branch
'jkeenan-typo-omitted-20180814'
- (Abe Timmerman, Fri, 17 Aug 2018 19:12:16 +0100) Merge branch
'rtc-126045-exits-test-smoke-archiver-20180812' of
https://github.com/jkeenan/Test-Smoke into
jkeenan-rtc-126045-exits-test-smoke-archiver-20180812
- (Abe Timmerman, Fri, 17 Aug 2018 19:15:03 +0100) Merge branch
'jkeenan-rtc-126045-exits-test-smoke-archiver-20180812'
- (Abe Timmerman, Fri, 17 Aug 2018 19:23:29 +0100) Merge branch
'rt-125932-defective-json' of https://github.com/jkeenan/Test-Smoke into
jkeenan-rt-125932-defective-json
- (Abe Timmerman, Fri, 17 Aug 2018 19:23:52 +0100) Merge branch
'jkeenan-rt-125932-defective-json'
- (James E Keenan, Fri, 17 Aug 2018 16:28:07 -0400) Correct one typo in
Test::Smoke::App::Base docs.
- (Abe Timmerman, Fri, 9 Nov 2018 08:02:52 +0100) Set autoflush on STDOUT
for all the scripts
- The output from the smoke-scripts was buffered, this resulted in
- incomplete log files for the archive function.
- (Abe Timmerman, Fri, 9 Nov 2018 08:19:11 +0100) Merge branch
'feature/jkeenan_merges'
- (Abe Timmerman, Fri, 9 Nov 2018 08:23:05 +0100) Clean-up t/config.sh
after tests
- (Abe Timmerman, Fri, 9 Nov 2018 09:43:58 +0100) Test::Smoke::Syncer::Git
- reset working dir before branch
- We update the `patchlevel.h` file before the smoke. In some cases this
- changed file doesn't allow us to checkout the branch we want to smoke,
- so now before we start updating the working dir we `reset --hard HEAD`
- See also RT#127577
- (Abe Timmerman, Fri, 9 Nov 2018 14:16:31 +0100) Autocommit for
distribution Test::Smoke 1.72_06 (same)
- (Christian Walde, Sun, 2 Sep 2018 18:02:22 +0200) prevent git syncer
from failing on trying to log clone progress with %s
- (Christian Walde, Sun, 2 Sep 2018 19:16:45 +0200) throw an error if no
makefile is known for the specified win32 make
- (Christian Walde, Sun, 2 Sep 2018 19:17:05 +0200) ensure newlines of the
munged win32 makefile don't get mangled
- (Christian Walde, Sun, 2 Sep 2018 19:17:33 +0200) allow using gmake on
win32
- (Christian Walde, Mon, 3 Sep 2018 14:01:32 +0200) recognize win32
gnumakefile assignments in makefile preparation
- (Christian Walde, Mon, 3 Sep 2018 14:02:06 +0200) don't munge makefile
lines that remain uncommented in win32 makefile prep
- (Christian Walde, Tue, 4 Sep 2018 14:56:56 +0200) split Smoker::_make
into os subs and silence stderr with nul on win32
- (Christian Walde, Tue, 4 Sep 2018 14:57:30 +0200) switch _run from qx to
Capture::Tiny
- (Christian Walde, Tue, 4 Sep 2018 14:57:46 +0200) on windows, dump make
stdout under gcc as well
- (Christian Walde, Tue, 4 Sep 2018 14:58:12 +0200) add binmode so
change_manifest doesn't mangle the makefile too badly
- (Christian Walde, Tue, 4 Sep 2018 14:59:37 +0200) [HACK] convert
set_skip_tests to replace skipped tests with print ok 1
- this works around a shitton of manifest- and manifest-test-related
issues,
- but also breaks the unsetting. might needs to work by copying the
skipped
- tests somewhere else.
- (Christian Walde, Tue, 4 Sep 2018 15:00:45 +0200) add w32args to general
options list for runsmoke
- not sure if this is useful, was an attempt to get cchome into the