File: Known_Bugs.html

package info (click to toggle)
dar 2.6.13-2
  • links: PTS
  • area: main
  • in suites: bullseye
  • size: 10,364 kB
  • sloc: cpp: 77,385; sh: 6,192; ansic: 776; makefile: 435; python: 242; csh: 95; perl: 43; sed: 16
file content (1091 lines) | stat: -rw-r--r-- 49,447 bytes parent folder | download | duplicates (8)
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
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html; charset=ISO-8859-1"
 http-equiv="content-type">
  <title>Known Bugs</title>
</head>
<body
 style="background-color: rgb(221, 221, 221); color: rgb(0, 0, 170);"
 link="#0000ff" vlink="#000055" alink="#ff0000">
<center>
<table
 style="width: 90%; text-align: left; margin-left: auto; margin-right: auto;"
 border="0" cellspacing="2" cellpadding="2">
  <tbody>
    <tr>
      <td style="vertical-align: top; width: 161px;"><a
 href="index.html"><img
 style="border: 0px solid ; width: 160px; height: 120px;"
 alt="Dar Documentation" src="dar_s_doc.jpg"></a><br>
      </td>
      <td style="vertical-align: top;">
      <h1 style="text-align: center;"><br>
      </h1>
      <h1 style="text-align: center;">Known Bugs</h1>
      </td>
    </tr>
  </tbody>
</table>
<div style="text-align: center;">
<span style="font-family: monospace;"><span style="font-weight: bold;"><br>
<br>
</span></span><code></code>
<hr style="width: 100%; height: 2px;"><code><br>
</code>
<table
 style="width: 90%; text-align: left; margin-left: auto; margin-right: auto;"
 border="0" cellspacing="2" cellpadding="2">
  <tbody>
    <tr>
      <td style="vertical-align: top;">
      <div style="text-align: justify;"><code><span
 style="font-weight: bold;"></span></code>Here follows the description
and status of old bugs. For recent bugs please consult the <a
 href="http://sourceforge.net/tracker/?group_id=65612&amp;atid=511612">bug
tracker</a> on sourceforge.<br>
      <code><span style="font-weight: bold;"></span></code><code><span
 style="font-weight: bold;"></span></code><br>
      <code><span style="font-weight: bold;"></span></code><br>
      <code><span style="font-weight: bold;">#1:</span> </code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> files
pointed to by a symbolic links get their times (mtime &amp; atime) not
properly restored</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">problem:</span> when restoring
time of a symbolic link the
utime() system call does not change the inode of
the symbolic link itself but the one of the file pointed to by the
symbolic link</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">solution:</span> as the utime()
call behaves the same on
many Unix platforms, (HP-UX, Solaris, Linux,
etc.), the times for the symbolic links are </code><code>no more
restored, thus files
pointed to by symbolic links are not </code><code>affected by the
utime()
strange behavior. Dar, behaves not the </code><code>same as touch,
tar, etc.</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code>affects
version: 1.0.0</code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code>fixed
in version: 1.0.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#2:</span> <span
 style="text-decoration: underline;"></span></code><br>
      <code><span style="text-decoration: underline;"></span></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> dar
immediately exits with message "file mode is
neither read nor write" while
creating an archive on 64 bits platforms.</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">problem:</span> missing mask with
O_ACCMODE for flags
returned by fcntl() </code><code>system call</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
O_ACCMODE mask added</code><br>
      <code> </code><br>
      <code>affects versions: 1.0.0, 1.0.1</code><br>
      <code>fixed in version: 1.0.2</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#3: </span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> making
differential backup with some directory
implicitly or </code><code>explicitly (-P) excluded,
sometime lead to full backup.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
when entering a directory that was ignored,
the comparison in the </code><code>catalogue of reference was
not going back to the parent directory, </code><code>leading the
comparison to be
no more synchronized with the file being </code><code>read.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
added a line to go back to parent
directory when treating an </code><code>ignored directory </code><br>
      <code> </code><br>
      <code>affects versions: 1.0.0, 1.0.1</code><br>
      <code>fixed in version: 1.0.2</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#4:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> on Linux 64
bits, when an interaction from the user is </code><code>required,
(escape or return
key), the keys have to be pressed four </code><code>time before dar
can read a
character.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span> </code><code>unknown,
same code works
fine on other platforms, and if bad character </code><code>remains in
the pipe, the
message would be displayed several time, </code><code>which is not the
case, here.
It seems that the user has to press four </code><code>key for the
first one to be
accessible to dar.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
none actually</code><br>
      <code> </code><br>
      <code>affects versions: 1.0.0, 1.0.1 , 1.0.2 and probably
also 1.1.0</code><br>
      <code>fixed in version: none</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#5:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> Dar stops one
of the two following error:</code><br>
      <code>--- exception nature = [alive]&nbsp; exception type =
[BUG] ----------</code><br>
      <code>[source]</code><br>
      <code>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; file
compressor.cpp line 244 : it seems to be a bug here</code><br>
      <code>[most outside call]</code><br>
      <code>&nbsp;-----------------------------------</code><br>
      <code> </code><br>
      <code>or</code><br>
      <code> </code><br>
      <code>--- exception nature = [alive]&nbsp; exception type =
[BUG] ----------</code><br>
      <code>[source]</code><br>
      <code>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; file
compressor.cpp line 254 : it seems to be a bug here</code><br>
      <code>[most outside call]</code><br>
      <code>-----------------------------------</code><br>
      <code> </code><br>
      <code style="text-decoration: underline;">problem:</code><code>
the zlib deflate() calls
returned an unexpected value (Z_STREAM_END) </code><code>and inflate()
in some rare
cases was called without data to process, </code><code>when reaching
end of file,
which made inflate() to return Z_BUF_ERROR </code><code>code.</code><br>
      <code> </code><br>
      <code style="text-decoration: underline;">solution:</code><code>
adding code for taking care
of the returned code Z_STREAM_END by </code><code>deflate(), and code
for
taking care of Z_BUF_ERROR returned at end </code><code>of file.</code><br>
      <code> </code><br>
      <code>affects version: 1.0.0, 1.0.1, 1.0.2</code><br>
      <code>fixed in version: 1.0.3</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#6:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> Dar does not
properly display dates on Linux 64 bits </code><code>systems.</code><br>
      <code> </code><br>
      <code style="text-decoration: underline;">problem:</code><code>
non portable cast from
'time_t' to 'unsigned long int'</code><br>
      <code> </code><br>
      <code style="text-decoration: underline;">solution:</code><code>
changed code not to have to
use 'unsigned long int'</code><br>
      <code> </code><br>
      <code>affects version: 1.0.x, 1.1.x</code><br>
      <code>fixed in version: 1.2.0</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#7:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> while saving,
Dar stops when a given inode could not
be read, </code><code>even if the inode had not to
be saved</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span> </code><code>Dar
reads all file in the
directory to save, and then applies filters </code><code>to their
filename. Error
appears before the filter, leading to a </code><code>global saving
error.</code><br>
      <code> </code><br>
      <code style="text-decoration: underline;">solution:</code><code>
read error are now caught
sooner a message is displayed that the file </code><code>or directory
will be
ignored, and saving process continues. As </code><code>this error
occurs before
testing filename against filters, an error </code><code>message is
still displayed
(which may seems strange when the file was </code><code>about to be
excluded by
filter mechanism) <br>
      </code><code></code><br>
      <code>Just reminds that dar reads a directory contents first, and
      </code><code>then check filenames against
filter. At the time of the message, Dar </code><code>does not know the
file would
not to be saved. Thus, don't worry if </code><code>you see this kind
of
message, they don't mean that dar was about to </code><code>save the
given file but
rather exactly what the message says, that </code><code>the file or
directory could
not be read. ;-)</code><br>
      <code> </code><br>
      <code>affects version: 1.0.x, 1.1.x</code><br>
      <code>fixed in version: 1.2.0</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#8:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> when saving
hard linked file using filter that does
not save </code><code>the first hard link, dar
stops with SEGFAULT.</code><br>
      <code> </code><br>
      <code style="text-decoration: underline;">problem:</code><code>
The first hard link is used
as reference for the following occurrences. </code><code>As it is not
covered by the
filter it is destroyed in filter module, </code><code>but filesystem
module keeps
using it as reference for following </code><code>hard links on that
inode. </code><br>
      <code> </code><br>
      <code style="text-decoration: underline;">solution:</code><code>
added a call in filesystem
module to "forget" a given reference. This </code><code>way, if a
file_etiquette is
not covered, filter module can inform </code><code>filesystem module
to delete
the corresponding entry on the hard link </code><code>map. Subsequent
hard link on
that inode will generate first a new </code><code>file_etiquette,
which will
be used as reference (unless it is not </code><code>covered by filter
too). </code><br>
      <code> </code><br>
      <code>affects version: 1.1.x</code><br>
      <code>fixed in version: 1.2.0</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#9:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> </code><code>failed
to read an archive
under windows NT and 2000</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span> </code><code>error
concerning the integer
macro definition (module integer.hpp)</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span> </code><code>replaced
the macro by a
proper definition.</code><code></code><br>
      <code> </code><br>
      <code>affects version: 1.2.0</code><br>
      <code>fixed in version: 1.2.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#10:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> </code><code>when
using dar on windows95
if giving c:/ or other driver to -R </code><code>option, dar stops
with the
following error:</code><br>
      <code> </code><br>
      <code>file filesystem.cpp line 186
: it seems to be a bug here</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span> </code><code>dar
checks first that c:/ is
a directory. Cygwin says yes this </code><code>dar continues : from
c:/ dar
completes the path with a dot thus </code><code>dar tries to open the
path
c:/. which is Cygwin does not report </code><code>as an existing
directory. </code><code><br>
      <br>
      </code><code><span style="text-decoration: underline;">solution:</span>
      </code><code>dar does not try to open
c:/. but tries to open c:/ . Moreover </code><code>additional code as
been
added to detect the incoherence if in </code><code>some other
conditions such
problem occurs. Dar will report the </code><code>exact problem (path
does not
exist / or path is not a directory)</code><br>
      <code></code><br>
      <code>affects version: 1.2.1, 1.2.0</code><br>
      <code>fixed in version: 1.3.0</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#11:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> when giving to
dar_manager
an absolute path to -r option dar_manager stops with
INTERNAL ERROR message.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span> a
check is missing to
display error message in such condition</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
added a error message, if an
absolute path is given to -r</code><br>
      <code> </code><br>
      <code>affects dar_manager version&nbsp;&nbsp; : 1.0.0</code><br>
      <code>fixed in dar_manager version&nbsp; : 1.0.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#12:</span></code><br
 style="text-decoration: underline;">
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> memory leakage
when using -Z
or -Y option. This bug has been found and explained by David
Rose, who proposed a good patch, which I have just a little
adapted.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
the dar code that call the
zlib when changing the compression algorithm did properly
released memory allocated by dar but forget to execute a call of
the zlib library that releases memory allocate by the library
itself. </code><br>
      <code></code><br>
      <code><span style="text-decoration: underline;">solution:</span>
solution, add call to
release memory used by zlib library when changing of compression
algorithm</code><br>
      <code> </code><br>
      <code>affects dar version&nbsp;&nbsp; : 1.2.0, 1.2.1</code><br>
      <code>fixed in dar version&nbsp; : 1.3.0</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#13:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> archive
created by dar under
Windows NP SP6, XP, 2000, cannot be read, tested. All action on
generated archive finishes by CRC error message, incoherent
catalogue structure, or badly formated infinint.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
under windows, Dar uses
Cygwin. As Unix and Window do not define end of line the same way (\n
for Unix and \r\n for Windows), Cygwin translates end of
line on the fly when reading a file (Windows to Unix), and when
writing (Unix to Windows). At writing time especially,
each \n anywhere in the archive is written \r\n, thus two
bytes get written in place of one. In another hand, to get the
global position in the whole archive dar counts bytes it writes
to file. Thus a shift occurs between the position it thinks it
writes the catalogue and the position it actually does. Thus
reading the archive, the given position is wrong, and archive cannot
be opened.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
deactivate the translation
when generating the archive. Cygwin adds a two special flags to
pass to open(2) which are O_TEXT and O_BINARY, that change the
way Cygwin handles file read(2) and write(2).</code><br>
      <code> </code><br>
      <code>affects all dar suite program version 1.0.x, 1.1.x,
1.2.x</code><br>
      <code>fixed in dar version 1.3.0</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#14:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> Dar aborts
with the
following message</code><br>
      <code> </code><br>
      <div style="margin-left: 40px;"><code>---- exception nature
= [alive] exception type = [BUG] ----------</code><br>
      <code>[source]</code><br>
      <code>&nbsp; &nbsp; file compressor.cpp
line 320 : it seems to be a bug here</code><br>
      <code>[most outside call]</code><br>
      <code>-----------------------------------</code><br>
      </div>
      <code> </code><br>
      <code>It appears while doing a
backup using compression and scrambling (could not (yet) be reproduced
twice).</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
this message is issued
because zlib library reported an inconsistency in call order or in call
arguments given to zlib calls. Actually I cannot figure out what is
wrong, nor how this bug may happens.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
none actually at the time of
2.0.0 release.</code><br>
      <code> </code><br>
      <code>But, if you are face to face
with this bug, thanks to try reproducing it, re-issuing the same
command in the same condition,&nbsp; and if possible to reproduce then
as much as possible reduce the necessary environment condition that
make this bug happen. Then please, contact me (either via Sourceforge
or via email).</code><br>
      <code> </code><br>
      <code>affects version 1.2.1 (at least)</code><br>
      <code>not yet fixed (check Sourceforge for fresh news),
nor reproduced.</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#15:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> Dar aborts
with the following
message when giving -y0 as argument</code><br>
      <code> </code><br>
      <div style="margin-left: 40px;"><code>---- exception nature =
[alive]&nbsp; exception type = [BUG] ----------</code><br>
      <code>[source]</code><br>
      <code>&nbsp;&nbsp;&nbsp;
file compressor.cpp line 96 : it seems to be a bug here</code><br>
      <code>[most outside call]</code><br>
      <code>-----------------------------------</code><br>
      </div>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
correct compression level for zlib
are from 0 to 9, 0 meaning no compression, while for libbz2
correct compression levels are from 1 to 9, thus zero is not a valid
value. libbz2 returns an error code which triggers a libdar internal
sanity check code, which throws this BUG exception.</code><br>
      <code> </code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
forbid the use of -z0 or -y0. If
you don't want compression, you will need to not give neither -z nor
-y. </code><br>
      <code> </code><br>
      <code>affects version 1.3.0</code><br>
      <code>fixed in version 2.0.0</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#16:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> Dar include
does not ignore
comments in included file as given by -B option</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
when a line with only one
carriage-return is put in a file, the following line is considered as a
normal line (not a comment), whatever it is, in particular if it starts
with a hash (#).</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
make the algorithm consider normal
line once another character than space, tab hash or carriage-return
is met. (added the carriage-return in the list).</code><br>
      <code> </code><br>
      <code>affects version 1.2.0, 1.2.1, 1.3.0</code><br>
      <code>fixed in version 2.0.0</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#17:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> Dar warns that
the archive is about to
save itself while a -P option is used to prevent this eventuality.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
testing against this situation did not
check that a parent directory of the path where is saved the archive is
excluded by a -P option. </code><br>
      <code></code><br>
      <code><span style="text-decoration: underline;">solution:</span>
updating the test for the warning only
appear in real auto-saving conditions.</code><br>
      <code> </code><br>
      <code>affects version 2.0.0</code><br>
      <code>fixed in version 2.0.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#17bis:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> save as #17,
but in the case where the
archive is in the under the current directory, and some -P options
exclude it.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
the current directory "." is not
substitued by its full path, and the testing algorithm does tread
"." as directly under the root "/" directory, no -P could thus
exclude it from the files to save.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
for testing and preventing
self-backup, "." is replaced by its full path counterpart. affects
version 2.0.0, 2.0.1, fixed in version 2.0.2</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#18:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> Dar does not
save files or just stores
empty directories when using -B option or having a ~/.darrc or
/etc/darrc file</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
the parsing code that strip the comments
from files is not counting empty lines ('\n')
between two lines of commands (i.e.: uncommented lines). This strips
the last uncommented line before the next line of comment, usually
leading to
the&nbsp; removal of an ending '\n' in the uncommented line, implying
the
fusion of the last command on this line and the next command on any
uncommented after the last line of comment two commands that were
initially
separated by a '\n' and one or several&nbsp; lines of comments. This
wrongly interpreted
command most of the time appear as a member of [list of path] and
thus restricts&nbsp; the backup to only a subdirectory that most of the
time does not exist. If -D option is used, all first level directory
are excluded from backup and thus stored as empty directories.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
bug fix in no_comment.cpp. A workaround
is to remove comments or empty lines from configuration files. </code><br>
      <code></code><br>
      <code>affects versions 2.0.0, 2.0.1</code><br>
      <code>fixed in versions 2.0.2</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#19:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> Dar hangs when
using --nodump with named
pipe to save</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
to get the nodump flag value, dar needs
to call the open() system call in read-only mode. For named pipes
this&nbsp; system call does return only when another process opens the
same pipe in
write mode.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
as proposed by Joshua Neal, the open
will be set to non blocking mode to return immediately.</code><br>
      <code> </code><br>
      <code>affects version 1.3.0, 2.0.0, 2.0.1</code><br>
      <code>fixed in version 2.0.2</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#20:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> dar_manager
did not completely removed
reference to an archive when it was removed from a database</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
data_dir is a C++ class that inherit
from data_tree. This one has a virtual method remove_all_from(), that
had was redefined in the data_dir class with a very little difference,
making it not be a redefinition of the inherited class. (argument was
"archive_num" in place of "const &amp; archive_num"). This redefined
method was never called, and the removal never recursed in
subdirectories.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
make the two signature identical in the
mother and inherited classes.</code><br>
      <code> </code><br>
      <code>affects version 1.3.0, 2.0.0, 2.0.1, 2.0.2</code><br>
      <code>fixed in version 2.0.3</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#21:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> under FreeBSD
using -B option leads to
parse error</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
under FreeBSD, getopt() call is not
reset just by setting optind to zero like under Linux, but the variable
optreset must be set to 1.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
configure script now check for the
presence of optreset variable which allows to properly reset the
getopt call according to the system nature.</code><br>
      <code> </code><br>
      <code>affects version 2.1.0 and older</code><br>
      <code>fixed in version 2.1.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#22:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> dar stops when
fed with an empty file
for -B option</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
the comments stripping class lack a
variable initialization</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
initialize variable before use</code><br>
      <code> </code><br>
      <code>affects version 2.1.0 and older</code><br>
      <code>fixed in version 2.1.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#23:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> -z -y -H and
-w do not work under FreeBSD</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
the getopt() version does not support
optional arguments</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
document this problem, and make possible
to reach any feature</code><br>
      <code> </code><br>
      <code>in the meanwhile, replace</code><br>
      <div style="margin-left: 40px;"><code>-z&nbsp;&nbsp;&nbsp;
by&nbsp;&nbsp; -z 9</code><br>
      <code>-y&nbsp;&nbsp;&nbsp; by&nbsp;&nbsp; -y 9</code><br>
      <code>-H&nbsp;&nbsp;&nbsp; by&nbsp;&nbsp; -H 1</code><br>
      <code>-w&nbsp;&nbsp;&nbsp; by&nbsp;&nbsp; -w d (added for the
circumstances)</code><br>
      </div>
      <code> </code><br>
      <code>affects all versions</code><br>
      <code>documented starting version 2.1.2 (see dar's man
page)</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#24:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> under windows
2000 SP4 failed to
properly save a text file with DOS carriage return. Testing the archive
reports CRC error.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
The cygwin adaptation layer for dar to
be able to run under windows, needs non standard flag mode to not
automatically translate between windows and unix carriage return. This
bug is related to bug #13 which fix was not properly applied in
one single place. </code><br>
      <code></code><br>
      <code><span style="text-decoration: underline;">solution:</span>
properly apply the fix of bug #13 in
this single place too.</code><br>
      <code> </code><br>
      <code>affects version 2.1.1 and older</code><br>
      <code>fixed in version 2.1.2</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#25:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> under some
hypothetical conditions (when
the archive is larger than 10^16 Terabytes, 1 followed by 16
zeros), dar is not able to read the archive.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
In this situation, the terminateur is
not properly formed and cannot fulfill its role: point to the start of
the catalogue. A byte 0x00 is missing near the end of the archive,
that indicate the end of the terminateur.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
properly write down terminateur string
of archive. Note that archive produced by earlier version are not
possible to read dar version that have the fix, but don't worry, if
you have a such big archive (more than 10^16 Terabytes), tell me I will
create a little program that fixes the archive ;-) . Note,
"terminateur" the french word for terminator, the word has been kept
here
because it is used in the source code.</code><br>
      <code> </code><br>
      <code>affects version 2.1.2 and older</code><br>
      <code>fixed in version 2.1.3</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#26:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> in
configuration file using conditional
syntax (see man page), when two targets follow each other on two
directly following lines (the first condition is thus empty), the
second condition is skipped. This condition also occurs if there is
only
comments between the two targets.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
when looking for the first target, an
extra character is read (the first of the second target), as there is
no
contents in this target no reposition (seek) in the file is done,
and the search for the next target start at that point (second
character of
the second target), which makes the second target not recognized
because
of the missing initial character.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
avoid reading this extra character so
the next character to read is the first character (not the second) of
the
following target.</code><br>
      <code> </code><br>
      <code>affects version 1.2.0 and older versions</code><br>
      <code>fixed in version 2.1.3</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#27:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> when restoring
only more recent files
(-r), with hourshift activated (-H) the -H feature was randomly not
activated.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span> a
test comparing the dates is made on
address's objects rather on objects themselves. This leads to the -H
test
being activated randomly depending on address of two objects to compare.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
make the test on objects rather than on
their addresses in memory.</code><br>
      <code> </code><br>
      <code>affects version from 2.0.0 to 2.1.3</code><br>
      <code>fixed in version 2.1.4</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#28:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> dar does not
restore root Extended
Attributes</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span> a
collision occurred in the naming of
the enumeration listing the extended attributes available domains,
and the variables carrying user choice to consider or not user or
root extended attributes for the operation (backup, restoration,
comparison, etc.)</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
rename both the domain enumeration and
the variables carrying user choice about extended attribute
comportment.</code><br>
      <code> </code><br>
      <code>affects version from 2.1.0 to 2.1.3</code><br>
      <code>fixed in version 2.1.4</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#29:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> dar does
Segmentation fault during
backup when hard links are involved</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
when an plain file inode linked several
time to the directory tree (a hard linked file), is not readable
due to a lack a permission, the filesystem module generates a C++
object and stores its address for further reference and hard link
handling. When the file opening comes an exception is throw, leading to
this object destruction, but the reference stays in the the filesystem
module, leading the next link to this inode being referenced by a
destroyed object.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
remove the reference when exception is
caught. This removal was properly done when such an inode was not to be
saved (and the corresponding C++ object was destroyed), the case of the
exception had been forgotten.</code><br>
      <code> </code><br>
      <code>affects version up to 2.1.4</code><br>
      <code>fixed in version 2.1.5</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#30:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> compilation
failed with syntax error
reported in libstdc++ header files (reported on Solaris and openBSD).</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
the configure scripts does detect the
presence of mbstate_t type, and redefines it. This conflicts the system
header files and lead to a syntax error.</code><br>
      <code> </code><br>
      <code>workaround 1:</code><br>
      <code></code></div>
      <div style="margin-left: 80px; text-align: justify;"><code>will
update the configure script to
check if that fixes the problem. Else there is a workaround. Once
configure
has finished, edit the config.h file adding the line</code><br>
      </div>
      <div style="margin-left: 120px; text-align: justify;"><code>#define
HAVE_MBSTATE_T 1</code><br>
      </div>
      <div style="margin-left: 80px; text-align: justify;"><code>and
commenting out the line</code><br>
      <div style="margin-left: 40px;"><code>#define mbstate_t int</code><br>
      </div>
      <code>Thanks to Steve Evans for this
workaround.</code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
undefine the mbstate_t in dar's source
code. This is implemented and seems to work. No action is required for
your part.</code><br>
      <code> </code><br>
      <code>may affects version up to 2.2.0 </code><br>
      <code>fixed in version 2.2.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code> </code><br>
      <code><span style="font-weight: bold;">#31:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> dar segfault
when restoring Extended
Attributes</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
to speed up the reading of the catalogue
from an archive an intermediate layer has been added thanks to an
"cache"&nbsp; object. This object is temporary instead of the object it
does caching
on. When the catalogue reading has finished, the reference given to
inode
becomes invalid because it is the caching object in place of the real
underlying archive object.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
keep doing archive reading through the
caching object but give the real archive object to the inode contained
in
the catalogue for they have a valid reference when it is time to
restore EA.</code><br>
      <code> </code><br>
      <code>affects version 2.2.0</code><br>
      <code>fixed in version 2.2.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#32:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> dar is unable
to read blowfish encrypted
archives under certain conditions</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
this bug appears when using infinint (no
--enable-mode in command-line). The implicit cast from a unsigned
integer to an infinint in a operation was not properly made by the
compiler.
Instead of converting the system integer to an infinint which is
possible
thanks to the infinint constructor class, the compiler down cast the
the
unsigned integer to unsigned char.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
use explicit cast conversion when doing
operations which concern both infinint and unsigned integers.</code><br>
      <code> </code><br>
      <code>affects version 2.2.0</code><br>
      <code>fixed in version 2.2.1</code><br>
      <code></code></div>
      <code> <br>
      </code>
      <div style="text-align: justify;"><code><span
 style="font-weight: bold;">#33:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">decription:</span> dar hangs when
the last line of a DCF file is
starting with a few spaces or tabs just following a comment.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
The layer that removes comments was
counting the bytes from the beginning while the first byte retained as
significative was the first non space of the line. leading to provide
to upper
layer a wrong information about the number of byte in the file.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
start counting bytes from the first
significative byte of the line instead of the first byte of the line.</code><br>
      <code> </code><br>
      <code>affects versions 2.1.x up to 2.2.0</code><br>
      <code>fixed in version 2.2.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#34:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> when restoring
files in a directory, this one
has all permission set for every body. This makes a security hole on
multi-user system.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
when restoring in a directory as non
privileged user it is necessary to have write access to that directory,
thus temporary change the permission. The permission is thus changed,
but the
temporary change is too permissive, thus opens a security hole,
allowing local
user delete/create/replace under the directory restored. </code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
only add the user write access (if necessary)
to the directory when dar is run as non privileged user.</code><br>
      <code> </code><br>
      <code>affects versions 1.x.x up to 2.2.0</code><br>
      <code>fixed in version 2.2.1</code><br>
      <code></code></div>
      <div style="text-align: justify;"><code> </code><br>
      <code><span style="font-weight: bold;">#35:</span></code><br>
      <code></code></div>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> when producing
archive to stdout sometimes the
message "Error while saving data: Bad file descriptor
given" is displayed and dar aborts its execution</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">problem:</span>
health check done on a uninitialized object
field in place of a constructor's argument that will be used to
set this field.</code><br>
      <code> </code><br>
      <code><span style="text-decoration: underline;">solution:</span>
test the method's argument in place of the
object field</code><br>
      <code> </code><br>
      <code>affects version 2.2.x up to 2.2.3</code><br>
      <code>fixed in version 2.2.4</code> <br>
      <br>
      </div>
      <code><span style="font-weight: bold;">#36: </span></code>
      <div>
      <div style="margin-left: 40px;"><code><span
 style="text-decoration: underline;">description:</span>
when using -I or -X with -l option directory names are also filtered
while they should not be, as described in the man page</code><br>
      <code> </code><br>
      <code> <span style="text-decoration: underline;">solution:</span>
avoid
having -X/-I filtering directories</code><br>
      <code> </code><br>
      <code>affected versions 2.2.x up to 2.2.5</code><br>
      <code>fixed in 2.2.6</code><br>
      </div>
      <code><br>
      <span style="font-weight: bold;">#37:</span><br>
      </code>
      <div style="margin-left: 40px; text-align: justify;"><code><span
 style="text-decoration: underline;">description:</span> when restoring
a file which all EA has been dropped since the archive of reference,
dar reports CRC error.</code><br>
      <code></code><br>
      <code><span style="text-decoration: underline;">problem:</span>
to store that EA has been dropped, dar saves an empty EA list, but did
not calculated the CRC value</code><br>
      <code></code><br>
      <code><span style="text-decoration: underline;">solution:</span>
make the CRC value set when storing an empty list of EA. Note that this
bug did not cause any problem except a CRC error message.</code><br>
      <code></code><br>
      <code>affects versions up to 2.2.6</code><br>
      <code>fixed in 2.2.7</code><br>
      </div>
      <br>
      </div>
      </td>
    </tr>
  </tbody>
</table>
<br>
</div>
<br>
<br>
</center>
</body>
</html>