File: Notes.html

package info (click to toggle)
dar 2.8.1-3
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid
  • size: 12,024 kB
  • sloc: cpp: 86,219; sh: 6,978; ansic: 895; makefile: 489; python: 242; csh: 115; perl: 43; sed: 16
file content (3484 lines) | stat: -rw-r--r-- 129,686 bytes parent folder | download 
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
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
 
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
  <head>
    <link href="style.css" rel="stylesheet">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <meta content="text/html; charset=ISO-8859-1" http-equiv="content-type">
    <title>DAR/LIBDAR Internals - Notes</title>
  </head>
  <body>

    <a name="top">&nbsp;</a>
    <div class=top>
      <img alt="Dar Documentation" src="dar_s_doc.jpg" style="float:left;">
      <h1>Dar/Libdar Internals - Notes</h1>
    </div>

    <div class=jump>
      <div class=menuitem>
	<a href="#top">Back to top</a>
      </div>
    </div>

    <div class=page>
      <div class=menuleft>
	<div class=menuitem>
	  <a href="#intro">Introduction</a>
	</div>
	<div class=menuitem>
	  <a href="#EA_and_diff">EA &amp; differential backup</a>
	</div>
	<div class=menuitem>
          <a href="#archive_structure">Archive structure in brief</a>
	</div>
	<div class=menuitem>
	  <a href="#overflow">Overflow in arithmetic integer operations</a>
	</div>
	<div class=menuitem>
	  <a href="#thread_safe">libdar and thread-safe requirement</a>
	</div>
	<div class=menuitem>
	  <a href="#NLS">Native Language Support</a>
	</div>
	<div class=menuitem>
	  <a href="#Dar_relase_Process_in_brief">Dar Release Process</a>
	</div>
	<div class=menuitem>
	  <a href="#Dar_version_naming">Dar's Versions</a>
	</div>
	<div class=menuitem>
          <a href="#scrambling">Scrambling (weak encryption)</a>
	</div>
	<div class=menuitem>
	  <a href="#strong_encryption">Strong encryption internals</a>
	</div>
	<div class=menuitem>
          <a href="#asym">Public key based encryption internals</a>
	</div>
	<div class=menuitem>
	  <a href="#multi-threaded">Multi-threading in libdar</a>
	</div>
	<div class=menuitem>
	  <a href="#block_compression_mode">Streaming versus block Compression mode</a>
	</div>

      </div>

      <div class=mainleft>

	<h2><a name="intro">Introduction</a></h2>
	<p>
	  Here takes place a collection of notes. These have been created after
	  implementation of a given feature, mainly for further
	  reference but also as user information. The idea behind
	  these notes are to remind some choices of implementation, the
	  arguments and reasons that lead to them, but also
	  let the user have a way to be informed about the
	  choices done and be able to bring his remarks without having
	  to deeply look into the code to learn dar's internals.
	</p>

	<h2><a name="EA_and_diff">EA &amp; differential backup</a></h2>

	<h3>Brief presentation of EA:</h3>

	<p>
	  EA stands for <i>Extended Attributes</i>.
	  In Unix filesystem a regular file is composed of a
          set of bytes (the data) and an <i>inode</i> (containing the <i>metadata</i>).
	  The inode add properties to the file, such as owner, group, permission, dates (last
          modification date of the data [mtime], last access date to
          data [atime], and last inode change date [ctime]), etc). Last,
          the name of the file is not contained in the inode, but in the
          directory(ies) it is linked to. When a file is linked more
          than once in the directory tree, we speak about "hard links".
          This way the same data and associated inode appears several
          time in the same or different directories. This is not the
          same as a symbolic links, which later one is a file that contains the
          path to another file (which may or may not exist). A symbolic
          link has its own inode. OK, now let's talk about EA:
	</p>

	<p>
	  Extended attributes is a recent feature of Unix file system (at the time
          of the writing, year 2002). They are not part of the
          inode, nor part of the data, nor part of a given directory. They
          are stored beside the inode and are a set of pair of <i>key</i> and
          <i>value</i>. The owner of the file can add or define any key and
          eventually associate it an arbitrary data value. The user has also
	  the means to list and remove a particular key for the EA.
	  What are they used for: Simply a more flexible way to associate information to a file.
	</p>
	<p>
          One particular interesting use of EA, is ACL: <i>Access Control
            List</i>. ACL are implemented using EA (Linux) and add a more fine grain
          in assigning access permission to file. For more information one
          EA and ACL, see the site of Andreas Grunbacher:
	</p>

	<p>
          <a href="http://acl.bestbits.at/">http://acl.bestbits.at/</a>
	</p>

	<p>
	  File forks under MACOS X also relies on EA, as well the security features
	  brought by SELinux.
	</p>


	<h3>EA &amp; Differential Backup</h3>
	<p>
	  to determine
          that an EA has changed, <i>dar</i> looks at the <i>ctime</i> value of the
	  inode. if <i>ctime</i>
          has changed, (due to EA change, but also due to permission or owner
          change) dar saves the EA. ctime also changes, if atime or mtime
          changes. So if you access a file or modify it, dar will consider
          that the EA have changed also. This is not really fair, I admit.
	</p>
	<p>
          It may sound better would be to compare EA one by one, and record
          those that have changed or have been deleted. But to be able to do
          so, all EA and their associated values  must reside in the archive
	  table of content (the <i>catalogue</i>), which is by design stored
	  in memory. As EA can grow up to 64 KB by file, this can lead to a
          quick saturation of the virtual memory, which is already enough
          solicited by other information from <i>catalogue</i>.
	</p>
	<p>
          These two schemes implies a different pattern for saving EA in
          archive. In the first case (no EA in memory except at time of
          operation on it), to avoid skipping in the archive (and ask the
          user to change of disks too often, or just put pressure on the cache and
	  disk I/O), EA must be stored beside the
          data of the file (if present). Thus they must be distributed all
          along the archive (except at the end that only contains the
          catalogue).
	</p>
	<p>
          In the second case (EA are loaded in memory for comparison), EA
          must reside beside or within the catalogue, in any case at the
          end of the archive, not to have to user to need all disks to
          just take an archive as reference.
	</p>
	<p>
          As the catalogue grows already fast with the number of file to
          save (from a few bytes for hard_link to 400 bytes around per
          directory inode), the memory saving option has been adopted.
	</p>
	<p>
          Thus, EA changes are based on the ctime change. Unfortunately,
          no system call permits to restore ctime. Thus, restoring a
          differential backup after its reference has been restored, will
          present restored inode as more recent than those in the
          differential archive, thus the <code>-r</code> option would prevent any EA
          restoration. In consequence, <code>-r</code> has been disabled for EA, it
          does only concern data contents. If you don't want to restore
          any EA but just more recent data, you can use the following :
	  <code> -r -u "*"</code>
	</p>

	<h2><a name="archive_structure">Archive structure in brief</a></h2>

	<h3>The Slice Level</h3>
	<p>
	  A slice is composed of a header,
          data and trailer (the trailer appeared with archive format
          version 8)
	</p>
	<code class=block>
	  +--------+-------------------------------------------+-------+
	  | header |  Data                                     |Trailer|
	  |        |                                           |       |
	  +--------+-------------------------------------------+-------+
	</code>

	<p>
          the <b>slice header</b> is composed of:
	</p>

	<ul>
          <li>a magic number that tells this is a dar slice</li>
          <li>
	    a internal_name which is unique to a given archive and
            shared by all slices
          </li>
          <li>
	    a flag that tells whether the slice is the last of the
            archive or whether a trailer is present that contains this
            info.
          </li>
          <li>
	    a extension flag, that was used in older archive but which
            now always set to 'T' telling that a TLV list follows
	  </li>
          <li>
	    A TLV (Type Length Value) list of item, it contains:
	  </li>
          <ul>
            <li>the slice size</li>
            <li>first slice size</li>
            <li>the data_name</li>
          </ul>
	</ul>

	<p>
	  The TLV list will receive any
          future new field related to slice header.
	</p>
	<code class=block>
	  +-------+----------+------+-----------+-------+
	  | Magic | internal | flag | extension | TLV   |
	  | Num.  | name     | byte | byte      | list  |
	  +-------+----------+------+-----------+-------+
	</code>
	<p>
          The header is the first thing
          to be written, and if the current slice is not the last slice
          (all data to write could not fit in it), before format 8, the
          flag field was changed indicating that another slice follows.
          Since archive format 8, the flag is set to a specific value
          indicating that the information telling whether the slice is
          the last or not is placed in a slice trailer, a new
          "structure" that appeared with that format and which is
          located at the end of each slice.
	</p>
	<p>
          The header is also the first part to be read.
	</p>
	<p>
          A TLV list is of course a list of TLV:
	</p>
	<code class=block>
	  +-------+----------+------+-----------+- ...-----+-------+
	  | Number| TLV 1    | TLV 2| TLV 3     |          | TLV n |
	  | of TLV|          |      |           |          |       |
	  +-------+----------+------+-----------+--...-----+-------+
	</code>
	<p>
          Each TLV item is, as commonly,
          defined as set of three fields:
	</p>
	<code class=block>
	  +---------+-------------------------+-------------------------+
	  | Type    | Length                  | Value                   |
	  |(2 bytes)| (arbitrary large value) | (arbitrary large data)  |
	  +---------+-------------------------+-------------------------+
	</code>

	<p>
          The 2 bytes type is large
          enough for today's need (65535 different types while only
          three used), however TLV 65535 is reserved for future use and
          will signal a new format for the type field.
	</p>

	<p>
	  To know in
          which slice and at which position to find a particular data, dar
          needs to know each file's size. This is the reason why each
          slice contains the slice size information, in particular the
          last slice. In older version, dar had to read the first slice
          first to get this slicing information. Then it could read the
          archive contents at the end of the last slice. Today, reading
          the last slice, dar can fetch the slicing scheme from the slice
          header (what we just detailed) and fetch the archive contents at
          the end of this same last slice.
	</p>
	<p>
          <b>The trailer</b> (which is one byte length) is new since archive
          format version 8 (released 2.4.0). It contains the value that
          was located in the header flag field in older archive format,
          telling whether the slice is the last of the archive or not.
          When writting down a single sliced archive (no -s option
          provided), both the header and the trailer tell that the slice
          is the last of the archive (duplicated information). However,
          when doing multi-sliced archive, it is not possible to known
          whether a slice is the last before reaching the requested amount
          of data per slice (which depends on the amount of byte to save,
          compression ratio, encryption overhead, etc.). Thus the header
          flag contains a value telling that to know whether the slice is
          the last or not, one must read the trailer.
	</p>
	<p>
          In older format, it was necessary to seek back to update the
          header with the correct information when a new slice had to be
          created. But, keeping this behavior, it would not have been
          possible to make a digest "on the fly" (see --hash option). The
          addition of the trailer was required for that feature: to
          compute a md5 or sha1, ... hash for each slice. But, this costs
          one byte per slice, yes.
	</p>

	<h3>Data Name</h3>
	<p>
	  As seen above in the header
          fields, we have among others the three following identifiers:
	</p>
	<ul>
          <li>magic number</li>
          <li>internal name</li>
          <li>data name</li>
	</ul>

	<p>
	  As already said, magic number
          is constant and let dar be (almost) sure a given file is a dar
          slice file, this is also based in particular on that field
          that the common unix 'file' command identifies an dar archive.
          Also briefly explained, the internal_name is a identifier that
          let dar be almost sure that several slices are from the same
          archive (problem car arise if two archives of same basename
          have their slices mixed together: dar will see that and report
          it to the user).
	</p>
	<p>
          The new and not yet described field is the "data_name". The
          data_name field is also present in the archive catalogue (the
          table of content) of each archive. It may be the same value as
          the one in the slice headers (normal archives) or another
          value if the archive results from a catalogue isolation
          process.
	</p>
	<p>
          Why this field? A new feature with release 2.4.0 is the
          ability to use an extracted catalogue to backup a internal
          catalogue of a given archive. Comparing the data_name value of
          the catalogue resulting from the isolation operation to the
          data_name value present int the slices of an archive to
          rescue, dar can be (almost) sure that the extracted catalogue
          matches the data present in the archive the user is trying to
          use it with.
	</p>
	<p>
          In brief:
	</p>
	<div class=table>
	  <table class=center>
            <tr>
              <th>Fields</th>
	      <th>Normal Archive</th>
              <th>Resliced Using dar_xform</th>
              <th>Resulting From Isolation</th>
	      <th>isolated archive resliced with dar_xform</th>
	    </tr>
            <tr>
              <th>Internal_name (slice header)</th>
              <td>A</td>
              <td>B</td>
              <td>C</td>
              <td>D</td>
            </tr>
            <tr>
              <th>data_name (slice header)</th>
              <td>A</td>
              <td>A</td>
              <td>C</td>
              <td>C</td>
            </tr>
            <tr>
              <th>data_name (archive catalogue)</th>
              <td>A</td>
              <td>A</td>
              <td>A</td>
              <td>A</td>
            </tr>
	  </table>
	</div>

	<h3>Archive Level</h3>
	<p>
	  The archive level describes
          the structure of the slice's data field (removing header and
          trailer of each slice), when they are all sticked together
          from slice to slice:
	</p>
	<code class=block>
	  +---------+----------------------------+-----------+--------+---------+--------+
	  | version |   Data                     | catalogue | term 1 | version | term 2 |
	  | header  |                            |           |        | trailer |        |
	  +---------+----------------------------+-----------+--------+---------+--------+
	</code>

	<h4>version header</h4>
	<p>
          The <b>version header</b> is an almost diplication of <i>version trailer</i>. It is
          used when reading an archive in sequential mode, to be able to
          prepare the proper compression layer, and known whether escape
          sequence mark are present in the archive. It is present by default but absent
	  when removing tape marks (<code>-at</code> option).
	</p>

	<h4>version trailer</h4>
	<p>
          the <b>version trailer</b> (which may still be called "version header" in
          some part of the documentation because it was originally located at
          the beginning of the archive in previous archive format) is
          composed of:
	</p>

	<ul>
          <li>edition version of the archive</li>
          <li>compression algorithm used</li>
          <li>
	    command line used for creating the archive, now known as
            "user comment"
          </li>
          <li>flag, telling:
            <ul>
              <li>whether the archive is encrypted,</li>
              <li>whether it has escape sequence marks,</li>
              <li>whether the header/trailer contains an encrypted key</li>
              <li>whether the header/trailer contains the initial offset field</li>
              <li>whether the archive is signed</li>
              <li>
		whether the header/trailer contains the slice
		layout of the archive of reference
	      </li>
              <li>
		whether the archive has salt+iteration count+hash algo for
		key derivation fonction (KDF)
              </li>
            </ul>
          <li>
	    initial offset (telling where starts the data in the
            archive, is only present in the trailer)
	  </li>
          <li>
	    crypto algorithm used (present only if flag tells that the
            archive is encrypted)
	  </li>
          <li>
	    size of the crypted key that follows (present only if the
            flag tells an encrypted key is present)
	  </li>
          <li>
	    encrypted key (encrypted by mean of GPG asymetric algorithm,
            present only if flag says so)
	  </li>
	  <li>
	    the slice layout of the backup of reference if the current archive
	    is an isolated catalogue
	  </li>
          <li>
	    eventually, salt, iteration_count and hash algo for key
            derivation fonction (KDF) (used since version 2.6.0 when
            strong encryption is set)
          </li>
          <li>
	    CRC (Cyclic Redundancy Check) computed on the whole version
            header or trailer
          </li>
	</ul>

	<code class=block>
	  +---------+------+---------------+------+--------+-------+----------+---------------+------------+------+------+-----------+-----+-------+
	  | edition | algo | command line  | flag | initial| crypto| crypted  | gnupg crypted |  reference | salt | salt | iteration |hash | CRC   |
	  |         |      |               |      | offset | algo  | key size | sym. key      |  slicing   | size | (KDF)| count(KDF)|(KDF)|       |
	  +---------+------+---------------+------+--------+-------+----------+---------------+------------+------+------+-----------+-----+-------+
	</code>

	<p>
	  The trailer is used when
          reading an archive in direct access mode, to build the proper
          compression layer, escape layer (it is needed if mark have
          been inserted in the archive to un-escape data that could else
          be taken as an escape sequence mark) and encryption layer.
	</p>

	<h4>Slice Layout</h4>

	<p>
	  <b>slice_layout</b> is a class holding information about the first slice size, other file size as well as size of slice headers.
	  It is used in <em>sar</em> class to store the slicing information read from the class <em>header</em> which for historical reason
	  does not use the slice_layout class but read/write those fields as independent ones. Class <em>sar</em> thus relies on class <em>header</em>
	  to read or write this information to/from an archive.
	</p>
	<p>
	  This class is also used in class <em>header_version</em> which is used to read/write the archive header/trailer at the beginning and
	  at the end of the backup. The use of the class <em>slice_layout</em> in that structure is only present for isolated catalogues and contains
	  the slicing information of the archive of reference. The class <em>header_version</em> relies on the <em>slice_layout::read()</em>
	  and <em>slice_layout::write()</em> methods to read and write those fields to/from an archive's trailer/header.
	  Here, this information is needed for the <b>-Tslice</b> option to work when applied to an isolated catalogue.
	</p>
	<p>
	  <b>The isolation process:</b> At the end of archive creation, either read from filesystem (normal isolation) or just written to
	  filesystem (on-fly isolation), the field <em>archive::i_archive::slices</em> holds the slice layout of the current archive. It has been
	  read for the <em>sar</em> layer if the archive is sliced or is set to a default value if there is no <em>sar</em> layer and thus no slicing.
	  The isolation process is defined in <em>archive::i_archive::op_isolate()</em>: it transmits the <em>slices</em> information to
	  <em>macro_tools_create_layers()</em> which, in particular, uses it to setup the <em>header_version</em>. Once the layers are built
	  by this routine, this <em>header_version</em> structure is written at the beginning of the archive (version header).
	  After the catalogue has been written to these layers, <em>op_isolate()</em> calls <em>macro_tools_close_layers()</em> giving it
	  this same generated <em>header_version</em> (containing the slice_layout of the current archive, which will be the reference for the isolated
	  catalogue under construction), for it be written at the end of the archive (this is then the "version trailer").
	</p>

	<h4>The data</h4>

	<p>
          <b>The data</b> is a suite of file contents, with EA and FSA if
          present. When tape mark is used, a copy of the CRC is placed
          after's file Data and file's EA, to be used when reading the
          archive in sequential mode. This CRC is also dropped into the
          catalogue which takes place at the end of the archive to be used
          when reading the archive in direct access mode (the default).
          Last when delta binary is used, a file signature may follow the
          file's data:
	</p>

	<code class=block>
	  &nbsp;....--+---------------------+-------+----+----+------------+------------+----+-----------+-----+-....
	  &nbsp;      |  file1 data         | delta | EA | FSA| file2 data | file3 data | EA | file4     | FSA |
	  &nbsp;      | (may be compressed) | sig   |    |    |(no EA/FSA) |            |    | delta sig |     |
	  &nbsp;....--+---------------------+-------+----+----+------------+------------+----+-----------+-----+-....
	</code>

	<p>
          In the previous archive example,
          we find:
	</p>
	<ul>
          <li>
	    for file1: his data, a delta signature of this data,
            Extended Attributes, and File Specific Attributes
	  </li>
          <li>
	    for file2: only his data, he has no delta signature, no
            EA no FSA
	  </li>
          <li>
	    for file3: data and EA
	  </li>
          <li>
	    for file4: no data, only a delta signature and FSA
	  </li>
	</ul>

	<p>
          file1 shows all fields that can be associated with an inode,
          but none is mandatory, though if present they always follow
          this order:
	</p>
	<ul>
          <li><b>Data</b> or <b>delta patch</b></li>
          <li>followed by data/delta patch CRC when tape marks are set</li>
          <li><b>Delta signature</b></li>
          <li>followed by delta signature CRC when tape marks are set</li>
          <li><b>Extended Attributes</b></li>
          <li>followed by EA CRC when tape maks are set</li>
          <li><b>File Specific Attributes</b></li>
          <li>followed by FSA CRC when tape marks are set</li>
	</ul>

	<p>
	  when tape marks are not set, the CRC listed above are not present
	  inline in the archive, though they are still stored in the <i>catalogue</i>
	  at the end of the archive (see below)
	</p>

	<p>
          More precisely about <b>delta signature</b>
	  combined with tape marks, there is
          additional fields present than just the delta sig and its CRC:
	</p>

	<code class=block>
	  +------+------+------+---------------+----------+--------+
	  | base | sig  | sig  | sig data      | data CRC | result |
	  | CRC  | size | block| (if size > 0) |    if    |  CRC   |
	  | (*)  |      | len  |               | size > 0 |        |
	  +------+------+------+---------------+----------+--------+
	</code>

	<ul>
          <li>
	    <b>base CRC</b> is the
            CRC of the file that delta signature has been based on,
            used at restoration time before applying a patch
	    <br>
	    <dl>
	      <dt class>(*)</dt><dd>
		Since format 11.2 (release 2.7.9) this field has been moved
		with the inode information in the catalogue as well as inlined
		along the backup following seqt_file (tape) mark. This CRC
		is only present if the data for a file contains a binary patch.
	      </dd>
	    </dl>

	  </li>
          <li>
	    <b>sig size</b> gives
            the size of the next two fields (sig block len exists only
	    since archive format 10,1). This may be zero if the file
            does not has signature associated but the file's data is a
            delta patch and thus the base and the result CRCs are needed.
	    When sig size is zero, neither "sig block len" nor "sig data"
	    fields are present.
          </li>
	  <li>
	    <b>sig block len</b>
	    since format 10,1 the block size used for signature calculation
	    is variable (it was fixed to 2048 bytes before) depending on the
	    file size under consideration.
	  </li>
          <li>
	    <b>sig data</b> is the
            delta signature data. This field and the following are the
            only one present outside the ending catalogue about delta
            signatures when tape marks are not set.
          </li>
          <li>
	    <b>data CRC</b> is the
            CRC on the previous "sig block len" + "sig data" fields.
          </li>
          <li>
	    <b>result CRC</b> is
            the CRC of the resulting file once patched (to check
            the patching was successful)
	  </li>
	</ul>

	<h4>The catalogue</h4>

	<p>
          <b>the catalogue</b>, contains all inode, directory structure and
          hard_links information as well as data and EA CRC. The directory
          structure is stored in a simple way: the inode of a directory
          comes, then the inode of the files it contains, then a special
          entry named "EOD" for End of Directory. Considering the
          following tree:
	</p>

	<code class=block>
	  &nbsp;- toto
	  &nbsp;   | titi
	  &nbsp;   | tutu
	  &nbsp;   | tata
	  &nbsp;   |   | blup
	  &nbsp;   |   +--
	  &nbsp;   | boum
	  &nbsp;   | coucou
	  &nbsp;   +---
	</code>

	<p>
	  it would generate the following sequence for catalogue storage:
	</p>
	<code class=block>
	  +-------+------+------+------+------+-----+------+--------+-----+
	  |  toto | titi | tutu | tata | blup | EOD | boum | coucou | EOD |
	  |       |      |      |      |      |     |      |        |     |
	  +-------+------+------+------+------+-----+------+--------+-----+
	</code>

	<p>
          the <i>EOD</i> entries take on byte. Doing this way, there is
	  no need to store the full path
          of each file, just the filename is recorded. The file order and the
	  EOD can be used to find the relative path of each entry.
	</p>

	<p>
	  To be complete, the previous sequence is preceeded by:
	</p>
	<ul>
	  <li>
	    the <b>data_name</b> as describe above,
	  </li>
	  <li>
	    the <b>in-place path</b> which is the root path (-R option) used at backup time that can be used
	    at restoration time by mean of -ap option in place of any -R option.
	  </li>
	</ul>
	<p>
	  This sequence is then followed by a CRC calculated on the whole
	  catalogue dumped data:
	</p>

	<code class=block>
	  +------------+----------+------------------------------------+-----+
	  |  data_name | in_place |<--------catalogue content--------->| CRC |
	  |            |          | toto|titi|tata|blup|EOD|boum|...   |     |
	  +------------+----------+------------------------------------+-----+
	</code>

	  <h4>The terminator</h4>
	<p>
	  <b>the terminator</b> stores the position of the beginning of the
          catalogue, it is the last thing to be written. Thus dar first
          reads the terminator, then the catalogue. Well, there is now two
          terminators, both are meant to be read backward. The second
          terminator points to the beginning of the "trailer version"
          which is read first in direct access mode. The first terminator
          points to the start of the catalogue, which is read once the
          adhoc compression and encryption layers has been built based on
          the information found on the "trailer version"
	</p>

	<h3>All Together</h3>
	<p>
	  Here is an example of how data can be structured in a
	  four sliced archive:
	</p>
	<code class=block>
	  +--------+--------+------------------------+--+
	  | slice  | version|  file data + EA        |Tr|
	  | header | header |                        |  |
	  +--------+--------+------------------------+--+
	</code>
	<p>
          the first slice (just above) has been defined smaller using the
          -S option
	</p>
	<code class=block>
	  +--------+-----------------------------------------------------------------+--+
	  | slice  |           file data + EA                                        |Tr|
	  | header |                                                                 |  |
	  +--------+-----------------------------------------------------------------+--+

	  +--------+-----------------------------------------------------------------+--+
	  | slice  |           file data + EA                                        |Tr|
	  | header |                                                                 |  |
	  +--------+-----------------------------------------------------------------+--+

	  +--------+---------------------+-----------+------ +---------+--------+--+
	  | slice  |   file data + EA    | catalogue | term 1| version | term 2 |Tr|
	  | header |                     |           |       | trailer |        |  |
	  +--------+---------------------+-----------+-------+---------+--------+--+
	</code>

	<p>
	  the last slice is smaller because there was not enough data to make it
          full.
	</p>

	<h3>Other Levels</h3>

	<p>
	  Things get a bit more complicated if we consider compression and encryption.
          The way the problem is addressed in dar's code is a bit like
          networks are designed in computer science, using the notion of
	  <b>layers</b>.
	</p>
	<p>
	  Here, there
          is a additional constraint, a given layer may or may not be
          present (encryption, compression, slicing for example). So all
          layer must have the same interface for serving the layer above
          them.
	</p>

	<p>
	  This interface is defined by the pure virtual class name
          <i>generic_file</i>, which provides generic methods for reading,
          writing, skipping, getting the current offset when writing or
          reading data to any layer. Here follows some example of implementation:
	</p>

	<ul>
	  <li>
	    <p>
	      For example the <b>compressor</b> class
              acts as a file which compresses data wrote to it and writes
              compressed data to another <i>generic_file</i> below it.
	    </p>
	  </li>
	  <li>
	    <p>
	      The <b>strong encryption</b> and <b>scramble</b> classes
	      act the same but in place of
              compressing/uncompressing they encrypt/decrypt the data to/from
              another generic_file object "below" them.
	    </p>
	  </li>
	  <li>
	    <p>
	      The <b>sar</b> class which <b>s</b>egment <b>a</b>nd
	      <b>r</b>eassmbles (perfoming thus the slicing)
              follows the same principle:
              it transfers data wrote to it to several fichier
	      [mening <i>file</i> in French] objects.
	    </p>
	  </li>
	  <li>
	    <p>
              Class <b>fichier</b> also inherit from <i>generic_file</i> class,
	      and is just a wrapper for the plain file system calls.
	    </p>
	  </li>
	  <li>
	    <p>
	      Some new classes
              have been added with format 8, in particular the <b>escape</b> class,
              which inserts escape sequence mark at requested position, and
              modifies data wrote for it never looks like an escape sequence
              mark. To reduce the level of context switch when reading the
              catalogue (which makes a ton of small read)
	    </p>
	  </li>
	  <li>
	    <p>
	      a <b>cache</b> class is also present, it gather small writes made
	      to it into larger writes, and pre-reads a large amount of data
	      to answer to the many small reads when building the catalogue
	      in memory from the archive.
	    </p>
	  </li>
	  <li>
	    <p>
	      <b>fichier_libcurl</b> is acts like the class <i>fichier</i>
	      but as it relies on <i>libcurl</i> it let libdar read and write
	      files to remote repository (using SFTP or FTP as of today).
	    </p>
	  </li>
	  <li>
	    <p>
	      class <b>generic_rsync</b> computes the delta signature, or patch a file
	      using <i>librsync</i>
	    </p>
	  </li>
	  <li>
	    <p>
	      class <b>hash_fichier</b> computes a hash (md5, sha1, sh512,...) of the
	      data it is written to before transmitting it unchanged to the layer below it
	    </p>
	  </li>
	  <li>
	    <p>
	      class <b>memory_file</b> is used to store data in memory.
	      <b>secu_memory_file</b> does the same but in secured memory (relying on
	      <i>libgcrypt</i>. It is used to store sensible data like keys and passwords
	    </p>
	  </li>
	  <li>
	    <p>
	      class <b>null_file</b> acts as the well known <code>/dev/null</code> block
	      device of Unix system (drops all what is written to it and reads not data, like
	      an empty file)
	    </p>
	  </li>
	  <li>
	    <p>
	      class <b>spares_file</b> replaces long sequence of zeroed bytes by a special
	      and small structure (a hole) the contains the number of replaced zeroed bytes
	      when you write data to it. The resulting is transmitted to the underlying layer.
	      At reading time, depending on its mode of operation,  it fetches data from
	      the underlying data and either reconstructs the hole on the filesystem
	      (skipping the write cursor after the hole) or replaces the read <i> hole</i>
	      structure by the zeroed bytes it was replacing.
	    </p>
	  </li>
	</ul>
	<p>
          Here are below is an example of possible layer stacking libdar is using to address a particular combinaison of commands and options:
	</p>

	<code class=block>

	  &nbsp;              +----+--+----+-................+---------+
	  &nbsp;archive       |file|EA|file|                 |catalogue|
	  &nbsp;layout        |data|  |data|                 |         |
	  &nbsp;              +----+--+----+-................+---------+
	  &nbsp;                |   |    |      |              |
	  &nbsp;            +-----+ | +-------+ |              |
	  &nbsp;sparse      |spars| | |sparse | |              |
	  &nbsp;file        |file | | |file   | |              |
	  &nbsp;detection   |detec| | |detect.| |              |
	  &nbsp;layer       +-----+ | +-------+ |              |
	  &nbsp;(optional)      |   |    |      |              |
	  &nbsp;                |   |    |      |              |
  	  &nbsp;            +-----+ |    |   +-----+           |
	  &nbsp;            |delta| |    |   |delta|           |
	  &nbsp;            |sig  | |    |   |sig  |           |
	  &nbsp;            +-----+ |    |   +-----+           |
	  &nbsp;                |   |    |      |              |
	  &nbsp;                |   |    |      |              |
	  &nbsp;                V   V    V      V              V
	  &nbsp;              +-----------------------------------------+
	  &nbsp;compression   |         (compressed)  data              |
	  &nbsp;              +-----------------------------------------+
  	  &nbsp;                     |                      |
	  &nbsp;                     |                      |
	  &nbsp;                     V                      V
	  &nbsp;              +-----------------------------------------+
	  &nbsp;escape layer  |   escaped data / escape sequences       |
	  &nbsp;(optional)    +-----------------------------------------+
	  &nbsp;                    |                      |                 / First Terminateur
	  &nbsp;                    |                      |                 |
	  &nbsp;                    |                      |                 V
	  &nbsp;elastic  +---+      |                      |              +----+---+
	  &nbsp;buffers  |EEE|      |                      |              | T1 |EEE|
	  &nbsp;         +---+      |                      |              +----+---+
	  &nbsp;           |        |                      |                    |           Second
	  &nbsp;           V        V                      V                    V         Terminator
	  &nbsp;         +---------------------------------------------------------+            |
	  &nbsp;cipher   |        (encrypted) data / cache if no encryption        |            |
	  &nbsp;         +---------------------------------------------------------+            V
	  &nbsp;                    |                         |                     +---------+----+
	  &nbsp;+-------+           |                         |                     | trailer | T2 |
	  &nbsp;| header|           |                         |                     +---------+----+
	  &nbsp;+-------+           |                         |                          |      |
	  &nbsp;    |               |                         |                          |      |
	  &nbsp;    V               V                         V                          V      v
	  &nbsp;+----------------------------------------------------------------------------------+
	  &nbsp;|                  data                                                            |
	  &nbsp;+----------------------------------------------------------------------------------+
	  &nbsp;        |         |  |         |   |        |   |        |  |        |  |        |
	  &nbsp;+---------+  +---------+  +---------+  +---------+  +---------+  +---------+  +----+
	  &nbsp;|hash_file|  |hash_file|  |hash_file|  |hash_file|  |hash_file|  |hash_file|  |hash|
	  &nbsp;+---------+  +---------+  +---------+  +---------+  +---------+  +---------+  +----+
	  &nbsp;        |        (hash_file are optional)   |   |        |  |        |  |        |
	  &nbsp;slice   |         |  |         |   |        |   |        |  |        |  |        |
	  &nbsp;headers |         |  |         |   |        |   |        |  |        |  |        |
	  &nbsp; |  |   |         |  |         |   |        |   |        |  |        |  |        |
	  &nbsp; |  +---|------\  |  |         |   |        |   |        |  |        |  |        |
	  &nbsp; |      |      |  |  |         |   |        |   |        |  |        |  |        |
	  &nbsp; V      V      V  V  V         V   V        V   V        V  V        V  V        V
	  &nbsp;+---------+  +---------+  +---------+  +---------+  +---------+  +---------+  +----+
	  &nbsp;|HH| data |  |HH| data |  |HH| data |  |HH| data |  |HH| data |  |HH| data |  |HH| |
	  &nbsp;+---------+  +---------+  +---------+  +---------+  +---------+  +---------+  +----+
	  &nbsp; slice 1      slice 2      slice 3      slice 4       slice 5      slice 6    slice 7
	</code>

	<p>
          The <b>elastic buffers</b> are here to prevent plain text attack, where
          one knows which data is expected at a given place, an trying
          to guess the cipher comparing the expected data and the
          encrypted one. As dar generates structured archives, there
          would have some possibility that one use this attack to crack
          an archive encryption. To overcome this problem, elastic
          buffers have been added at the beginning and at the end of
          encrypted data. This way it is not possible to know where is
          located a given archive structure within the encrypted data.
          The elastic buffers are random data that contain at a random
          place a pattern that tells the overall size of the buffer
          (which size is randomly chosen during archive creation). The
          pattern is of the form "&gt;###&lt;" where the hash field
          (###) contains the elastic buffer size in binary. Small
          elastic buffer can be "&gt;&lt;" for two bytes or "X" for one
          byte, but as it is encrypted beside archive data, it is not
          possible to determine its size for one that does not hold the
          archive encryption key. Elastic buffer are usually several
          kilobyte long. Here follows an example of elastic buffer:
	</p>

	<code class=block>
	  972037219&gt;20&lt;8172839
	</code>

	<p>
	  For clarity,
	  the size field between '&gt;' and '&lt;' has been written in
	  decimal instead of binary, as well as the random data inside the
	  elastic buffer. The location of the size field '&gt;20&lt;' is
	  also randomly chosen at creation time.
	</p>

	<p>
	  <b>Teminateur</b>
	  is short structure that is intended to be read backward. It
	  gives the absolute position of a given item within the archive:
	  The second terminateur let dar skip at the beginning of the
	  archive trailer. The first terminateur (eventually encrypted)
	  let dar skip at the beginning of the catalogue).
	</p>

	<h2><a name="overflow">Overflow in arithmetic integer operations</a></h2>

	<p>
	  Some code explanation about the
          detection of integer arithmetic operation overflows. We speak
          about *unsigned* integers, and we have only portable standard
          ways to detect overflows when using 32 bits or 64 bits integer
          in place of infinint.
	</p>
	<p>
          Written in binary notation, a number is a finite suite of digits (0 or
          1). To obtain the original number from its binary
          representation, we must multiply each digit by successive powers of two.
          Example the binary representation "101101" designs the number N
          where:
	  <code class=block>
            N = 2^5 + 2^3 + 2^2 + 2^0
	  </code>
	<p>
          In that context we will say that 5 is the maximum power of N
          (the power of the higher non null binary digit).
	</p>
	<p>
          For the addition operation ("+"), if an overflow occurs, the result
          is less than one or both operands, so overflow is not difficult to
          detect. To convince you, let's assume that the result is greater
          than both operands while it has overflowed. Thus the real result
          (without overflow) less the first operands should gives the second
          argument, but here we get a value that is greater than the all 1
          bits integer (because there was an overflow and the resulting
          overflowed value is greater than the second and the first
          operand), so this is absurd, and in case of overflow the resulting
          value is less than one of the operands.
	</p>
	<p>
          For substraction operation ("-"), if the second operand is greater
          than the first there will be an overflow (result must be
          unsigned thus positive) else there will not be any overflow.
          Thus detection is even more simple.
	</p>
	<p>
          for division ("/") and modulo ("%") operations, there is never an
          overflow (there is just the illicit division by zero).
	</p>
	<p>
          for multiplication operation ("*"), a heuristic has been chosen to
          quickly detect overflow, the drawback is that it may triggers
          false overflow when number get near the maximum possible integer
          value. Here is the heuristic used:
	</p>
	<p>
          given A and B two integers, which max powers are m and n
          respectively, we have:
	</p>
	<code class=block>
	  A &lt; 2^(m+1)
	</code>
	<p>
          and
	</p>
	<code class=block>
          B &lt; 2^(n+1)
	</code>
	<p>
          thus we also have:
	</p>
	<code class=block>
          A.B &lt; 2^(m+1).2^(n+1)
	</code>
	<p>
          which is:
	</p>
	<code class=block>
          A.B &lt; 2^(m+n+2)
	</code>
	<p>
          In consequences we know that the maximum power of the product of A
          by B is at most m+n+1 and while m+n+1 is less than or equal to the
          maximum power of the integer field there will not be overflow else
          we consider there will be an overflow even if it may not be always
          the case (this is an heuristic algorithm).
	</p>

	<h2><a name="thread_safe">libdar and thread-safe requirement</a></h2>

	<p>
          The following should only concern those who plan to
          use libdar in their own programs.
	</p>

	<p>
          If expect to only have one thread using libdar, there is no
          problem, of course, you will however have to call one of the
          get_version() first, as usual. Thing change if you intend to
          have several concurrent threads using libdar library in a
	  same process.
	</p>
	<p>
          libdar is thread-safe but you need to respect certain conditions:
	</p>
	<p>
          Thread-save support must have been activated at libdar compilation time.
	  Several 'configure' options have an impact this thread-safe support:
	</p>
	<dl>
	  <dt class=void>--enable-test-memory</dt><dd>
	    is a <u>debug option</u> that avoid libdar to be
            thread-safe, so don't use it unless debugging non thread relative
	    behaviors.
	  </dd>
	  <dt class=void>--disable-thread-safe</dt><dd>
	    this disable thread safe support in libdar, it may be used for
	    debugging purposes or to marginally speed up the libdar execution
	  </dd>
	</dl>

	<p>
	  If you want to rely in multi-thread support in libdar, a good practice
	  is to check that a call to <code>libdar::compile_time::thread_safe()</code>
	  returns <code>true</code>. If not your program should behave accordingly
	  either abort with an error or have all libdar interaction done in a single
	  thread at a time.
	</p>

	<dl>
	  <dt class=void>IMPORTANT</dt><dd>
            It is vital to to call one of the
	    <code>get_version(...)</code> as first call to <i>libdar</i> and wait for its return
	    before assuming multi-thread support is ready for use.
	  </dd>
	</dl>

	<p>
          For more information about libdar and its API, check the <a href="api_tutorial.html">doc/api_tutorial.html</a>
          document and the API reference manual under <a href="man/index.html">doc/html/index.html</a>
	</p>


	<h2><a name="NLS">Native Language Support</a></h2>

	<p>
	  Native Language Support (NLS) is
          the ability a given program has to display its messages in different
          languages. For dar, this is implemented using the <b>gettext</b> tools.
          This tool must be installed on the system for dar can be able to
          display messages in another language than English (some would
	  joke telling that this is more Frenglish than good old English, they might
	  be right, so feel free to report any too Frenchy syntax, spelling or grammar)
	</p>
	<p>
	  The dar behavior is the following:
	</p>
	<ul>
	  <li>
            On a system without gettext dar will not use gettext at all.
            All messages will be in English
	  </li>
	  <li>
            On a system with gettext dar will use the system's gettext,
            unless you use <code>--disable-nls</code> option with the configure script.
	  </li>
	</ul>

	<p>
          If NLS is available you just have to set the <code>LANG</code> environment
          variable to your locale settings to change the language in which
          dar displays its messages (see ABOUT-NLS for more about the LANG
          variable).
	</p>
	<p>
          Just for information, gettext() is the name of the call that
          makes translations of string in the program. This call is
          implemented in the library called 'libintl' (intl for
          Internationalization).
	</p>
	<p>
          Refer to the ABOUT-NLS file at the root of the source package to
          learn more about the way to display dar's messages in your own
          language. Note that not all languages are yet supported, this is
          up to you to send me a translation in your language and/or
          contact a translating team as explained in ABOUT-NLS.
	</p>
	<p>
          To know which languages are supported by dar, read the
          po/LINGUAS file and check out for the presence of the
          corresponding *.po files in this directory.
	</p>

	<h2><a name="Dar_relase_Process_in_brief"></a>Dar Release Process</h2>

	<h3>General view</h3>
	<p>
	  Dar does not follow the <i>so called modern continuous development</i> also known
	  as <i>agile</i>. First <i>dar</i> is not a devops code while it can be used
	  <b>for devops</b>. <i>dar</i> is rather a (free) product outside
	  any particuar customer specific context. Second keeping the good old
	  development process separating development from quality engineering in different
	  phases over time brings the expected robustness a backup tool must have.
	</p>
	<p>
	  This separation takes the form of branches of a tree where the <b>trunk</b> is
	  the development code, which grows when receiving new features, and the <b>branches</b>
	  carry the so called <i>releases</i>. Several releases can be found on a
	  given branch but branches only receive bug fixes between two releases.
	</p>
	<p>
	  Several branches may be active at a time, leading to several concurrent releases
	  (for examples releases 2.5.22 and 2.6.5 have the same date).
	</p>
	<p>
	  Last all bug fixes from an older and active branch are merged into the more recent branches
	  up to the trunk, which is the development code. But the modification of code in
	  branches is kept as minimal as possible.
	</p>

	<h3>Phasing</h3>
	<h4>Development Phase:</h4>
	<p>
          During that phase, dar receives new features. At this
          stage sources are modified and tested unitary after each feature addition. These
	  modifications and additions take place on the <i>trunk</i>.
	</p>

	<h4>Frozen API Phase:</h4>
	<p>
	  At this stage, no new feature that would change the API are added.
	  The API shall be documented enough to let API users give their feedback about
          the design and its implementation.
	</p>
	<p>
	  During this time, development may continue, whatever is necessary while it does not changes the
          API, like documentation of the whole project, problem fixes in
          libdar, new features in command-line part of the source, and so
          on. A set of non regression test is also updated and run to check that new features
	  works as expected with old ones and old ones still work fine toghether.
	</p>

	<h4>Pre-release Phase:</h4>
	<p>
	  This phase is announced on <a href="https://sourceforge.net/projects/dar/lists/dar-news">dar-news mailing-list</a>
	  which you can subscribe to also be informed about new releases, security issues, and other major problems.
	</p>
	<p>
	  The goal of this phrase is to let anyone test the release candidate in their own
	  environment and report any code building problem and any bug met to the
	  <a href="http://lists.sourceforge.net/lists/listinfo/dar-pre-release">pre-release mailing-list</a>.
	</p>
	<p>
	  Usually this code is on the trunk (the <i>master</i> in GIT)
	  but <a href="http://dar.linux.free.fr/pre-release/">pre-built packages</a> are also provided daily for testing.
	</p>

	<h4>Release Phase:</h4>
	<p>
          When the pre-release ends, the first official release is provided
          (its last number version is zero like for release 2.7.0). It is available
	  broadly by mean of
	  <a href="http://sourceforge.net/project/showfiles.php?group_id=65612">sourceforge mirrors</a>.
	</p>
	<p>
	  A new branch (like branch_2.7.x) is created to hold the released code, and will receive
	  any further releases by mean of bug fixes (releases 2.7.1, releases 2.7.2,...).
	</p>
	<p>
	  As for any new release, an email is sent with the list of Changes to
          the dar-news mailing-list. During that phase, users are welcome to report
          bugs/problem either first asking for support on the dar-support mailing-list
	  or openning a <a href="http://dar.linux.free.fr/feedback.html">bug report</a>.
	</p>

	<h4>Interim Releases:</h4>
	<p>
	  This type of release is fast to setup and will later become an official release. Doing so
	  saves time to update the website, time to generate the windows binaries, time to push
	  data to the sourceforge mirror update, and so on. Interim releases
	  are provided for validating bug fixes, and are setup from the latest code of a branch.
	  They are usually named after the future release they will produce suffixed by RC#
	  (for <b>R</b>elease <b>C</b>andidate).
	  For example, we got version <i>2.6.6.RC2</i> in the past. Note that interim releases
	  are deleted once the official release is done (release 2.6.6 in our example).
	</p>

	<h3><a name="Dar_version_naming"></a>Dar's Version<u>s</u></h3>

	<h4>package release version</h4>
	<p>
	  Dar packages are release during
          the pre-release phase
	  (<a href="#Dar_relase_Process_in_brief">see above</a>).
	  Each version is identified by three number separated
	  by dot like for example, version 2.3.0:
	</p>
	<ul>
	  <li>
	    <b>The last number</b> is
            incremented between releases that take place in the same release
            phase (keeping trace of bug fixes),
	  </li>
	  <li>
	    <b>the middle number</b> increments at
            each pre-release phase and designate a branch.
	  </li>
	  <li>
	    <b>the first number</b> is incremented when
            a major change in the software structure takes place, like for
	    version 2.0.0 which has seen the monolithic <i>dar</i> software
	    being split in two:  the <i>libdar</i> and its API on one side and
	    on the other side <i>dar</i> reduced to translating
	    between command-line and the API
	  </li>
	</ul>
	<p>
          Note that release versionning is completely different from what is
          done for the Linux kernel, here for dar all versionnized packages
          are stable released software and thus stability increases with the
          last number of the version.
	</p>

	<h4>Libdar version</h4>
	<p>
	  Unfortunately, the release version
          does not give much information about the compatibility of
          different libdar version, from the point of view of an external
          application, that thus has not been released with libdar and may
          be faced to different libdar versions. So, libdar has its own
          version. It is also a three number version, (for example,
          libdar version 6.2.7), but each number has a different
          meaning:
	</p>
	<ul>
	  <li>
	    <b>The last number</b> increases with a new version that only
	    fixes bugs (same meaning as package versioning)
	  </li>
	  <li>
	    <b>the middle number increases</b> with when new features has
            been added but older features can be used the same (a software that
	    relied on the previous middle number should still work with a more
	    recent one)
	  </li>
	  <li>
	    <b>the first number</b> changes when the API had to be modified
	    in a way that the use of old feature has changed
	  </li>
	</ul>

	<h4>Other versions</h4>
	<p>
          Beside the libdar library, you can find several command-line
          applications: dar, dar_xform, dar_slave, dar_manager, dar_cp and dar_split.
          These have their own version which is, here too, made of
          three numbers. Their meaning is the same as the meaning for the
          package release version: The last number increases upon bug fix,
          the middle upon new feature, the first upon major architecture
          changes.
	</p>

	<h4>Archive format version</h4>
	<p>
	  When new features come, it is sometime necessary to change the
	  structure of the archive. To be able to know the format used in
	  the archive, a field is present in each archive that defines this
	  format.
	</p>

	<p>
	  Currently the archive format version is <i>11</i> meaning there is
	  around 11 different version of dar archive format, plus the version
	  increment that were needed to fix a bug (like version 8.1 and 10.1), but
	  <i>dar</i> and <i>libdar</i> have been developped to be able to read
	  them all (all those that existed when the version was released).
	</p>

	<h4>Cross reference matrix</h4>
	<p>
	  OK, you may now find that this is
          a bit complex so a list of version is give below. Just remember
          that there are two points of view: The command-line user and the
          external application developer.
	</p>

	<div class=table>
	  <table class=center>
            <tr>
              <th style="min-width:200px">Date</th>
              <th>release (and dar version)</th>
              <th>Archive format</th>
              <th>Database Format</th>
              <th>libdar version</th>
              <th>dar_xform</th>
              <th>dar_slave</th>
              <th>dar_manager</th>
              <th>dar_cp</th>
              <th>dar_split</th>
            </tr>
            <tr>
              <td>April 2nd, 2002</td>
              <th>1.0.0</th>
              <td>01</td>
              <td>----- </td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>April 24th, 2002</td>
              <th>1.0.1</th>
              <td>01</td>
              <td>----- </td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>May 8th, 2002</td>
              <th>1.0.2</th>
              <td>01</td>
              <td>----- </td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>May 27th, 2002</td>
              <th>1.0.3</th>
              <td>01</td>
              <td>----- </td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>June 26th, 2002</td>
              <th>1.1.0</th>
              <td>02</td>
              <td>----- </td>
              <td>-----</td>
              <td>1.0.0</td>
              <td>1.0.0</td>
              <td>-----</td>
              <td>-----</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Nov. 4th, 2002</td>
              <th>1.2.0</th>
              <td>03</td>
              <td>01 </td>
              <td>-----</td>
              <td>1.1.0</td>
              <td>1.1.0</td>
              <td>1.0.0</td>
              <td>-----</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Jan. 10th, 2003</td>
              <th>1.2.1</th>
              <td>03</td>
              <td>01 </td>
              <td>-----</td>
              <td>1.1.0</td>
              <td>1.1.0</td>
              <td>1.0.0</td>
              <td>-----</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>May 19th, 2003</td>
              <th>1.3.0</th>
              <td>03</td>
              <td>01 </td>
              <td>-----</td>
              <td>1.1.0</td>
              <td>1.1.0</td>
              <td>1.1.0</td>
              <td>-----</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Nov. 2nd, 2003</td>
              <th>2.0.0</th>
              <td>03</td>
              <td>01 </td>
              <td>1.0.0</td>
              <td>1.1.0</td>
              <td>1.1.0</td>
              <td>1.2.0</td>
              <td>1.0.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Nov. 21th, 2003</td>
              <th>2.0.1</th>
              <td>03</td>
              <td>01 </td>
              <td>1.0.1</td>
              <td>1.1.0</td>
              <td>1.1.0</td>
              <td>1.2.0</td>
              <td>1.0.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Dec. 7th, 2003</td>
              <th>2.0.2</th>
              <td>03</td>
              <td>01 </td>
              <td>1.0.2</td>
              <td>1.1.0</td>
              <td>1.1.0</td>
              <td>1.2.0</td>
              <td>1.0.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Dec. 14th, 2003</td>
              <th>2.0.3</th>
              <td>03</td>
              <td>01 </td>
              <td>1.0.2</td>
              <td>1.1.0</td>
              <td>1.1.0</td>
              <td>1.2.1</td>
              <td>1.0.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Jan. 3rd, 2004</td>
              <th>2.0.4</th>
              <td>03</td>
              <td>01 </td>
              <td>1.0.2</td>
              <td>1.1.0</td>
              <td>1.1.0</td>
              <td>1.2.1</td>
              <td>1.0.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Feb. 8th, 2004</td>
              <th>2.1.0</th>
              <td>03</td>
              <td>01 </td>
              <td>2.0.0</td>
              <td>1.2.0</td>
              <td>1.2.0</td>
              <td>1.2.1</td>
              <td>1.0.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>March 5th, 2004</td>
              <th>2.1.1</th>
              <td>03</td>
              <td>01 </td>
              <td>2.0.1</td>
              <td>1.2.1</td>
              <td>1.2.1</td>
              <td>1.2.2</td>
              <td>1.0.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>March 12th, 2004</td>
              <th>2.1.2</th>
              <td>03</td>
              <td>01 </td>
              <td>2.0.2</td>
              <td>1.2.1</td>
              <td>1.2.1</td>
              <td>1.2.2</td>
              <td>1.0.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>May 6th, 2004</td>
              <th>2.1.3</th>
              <td>03</td>
              <td>01 </td>
              <td>2.0.3</td>
              <td>1.2.1</td>
              <td>1.2.1</td>
              <td>1.2.2</td>
              <td>1.0.1</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>July 13th, 2004</td>
              <th>2.1.4</th>
              <td>03</td>
              <td>01 </td>
              <td>2.0.4</td>
              <td>1.2.1</td>
              <td>1.2.1</td>
              <td>1.2.2</td>
              <td>1.0.1</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Sept. 12th, 2004</td>
              <th>2.1.5</th>
              <td>03</td>
              <td>01 </td>
              <td>2.0.5</td>
              <td>1.2.1</td>
              <td>1.2.1</td>
              <td>1.2.2</td>
              <td>1.0.1</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Jan. 29th, 2005</td>
              <th>2.1.6</th>
              <td>03</td>
              <td>01 </td>
              <td>2.0.5</td>
              <td>1.2.1</td>
              <td>1.2.1</td>
              <td>1.2.2</td>
              <td>1.0.1</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Jan. 30th, 2005</td>
              <th>2.2.0</th>
              <td>04</td>
              <td>01 </td>
              <td>3.0.0</td>
              <td>1.3.0</td>
              <td>1.3.0</td>
              <td>1.3.0</td>
              <td>1.0.1</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Feb. 20th, 2005</td>
              <th>2.2.1</th>
              <td>04</td>
              <td>01 </td>
              <td>3.0.1</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.0.1</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>May 12th, 2005</td>
              <th>2.2.2</th>
              <td>04</td>
              <td>01 </td>
              <td>3.0.2</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.0.2</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Sept. 13th, 2005</td>
              <th>2.2.3</th>
              <td>04</td>
              <td>01 </td>
              <td>3.1.0</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.0.2</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Nov. 5th, 2005</td>
              <th>2.2.4</th>
              <td>04</td>
              <td>01 </td>
              <td>3.1.1</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.0.2</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Dec. 6th, 2005</td>
              <th>2.2.5</th>
              <td>04</td>
              <td>01 </td>
              <td>3.1.2</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.0.2</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Jan. 19th, 2006</td>
              <th>2.2.6</th>
              <td>04</td>
              <td>01 </td>
              <td>3.1.3</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.0.3</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Feb. 24th, 2006</td>
              <th>2.2.7</th>
              <td>04</td>
              <td>01 </td>
              <td>3.1.4</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.3.1</td>
              <td>1.0.3</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Feb. 24th, 2006</td>
              <th>2.3.0</th>
              <td>05</td>
              <td>01 </td>
              <td>4.0.0</td>
              <td>1.4.0</td>
              <td>1.3.2</td>
              <td>1.4.0</td>
              <td>1.1.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>June 26th, 2006</td>
              <th>2.3.1</th>
              <td>05</td>
              <td>01 </td>
              <td>4.0.1</td>
              <td>1.4.0</td>
              <td>1.3.2</td>
              <td>1.4.0</td>
              <td>1.1.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Oct. 30th, 2006</td>
              <th>2.3.2</th>
              <td>05</td>
              <td>01 </td>
              <td>4.0.2</td>
              <td>1.4.0</td>
              <td>1.3.2</td>
              <td>1.4.0</td>
              <td>1.1.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Feb. 24th, 2007</td>
              <th>2.3.3</th>
              <td>05</td>
              <td>01 </td>
              <td>4.1.0</td>
              <td>1.4.0</td>
              <td>1.3.2</td>
              <td>1.4.1</td>
              <td>1.2.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>June 30th, 2007</td>
              <th>2.3.4</th>
              <td>06</td>
              <td>01 </td>
              <td>4.3.0</td>
              <td>1.4.0</td>
              <td>1.3.2</td>
              <td>1.4.1</td>
              <td>1.2.0</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Aug. 28th, 2007</td>
              <th>2.3.5</th>
              <td>06</td>
              <td>01 </td>
              <td>4.4.0</td>
              <td>1.4.1</td>
              <td>1.3.3</td>
              <td>1.4.2</td>
              <td>1.2.1</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Sept. 29th, 2007</td>
              <th>2.3.6</th>
              <td>06</td>
              <td>01 </td>
              <td>4.4.1</td>
              <td>1.4.1</td>
              <td>1.3.3</td>
              <td>1.4.2</td>
              <td>1.2.1</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Feb. 10th, 2008</td>
              <th>2.3.7</th>
              <td>06</td>
              <td>01 </td>
              <td>4.4.2</td>
              <td>1.4.2</td>
              <td>1.3.4</td>
              <td>1.4.3</td>
              <td>1.2.2</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>June 20th, 2008</td>
              <th>2.3.8</th>
              <td>07</td>
              <td>01 </td>
              <td>4.4.3</td>
              <td>1.4.2</td>
              <td>1.3.4</td>
              <td>1.4.3</td>
              <td>1.2.2</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>May 22nd, 2009</td>
              <th>2.3.9</th>
              <td>07</td>
              <td>01 </td>
              <td>4.4.4</td>
              <td>1.4.2</td>
              <td>1.3.4</td>
              <td>1.4.3</td>
              <td>1.2.2</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>April 9th, 2010</td>
              <th>2.3.10</th>
              <td>07</td>
              <td>01 </td>
              <td>4.4.5</td>
              <td>1.4.2</td>
              <td>1.3.4</td>
              <td>1.4.3</td>
              <td>1.2.2</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>March 13th, 2011</td>
              <th>2.3.11 </th>
              <td>07 </td>
              <td>01 </td>
              <td>4.5.0 </td>
              <td>1.4.3 </td>
              <td>1.3.4 </td>
              <td>1.4.3 </td>
              <td>1.2.2 </td>
              <td>-----</td>
            </tr>
            <tr>
              <td>February 25th, 2012</td>
              <th>2.3.12</th>
              <td>07</td>
              <td>01</td>
              <td>4.5.1</td>
              <td>1.4.3</td>
              <td>1.3.4</td>
              <td>1.4.3</td>
              <td>1.2.2</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>June 2nd, 2011</td>
              <th>2.4.0</th>
              <td>08</td>
              <td>02</td>
              <td>5.0.0</td>
              <td>1.5.0</td>
              <td>1.4.0</td>
              <td>1.5.0</td>
              <td>1.2.3</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>July 21st, 2011</td>
              <th>2.4.1</th>
              <td>08</td>
              <td>02</td>
              <td>5.1.0</td>
              <td>1.5.0</td>
              <td>1.4.0</td>
              <td>1.6.0</td>
              <td>1.2.3</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Sept. 5th, 2011</td>
              <th>2.4.2</th>
              <td>08</td>
              <td>02</td>
              <td>5.1.1</td>
              <td>1.5.0</td>
              <td>1.4.0</td>
              <td>1.6.0</td>
              <td>1.2.3</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>February 25th, 2012</td>
              <th>2.4.3</th>
              <td>08</td>
              <td>03</td>
              <td>5.2.0</td>
              <td>1.5.0</td>
              <td>1.4.0</td>
              <td>1.7.0</td>
              <td>1.2.3</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>March 17th, 2012</td>
              <th>2.4.4</th>
              <td>08</td>
              <td>03</td>
              <td>5.2.1</td>
              <td>1.5.0</td>
              <td>1.4.0</td>
              <td>1.7.1</td>
              <td>1.2.3</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>April 15th, 2012</td>
              <th>2.4.5</th>
              <td>08</td>
              <td>03</td>
              <td>5.2.2</td>
              <td>1.5.1</td>
              <td>1.4.1</td>
              <td>1.7.2</td>
              <td>1.2.4</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>June 24th, 2012</td>
              <th>2.4.6</th>
              <td>08</td>
              <td>03</td>
              <td>5.2.3</td>
              <td>1.5.2</td>
              <td>1.4.2</td>
              <td>1.7.3</td>
              <td>1.2.5</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>July 5th, 2012</td>
              <th>2.4.7</th>
              <td>08</td>
              <td>03</td>
              <td>5.2.4</td>
              <td>1.5.2</td>
              <td>1.4.3</td>
              <td>1.7.3</td>
              <td>1.2.5</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>September 9th, 2012</td>
              <th>2.4.8</th>
              <td>08</td>
              <td>03</td>
              <td>5.3.0</td>
              <td>1.5.3</td>
              <td>1.4.4</td>
              <td>1.7.4</td>
              <td>1.2.6</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>January 6th, 2013</td>
              <th>2.4.9</th>
              <td>08</td>
              <td>03</td>
              <td>5.3.1</td>
              <td>1.5.3</td>
              <td>1.4.4</td>
              <td>1.7.4</td>
              <td>1.2.7</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>March 9th, 2013</td>
              <th>2.4.10</th>
              <td>08</td>
              <td>03</td>
              <td>5.3.2</td>
              <td>1.5.3</td>
              <td>1.4.4</td>
              <td>1.7.4</td>
              <td>1.2.7</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>Aug. 26th, 2013</td>
              <th>2.4.11</th>
              <td>08</td>
              <td>03</td>
              <td>5.4.0</td>
              <td>1.5.4</td>
              <td>1.4.5</td>
              <td>1.7.5</td>
              <td>1.2.8</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>January 19th, 2014</td>
              <th>2.4.12</th>
              <td>08</td>
              <td>03</td>
              <td>5.5.0</td>
              <td>1.5.4</td>
              <td>1.4.5</td>
              <td>1.7.6</td>
              <td>1.2.8</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>April 21st, 2014</td>
              <th>2.4.13</th>
              <td>08</td>
              <td>03</td>
              <td>5.6.0</td>
              <td>1.5.5</td>
              <td>1.4.5</td>
              <td>1.7.7</td>
              <td>1.2.8</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>June 15th, 2014</td>
              <th>2.4.14</th>
              <td>08</td>
              <td>03</td>
              <td>5.6.1</td>
              <td>1.5.5</td>
              <td>1.4.5</td>
              <td>1.7.7</td>
              <td>1.2.8</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>September 6th, 2014</td>
              <th>2.4.15</th>
              <td>08</td>
              <td>03</td>
              <td>5.6.2</td>
              <td>1.5.6</td>
              <td>1.4.6</td>
              <td>1.7.8</td>
              <td>1.2.8</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>January 18th, 2015</td>
              <th>2.4.16</th>
              <td>08</td>
              <td>03</td>
              <td>5.6.3</td>
              <td>1.5.6</td>
              <td>1.4.6</td>
              <td>1.7.8</td>
              <td>1.2.8</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>January 31th, 2015</td>
              <th>2.4.17</th>
              <td>08</td>
              <td>03</td>
              <td>5.6.4</td>
              <td>1.5.6</td>
              <td>1.4.6</td>
              <td>1.7.8</td>
              <td>1.2.8</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>August 30th, 2015</td>
              <th>2.4.18 </th>
              <td>08.1 </td>
              <td>03</td>
              <td>5.6.5</td>
              <td>1.5.6</td>
              <td>1.4.6</td>
              <td>1.7.8</td>
              <td>1.2.8</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>October 4th, 2015</td>
              <th>2.4.19</th>
              <td>08.1</td>
              <td>03</td>
              <td>5.6.6</td>
              <td>1.5.6</td>
              <td>1.4.6</td>
              <td>1.7.8</td>
              <td>1.2.8</td>
              <td>----- </td>
            </tr>
            <tr>
              <td>November 21th, 2015</td>
              <th>2.4.20</th>
              <td>08.1</td>
              <td>03</td>
              <td>5.6.7</td>
              <td>1.5.8</td>
              <td>1.4.8</td>
              <td>1.7.10</td>
              <td>1.2.10</td>
              <td>----- </td>
            </tr>
            <tr>
              <td>April 24th, 2016</td>
              <th>2.4.21</th>
              <td>08.1</td>
              <td>03 </td>
              <td>5.6.8 </td>
              <td>1.5.9 </td>
              <td>1.4.9 </td>
              <td>1.7.11 </td>
              <td>1.2.10</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>June 5th, 2016</td>
              <th>2.4.22</th>
              <td>08.1</td>
              <td>03 </td>
              <td>5.6.9</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.11</td>
              <td>1.2.10</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>October 29th, 2016</td>
              <th>2.4.23</th>
              <td>08.1</td>
              <td>03 </td>
              <td>5.6.9</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.11</td>
              <td>1.2.10</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>January 21st, 2017</td>
              <th>2.4.24</th>
              <td>08.1</td>
              <td>03 </td>
              <td>5.6.10</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.11</td>
              <td>1.2.10</td>
              <td>-----</td>
            </tr>
            <tr>
              <td>October 4th, 2015</td>
              <th>2.5.0</th>
              <td>09</td>
              <td>04</td>
              <td>5.7.0</td>
              <td>1.5.7</td>
              <td>1.4.7</td>
              <td>1.7.9</td>
              <td>1.2.9</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>October 17th, 2015</td>
              <th>2.5.1</th>
              <td>09</td>
              <td>04</td>
              <td>5.7.1</td>
              <td>1.5.8</td>
              <td>1.4.8</td>
              <td>1.7.10</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>November 21st, 2015</td>
              <th>2.5.2</th>
              <td>09</td>
              <td>04</td>
              <td>5.7.2</td>
              <td>1.5.8</td>
              <td>1.4.8</td>
              <td>1.7.10</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>January 4th, 2016</td>
              <th>2.5.3</th>
              <td>09</td>
              <td>04</td>
              <td>5.7.3</td>
              <td>1.5.8</td>
              <td>1.4.8</td>
              <td>1.7.10</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>April 24th, 2016</td>
              <th>2.5.4</th>
              <td>09 </td>
              <td>04 </td>
              <td>5.8.0 </td>
              <td>1.5.9 </td>
              <td>1.4.9 </td>
              <td>1.7.11 </td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>June 5th, 2016</td>
              <th>2.5.5</th>
              <td>09</td>
              <td>04</td>
              <td>5.8.1</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.11</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>September 10th, 2016</td>
              <th>2.5.6</th>
              <td>09</td>
              <td>04</td>
              <td>5.8.2</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.11</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>October 29th, 2016</td>
              <th>2.5.7</th>
              <td>09</td>
              <td>04</td>
              <td>5.8.3</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.11</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>January 2nd, 2017</td>
              <th>2.5.8</th>
              <td>09</td>
              <td>04</td>
              <td>5.8.4</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.11</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>January 21st, 2017</td>
              <th>2.5.9</th>
              <td>09</td>
              <td>04</td>
              <td>5.9.0</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.11</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>April 4th, 2017</td>
              <th>2.5.10</th>
              <td>09</td>
              <td>04</td>
              <td>5.10.0</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.11</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>June 23rd, 2017</td>
              <th>2.5.11</th>
              <td>09</td>
              <td>04</td>
              <td>5.11.0</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.12</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>September 2nd,2017</td>
              <th>2.5.12</th>
              <td>09</td>
              <td>04</td>
              <td>5.11.1</td>
              <td>1.5.9</td>
              <td>1.4.9</td>
              <td>1.7.12</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>October 28th, 2017</td>
              <th>2.5.13</th>
              <td>09</td>
              <td>04</td>
              <td>5.12.0</td>
              <td>1.5.10</td>
              <td>1.4.10</td>
              <td>1.7.13</td>
              <td>1.2.10</td>
              <td>1.0.0</td>
            </tr>
            <tr>
              <td>December 20th, 2017</td>
              <th>2.5.14</th>
              <td>09</td>
              <td>04</td>
              <td>5.12.1</td>
              <td>1.5.10</td>
              <td>1.4.10</td>
              <td>1.7.13</td>
              <td>1.2.10</td>
              <td>1.1.1</td>
            </tr>
            <tr>
              <td>April 28th, 2018</td>
              <th>2.5.15</th>
              <td>09</td>
              <td>04</td>
              <td>5.12.2</td>
              <td>1.5.10</td>
              <td>1.4.10</td>
              <td>1.7.13</td>
              <td>1.2.10</td>
              <td>1.1.1</td>
            </tr>
            <tr>
              <td>July 19th, 2018</td>
              <th>2.5.16</th>
              <td>09</td>
              <td>04</td>
              <td>5.12.3</td>
              <td>1.5.10</td>
              <td>1.4.10</td>
              <td>1.7.13</td>
              <td>1.2.10</td>
              <td>1.1.1</td>
            </tr>
            <tr>
              <td>September 30th, 2018</td>
              <th>2.5.17</th>
              <td>09</td>
              <td>04</td>
              <td>5.13.0</td>
              <td>1.5.10</td>
              <td>1.4.10</td>
              <td>1.7.13</td>
              <td>1.2.10</td>
              <td>1.1.1</td>
            </tr>
            <tr>
              <td>December 8th, 2018</td>
              <th>2.5.18</th>
              <td>09</td>
              <td>04</td>
              <td>5.13.1</td>
              <td>1.5.10</td>
              <td>1.4.10</td>
              <td>1.7.13</td>
              <td>1.2.10</td>
              <td>1.1.1</td>
            </tr>
            <tr>
              <td>January 19th, 2019</td>
              <th>2.5.19</th>
              <td>09</td>
              <td>04</td>
              <td>5.13.2</td>
              <td>1.5.11</td>
              <td>1.4.11</td>
              <td>1.7.14</td>
              <td>1.2.10</td>
              <td>1.1.1</td>
            </tr>
	    <tr>
              <td>February 9th, 2019</td>
              <th>2.5.20</th>
              <td>09</td>
              <td>04</td>
              <td>5.13.3</td>
              <td>1.5.11</td>
              <td>1.4.11</td>
              <td>1.7.14</td>
              <td>1.2.10</td>
              <td>1.1.1</td>
	    </tr>
	    <tr>
              <td>May 25th, 2019</td>
              <th>2.5.21</th>
              <td>09</td>
              <td>04</td>
              <td>5.13.4</td>
              <td>1.5.11</td>
              <td>1.4.11</td>
              <td>1.7.14</td>
              <td>1.2.10</td>
              <td>1.1.1</td>
            </tr>
	    <tr>
              <td>July 6th, 2019</td>
              <th>2.5.22</th>
              <td>09</td>
              <td>04</td>
              <td>5.13.5</td>
              <td>1.5.11</td>
              <td>1.4.11</td>
              <td>1.7.14</td>
              <td>1.2.10</td>
              <td>1.1.1</td>
            </tr>
            <tr>
              <td>December 16th, 2018</td>
              <th>2.6.0</th>
              <td>10</td>
              <td>05</td>
              <td>6.0.0</td>
              <td>1.6.0</td>
              <td>1.5.0</td>
              <td>1.8.0</td>
              <td>1.2.11</td>
              <td>1.1.2</td>
            </tr>
            <tr>
              <td>January 19th, 2019</td>
              <th>2.6.1</th>
              <td>10</td>
              <td>05</td>
              <td>6.0.1</td>
              <td>1.6.0</td>
              <td>1.5.0</td>
              <td>1.8.0</td>
              <td>1.2.11</td>
              <td>1.1.2</td>
            </tr>
	    <tr>
              <td>February 9th, 2019</td>
              <th>2.6.2</th>
              <td>10</td>
              <td>05</td>
              <td>6.0.2</td>
              <td>1.6.1</td>
              <td>1.5.1</td>
              <td>1.8.1</td>
              <td>1.2.12</td>
              <td>1.1.2</td>
            </tr>
	    <tr>
	      <td>March 30th, 2019</td>
              <th>2.6.3</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.1.0</td>
              <td>1.6.1</td>
              <td>1.5.1</td>
              <td>1.8.1</td>
              <td>1.2.12</td>
              <td>1.1.2</td>
            </tr>
	    <tr>
	      <td>May 25th, 2019</td>
              <th>2.6.4</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.1.1</td>
              <td>1.6.1</td>
              <td>1.5.1</td>
              <td>1.8.1</td>
              <td>1.2.12</td>
              <td>1.1.2</td>
	    </tr>
	    <tr>
	      <td>July 6th, 2019</td>
              <th>2.6.5</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.1.2</td>
              <td>1.6.1</td>
              <td>1.5.1</td>
              <td>1.8.1</td>
              <td>1.2.12</td>
              <td>1.1.2</td>
	    </tr>
	    <tr>
	      <td>September 21st, 2019</td>
              <th>2.6.6</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.0</td>
              <td>1.6.2</td>
              <td>1.5.2</td>
              <td>1.8.2</td>
              <td>1.2.12</td>
              <td>1.1.2</td>
	    </tr>
	    <tr>
	      <td>January 12th, 2020</td>
              <th>2.6.7</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.1</td>
              <td>1.6.2</td>
              <td>1.5.2</td>
              <td>1.8.2</td>
              <td>1.2.12</td>
              <td>1.1.2</td>
	    </tr>
	    <tr>
	      <td>February 8th, 2020</td>
              <th>2.6.8</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.2</td>
              <td>1.6.2</td>
              <td>1.5.2</td>
              <td>1.8.2</td>
              <td>1.2.12</td>
              <td>1.1.2</td>
	    </tr>
	    <tr>
	      <td>March 22nd, 2020</td>
              <th>2.6.9</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.3</td>
              <td>1.6.2</td>
              <td>1.5.2</td>
              <td>1.8.2</td>
              <td>1.2.12</td>
              <td>1.1.2</td>
	    </tr>
	    <tr>
	      <td>May 31st, 2020</td>
              <th>2.6.10</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.4</td>
              <td>1.6.3</td>
              <td>1.5.3</td>
              <td>1.8.3</td>
              <td>1.2.13</td>
              <td>1.1.3</td>
	    </tr>
	    <tr>
	      <td>September 5th, 2020</td>
              <th>2.6.11</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.5</td>
              <td>1.6.3</td>
              <td>1.5.3</td>
              <td>1.8.3</td>
              <td>1.2.13</td>
              <td>1.1.3</td>
	    </tr>
	    <tr>
	      <td>September 11th, 2020</td>
              <th>2.6.12</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.6</td>
              <td>1.6.3</td>
              <td>1.5.3</td>
              <td>1.8.3</td>
              <td>1.2.13</td>
              <td>1.1.3</td>
	    </tr>
	    <tr>
	      <td>November 8th, 2020</td>
              <th>2.6.13</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.7</td>
              <td>1.6.3</td>
              <td>1.5.3</td>
              <td>1.8.3</td>
              <td>1.2.13</td>
              <td>1.1.3</td>
	    </tr>
	    <tr>
	      <td>March 14th, 2021</td>
              <th>2.6.14</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.8</td>
              <td>1.6.3</td>
              <td>1.5.3</td>
              <td>1.8.3</td>
              <td>1.2.13</td>
              <td>1.1.3</td>
	    </tr>
	    <tr>
	      <td>May 13th, 2021</td>
	      <th>2.6.15</th>
	      <td>10.1</td>
	      <td>05</td>
	      <td>6.2.9</td>
	      <td>1.6.3</td>
	      <td>1.5.3</td>
	      <td>1.8.3</td>
	      <td>1.2.13</td>
	      <td>1.1.3</td>
	    </tr>
	    <tr>
	      <td>January 2nd, 2022</td>
              <th>2.6.16</th>
              <td>10.1</td>
              <td>05</td>
              <td>6.2.10</td>
              <td>1.6.3</td>
              <td>1.5.3</td>
              <td>1.8.3</td>
              <td>1.2.13</td>
              <td>1.1.3</td>
	    </tr>
	    <tr>
	      <td>April 24th, 2021</td>
              <th>2.7.0</th>
              <td>11</td>
              <td>06</td>
              <td>6.3.0</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.2.13</td>
              <td>1.2.0</td>
	    </tr>
	    <tr>
	      <td>May 13th, 2021</td>
              <th>2.7.1</th>
              <td>11.1</td>
              <td>06</td>
              <td>6.4.0</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.2.13</td>
              <td>1.2.0</td>
	    </tr>
	    <tr>
	      <td>September 25th, 2021</td>
              <th>2.7.2</th>
              <td>11.1</td>
              <td>06</td>
              <td>6.4.1</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.2.13</td>
              <td>1.2.0</td>
	    </tr>
	    <tr>
	      <td>January 2nd, 2022</td>
              <th>2.7.3</th>
              <td>11.1</td>
              <td>06</td>
              <td>6.4.2</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.2.13</td>
              <td>1.2.1</td>
	    </tr>
	    <tr>
	      <td>March 13th, 2022</td>
              <th>2.7.4</th>
              <td>11.1</td>
              <td>06</td>
              <td>6.4.3</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.2.13</td>
              <td>1.2.1</td>
	    </tr>
	    <tr>
	      <td>April 13th, 2022</td>
              <th>2.7.5</th>
              <td>11.1</td>
              <td>06</td>
              <td>6.4.4</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.2.13</td>
              <td>1.2.1</td>
	    </tr>
	    <tr>
	      <td>June 19th, 2022</td>
              <th>2.7.6</th>
              <td>11.1</td>
              <td>06</td>
              <td>6.4.5</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>August 7th, 2022</td>
              <th>2.7.7</th>
              <td>11.1</td>
              <td>06</td>
              <td>6.4.6</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>November 29th, 2022</td>
              <th>2.7.8</th>
              <td>11.1</td>
              <td>06</td>
              <td>6.5.0</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>March 26th, 2023</td>
              <th>2.7.9</th>
              <td>11.2</td>
              <td>06</td>
              <td>6.5.1</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>June 25th, 2023</td>
              <th>2.7.10</th>
              <td>11.2</td>
              <td>06</td>
              <td>6.6.0</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>August 5th, 2023</td>
              <th>2.7.11</th>
              <td>11.2</td>
              <td>06</td>
              <td>6.6.1</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>September 3rd, 2023</td>
              <th>2.7.12</th>
              <td>11.3</td>
              <td>06</td>
              <td>6.7.0</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>October 1st, 2023</td>
              <th>2.7.13</th>
              <td>11.3</td>
              <td>06</td>
              <td>6.7.1</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>March 23rd, 2024</td>
              <th>2.7.14</th>
              <td>11.3</td>
              <td>06</td>
              <td>6.7.2</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>June 6th, 2024</td>
              <th>2.7.15</th>
              <td>11.3</td>
              <td>06</td>
              <td>6.7.3</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>December 9th, 2024</td>
              <th>2.7.16</th>
              <td>11.3</td>
              <td>06</td>
              <td>6.8.0</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>March 22nd, 2025</td>
              <th>2.7.17</th>
              <td>11.3</td>
              <td>06</td>
              <td>6.8.1</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>May 20th, 2025</td>
              <th>2.7.18</th>
              <td>11.3</td>
              <td>06</td>
              <td>6.8.2</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>july 31st, 2025</td>
              <th>2.7.19</th>
              <td>11.3</td>
              <td>06</td>
              <td>6.8.3</td>
              <td>1.7.0</td>
              <td>1.6.0</td>
              <td>1.9.0</td>
              <td>1.3.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>August 2nd, 2025</td>
              <th>2.8.0</th>
              <td>11.3</td>
              <td>06</td>
              <td>7.0.0</td>
              <td>1.7.0</td>
              <td>1.6.1</td>
              <td>1.10.0</td>
              <td>1.4.0</td>
              <td>1.2.2</td>
	    </tr>
	    <tr>
	      <td>September 24th, 2025</td>
              <th>2.8.1</th>
              <td>11.3</td>
              <td>06</td>
              <td>7.0.0</td>
              <td>1.7.0</td>
              <td>1.6.1</td>
              <td>1.10.0</td>
              <td>1.4.0</td>
              <td>1.2.2</td>
	    </tr>
	  </table>
	</div>

	<h2><a name="scrambling">Scrambling (weak encryption)</a></h2>

	<p>
          Before strong encryption was
          implemented, dar had only a very simple and weak encryption
          mechanism. It remains available in current release under the
          "scram" algorithm name. It mains advantage is that is does not
          rely on any external library, it is completely part of libdar.
	</p>

	<p>
          How does it works?
	</p>
	<p>
          Consider the pass phrase as a
          string, thus a sequence of bytes, thus a sequence of integer
          each one between 0 and 255 (including 0 and 255). The data to
          "scramble" is also a sequence of byte, usually much longer than
          the pass phrase. The principle is to add byte by byte the pass
          phrase to the data, modulo 256. The pass phrase is repeated all
          along the archive. Let's take an example:
	</p>
	<p>
          the pass phrase is "he\220lo" (where \220 is the character which
          decimal value is 220). The data is "example"
	</p>
	<p>
          Taken from ASCII standard we have:
	</p>
	<ul>
	  <li>h = 104</li>
          <li>l = 108</li>
          <li>o = 111</li>
          <li>e = 101</li>
          <li>x = 120</li>
          <li>a = 97</li>
          <li>m = 109</li>
          <li>p = 112</li>
	</ul>

	<code class=block>
          &nbsp;          e       x       a       m       p       l       e
          &nbsp;          101     120     97      109     112     108     101
          &nbsp;
          &nbsp;  +       h       e       \220    l       o       h       e
          &nbsp;          104     101     220     108     111     104     101
          &nbsp;
          &nbsp;  ---------------------------------------------------------------
          &nbsp;
          &nbsp;          205     221     317     217     223     212     202
          &nbsp;
          &nbsp;  ---------------------------------------------------------------
          &nbsp;  modulo
          &nbsp;  256 :   205     221     61      217     223     212     202
          &nbsp;          \205    \201    =       \217    \223    \212    \202
	</code>

	<p>
          thus the data "example" will be written in the archive as
          <code>\205\201=\217\223\212\202</code>
	</p>

	<p>
          This method allows to decode any
          portion without knowing the rest of the data. It does not
          consume much resources to compute, but it is terribly weak and
          easy to crack. Of course, the data is more difficult to retrieve
          without the key when the key is long. Today dar can also use
          strong encryption (blowfish and few others) and thanks to a
          encryption block can still avoid reading the whole archive to
          restore any single file.
	</p>



	<h2><a name="strong_encryption">Strong encryption internals</a></h2>

	<h3>Encryption per block</h3>
	<p>
	  To not completely break the
          possibility to directly access a file withing a <i>dar</i> backup,
	  the archive is not encrypted as a whole (as would probably do
	  an external program, like <i>openssl</i>). The
          encryption is done per large block of data. This way,
	  each block can be decrypted independently from the others,
	  and if you want to read some data somewhere
          you only need to decrypt the whole block(s) it is located in.
	</p>
	<p>
	  Such encryption block size can range from 10 bytes to 4 GB
	  (10 kiB is the default value). We will
	  design these as <b>libdar cipher block</b>, to differentiate with
	  what the underlying cipher algoritm uses as block.
	</p>
	<h3>CBC and IV</h3>
	<p>
	  Inside each libdar cipher block, <i>dar</i> relies on
	  the CBC mode (Cipher Block Chaining). In this mode, the data
	  of a libdar cipher block is split in small plaintext blocks (a few tens of
	  bytes as defined by the cipher algorithm). These plaintext blocks
	  are not ciphered independently, but the ciphering result is melt with
	  the next plaintext block before the ciphering operation.
	  The advantage of doing so, is that a repeated data pattern
	  does not lead to a repeated encryption pattern. However the first
	  plaintext block of a libdar cipher block is not melt with anything. To
	  cope with that an <i>Initial Vector</i> (IV) can be provided to 'randomize'
	  the encryption result, though one has to record what was the IV at time
	  of encryption to be able to decipher the data.
	</p>
	<p>
	  The <i>IV</i> for each <i>libdar cipher block</i> is derived from its number
	  inside the dar backup/archive and from the encryption key. The result
	  is that even if you have the same data in two libdar cipher blocks,
	  it will not result in the same encrypted data common to both (or to more
	  libdar cipher blocks)
	</p>
	<h3>Elastic buffer</h3>
	<p>
          An "elastic buffer" is introduced at the beginning and at the
          end of the archive, to protect against plain text attack.
          The elastic buffer size randomly varies and is defined at
          execution time. It is composed of randomly generated data
	  (using <code>gcry_create_nonce()</code>).
	</p>
	<p>
	  But to be able at reading time to determine the amount of random garbage
	  that has been put at the beginning and at the end of the archive by mean
	  of such elastic buffers, a small structure is randomly dropped inside
	  the garbage that tells the overall size of this random data:
	</p>
	<p>
	  Two marks characters '&gt;' and '&lt;' delimit a <b>size field</b>, which
	  indicate the byte size of the elastic buffer. The size field is
	  randomly placed in the elastic buffer. Last, the buffer is encrypted
	  with the rest of the data. Typical elastic buffer size range
	  from 1 byte to 10 kB, for both initial and terminal elastic
	  buffers.
	</p>
	<p>
          Elastic buffers may also be used inside libdar encryption blocks, when
	  the <i>libdar cipher block</i> size is not a multiple of the
	  the ciphering algorithm plaintext block size. Such elastic buffer
	  is added for padding at the end of the <i>libdar cipher block</i>.
	</p>

	<h3>Key Derivation Fonction - KDF</h3>
	<p>
	  Just above, we saw:
	</p>
	<ul>
	  <li>
	    how CBC inside a <i>libdar cipher block</i> avoided repetitive data
	    to be seen as repetitive encrypted data
	  </li>
	  <li>
	    how different IV could avoid exposing repetitive data between
	    <i>libdar cipher blocks</i> to be seen as repetitive encrypted data
	  </li>
	  <li>
	    We also see how <i>elastic buffer</i> randomly moved data structure
	    inside an archive to minimize code book attack.
	  </li>
	</ul>

	<p>
	  <b>First</b>, what if you use always the same password/passphrase between
	  different archives? Breaking one archive will let the attacker access
	  all archives or backups. Worse, an attacker has much more data
	  to work on and compute statistical decryption methods to get to its
	  goal.
	</p>
	<p>
          <b>Second</b>, directly using a human provided password would provide a
          weak encryption. Because in all the possible key values, humans choices lead
	  to a very restricted set of values: only letters, digits, space and eventually
	  punctuation characters, with a certain order at words level and
	  phrase level, even 5ubst1tu1n9 (substituing) letters by digits does
	  not bring much more randomness... Think that in the 256 different values
	  a byte can have we barely use 50 of them. Thing that in the may letter
	  combinaisons we have, we are far from using all of them (how many English words do you
	  know that contain 'xz'? ...Or even more unreadable larger sequences?)
	</p>
	<p>
	  To get one step further in securing your data, <i>libdar</i> makes use
	  of a <b>K</b>ey <b>D</b>erivation <b>F</b>unction (KDF):
	</p>
	<p>
	  Strong encryption uses secret key, to cipher/uncipher data, we
          will say "password" to simplify in the following.
	</p>
	<p>
	  To overcome this human weakness, the state of the art
          is to use a <i>Key Derivation Function</i>, which takes in input the
          human provided password plus a "salt" and outputs a stronger key
          (salt is described below). This function is based on a hash
          algoritm (sha1 by default, but this value can be modified since
          archive format 10/release 2.6.0) and an iteration count (2000
          before format 10 and 200000 interations since release 2.6.0 by default).
          The KDF mixes salt and passphrase then hash the
          result and applies the hashing algorithm on it repeatedly the number of
	  iteration count that has been selected, the final result is the <i>hashed password</i>
	</p>
	<p>
	  Note that as an
	  alternative to PBKDF2 (from PKCS#5 v2) which is the name of the KDF algorithm
	  we have just described, release 2.7.0 brings the modern <b>argon2</b> KDF,
	  which is the default, it has been built against libargon2 library.
	  Argon2 also makes use of a salt and an interation could, so what was described
	  above stays valid at a global view level.
	</p>
	<p>
	  So the KDF transform the well know human namespace of secret keys in a new namespace
	  for keys, not wider, but not that well known. Of course, anyone with the knownledge
	  of the salt, iteration count and algorithm is able to know it, but
	  <b>it costs a lot of effort in computing resource</b>! And this is the objective.
	  No need to hide or make secret the KDF parameters, when an attacker that
	  would requires hours or days to crack a password based on a dictionnary attack
	  now needs years or centuries to do so.
	</p>

	<h3>Choosing the salt</h3>

	<p>
          The <b>salt</b> is a randomly
          chosen value by <i>libdar</i> and stored in clear in the archive
          header, beside the iteration count and hash algorithm used for
          the KDF. Thus even if the user has the same password for
          different archive, the effective real key used for strong encryption
          will differ from archive to archive or from backup to backup,
	  making much more difficult
          for an attacker to crack an archive using statistical methods
          over a large number of archives sharing the same human provided
          password. And second point, a dictionnary attack is much more costly
	  to succeed, either you need hundred thouthand more time of hundred
	  thouthand more CPU power.
	</p>

	<h3>In summary</h3>

	<p>
	  In summary, <b>salt</b> randomize keys between archive, <b>elastic buffers</b>
	  randomize
	  data location inside the archive, <b>IV</b> randomize encryption between libdar
	  encryption blocks, and <b>CBC mode</B> randomize encryption withing a libdar encryption
	  block.
	</p>

	<p>
          Here follows a diagram of the way key, block number, cipher
          algorithm and Initial Vector (IV) interact together:
	</p>

	<code class=block>
	  &nbsp;           +------------------- [ algo in CBC mode ] -----------------------------> main key handle
	  &nbsp;algorithm -+                           ^                                                  |
	  &nbsp;           +---> max key len           |                                                  |
	  &nbsp;                     |                 |                                                  |
	  &nbsp;                     |                 |                                                  |
	  &nbsp;                     v                 |                                                  |
	  &nbsp;password ------> [  KDF  ] ------> hashed_key -------------+                              |
	  &nbsp;                     ^                                     |                              |
	  &nbsp;                     |                                     |                              |
	  &nbsp;salt ----------------+                                     v                              |
	  &nbsp;                     ^                              [ SHA1/SHA256 ]                       |
	  &nbsp;                     |                                     |                              |
	  &nbsp;hash algo -----------+                                     |                              |
	  &nbsp;                     ^                                     v                              |
	  &nbsp;                     |                                essiv_password                      |
	  &nbsp;iteration count -----+                                     |                              |
	  &nbsp;                                                           |                              |
	  &nbsp;                                                           v                              |
	  &nbsp;                                           [ Blowfish/AES256 in ECB mode ]                |
	  &nbsp;                                                           |                              |
	  &nbsp;                                                           |                              |
	  &nbsp;                                                           v                              |
	  &nbsp;. . . . . . . . . .                                 essiv key handle                      |
	  &nbsp;. Initialization  .                                        |                              |
	  &nbsp;. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .|. . . . . . . . . . . . . . . | . . . . . . . . . .
	  &nbsp;                                                           v                              |
	  &nbsp;libdar cipher block's number -----------------------> [ encrypt ] ------> IV -----+       |
	  &nbsp;                                                                                  |       |
	  &nbsp;                                                                                  |       |
	  &nbsp;                                                                                  v       v
	  &nbsp;libdar cipher block's data ---------------------------------------------------> [ encrypt/decrypt ] -----> data
	  &nbsp;
	</code>

	<h2><a name="asym"></a>Public key based encryption internals</h2>

	<p>
	  Dar does not encrypt the whole
          archive with a recipient's public key, but rather randomly chooses
          a (long) password for symmetric encryption (as seen above, except
          it does not need KDF and salt as it can select a much better
          random key than human can do), encrypts that password with the
          recipient's public keys (eventually signing it with your own
          private key) and drops a copy of this ciphered/signed data into
          the archive. At reading time, dar read the archive header to find
          the encrypted password, decrypt it using the user's private key
          then use that password, now in clear, to decrypt the rest of the
          archive with the adhoc symmetric algorithm.
	</p>
	<p>
          Signing is done on the randomly chosen ciphering key as seen
          above, but also on the catalogue (the table of content located at
          the end of the archive). More precisely the catalogue is
          sha512-hashed and the resulting hash is signed. The catalogue hash
          and its signature are stored encrypted in the archive right after
          the catalogue.
	</p>

	<h3>Why not using only asymmetric encryption end to end?</h3>
	<p>
	  First, for a given fixed amount of data, the resulting ciphered
          data size may vary. Thus sticking ciphered block together in the
          archive would not lead to easily know where a block starts and
          where is located its end. Second, doing that way allows an archive
          to have several different recipients, the password is ciphered for
          each of them and the archive is readable by any specified
          recipient, while they do not share any key. Doing that way has
          very little impact of archive size.
	</p>

	<h3>Why not using signing from end to end?</h3>
	<p>
	  First, this can be done without dar, just use gpg on the slices of
	  an archive, so there is no advantage in having such feature inside
          dar. The other reason is time, as signing the whole archive would
          be long. it would also be very painful to validate a given archive
          authenticity (time to read a whole archive), here we can leverage
          the fact the archive is encrypted (thus tampering the archive
          would be seen when unciphering it) and that dar uses of several
          mechanismes (compression, clear text CRC) to detect data
          corruption.
	</p>

	<h3>Security restriction when signing an archive for multiple recipients</h3>

	<p>
          For a multi recipient archive, any recipient has decode the
          signed and encrypted key. One could even reuses this same encryption key
          which may also be signed by the original sender to build a new
          archive with totally different content. Since 2.7.0 a warning is issued
	  when this condition is met, but the action is still allowed as it may make
	  sense when exchanging data between a set of people that trust each other,
	  but still need to validate the authenticity of the exchanged data
	  (over Internet for example).
	</p>
	<p>
	  But to exploit this security restriction the attack should still be difficult
	  as the table of content (the catalogue) should is also signed within the archive.
	  Touching any part of it would break this signature, thus the archive table of
	  content cannot be tampered if the signature algorithm is strong. This
	  <i>catalogue</i> contains the CRC of any file data, EA and FSA, as well as the
	  size of the compressed data. Thus the attacker may modify the file's data but must
	  manage for the resulting modification to still satisfy the CRC and have the same
	  lenght. Worse, if compression is used, the modification must not exceed the amount
	  of compressed data and must stay uncompressible by the same compression algorithm
	  that was used for that file in the original archive.
	</p>
	<p>
	  In short, if this security restriction exists and in exploitable in therory,
	  it is not so easy to realize it for a large portion of the archive.
	</p>

	<h2><a name="multi-threaded">Multi-threading in libdar</a></h2>

	<h3>Ciphering</h3>
	<p>
	  The ciphering in dar/libdar is performed by large block of data (see -# option
	  and consorts), it was
	  thus possible (added in 2.7.0) to parallelize the ciphering/deciphering operation assigning
	  different blocks to different threads without changing the archive format.
	  The measured performance gain is quite good, but usually hidden by the
	  compression load or disk I/O. More than 2 threads for ciphering/deciphering
	  does not bring much benefit to the
	  overall processing time.
	</p>
	<p>
	  To assign blocks
	  to worker threads (those that compute the ciphering algorithm), two helper
	  threads are needed: one that copy data into block of memory and scatters it to
	  workers for they can work in total autonomy, and another one that gather the resulting block of
	  memory in the correct order.
	<p/>
	<p>
	  If performance gain is present and no change in the archive format
	  was necessary to leverage multi-threading
	  for ciphering/deciphering, however, this has some impact on memory usage
	  required by dar. Each thread will need a pair of blocks (one for the clear
	  data, one for the ciphered equivalent). In addition, the gathering
	  and scattering structures of blocks to/from the workers (which we will globally
	  design under the term <i>ratelier</i> can hold each N block of memory
	  (N being the number of workers) in fact the size of these
	  structures has been chosen a bit wider: N+N/2. Last the helper thread that
	  feeds data to the worker may hold one pair of block under construction,
	  and the gathering thread may obtain from the ratelier_gather a whole structure (N+N/2 blocks)
	  to work with. For N worker, the number of memory block is N+N/2 for each ratelier
	  plus N for each worker, plus N+N/2 for the gathering thread and one for the scattering thread,
	  which makes around 7*N blocks of additional memory requirement. Concretely, having 5 threads with
	  encryption block of 10240 bytes (10 KiB) leads dar to allocate 350 KiB (7x5x10) of additional
	  memory compared to the single threaded implementation.
	</p>

	<h3>Compression</h3>
	<p>
	  For compression, things have been more complicated, as before release 2.7.0 compression was done
	  per file in <b>streaming mode</b>. In that mode <i>dar</i> has to provide a sequential flow of bytes to the compression
	  engine and at the same time retrieve the compression/decompression counterpart. The API provided
	  by the different libraries (libz, libbzip2, liblzo2, liblzma5, libzstd, liblz4) do not provide
	  multi-threaded approach for this mode of compressing data, however this gives the best compression
	  ratio you can expect for a given file, given compression algorithm and compression level.
	<p/>
	<p>
	  Before 2.7.0 parallelization had been tried but with little success by running compression and
	  encryption in different threads. Usually one of the threads spent most of its time waiting for
	  the other and the overall performance gain was little due to the added thread management and
	  synchronization overhead. This scheme
	  has been completely abandoned to the one exposed above for encryption. To extend this method
	  to compression, a new compression mode has to be added, to the legacy streaming one to:
	  the <b>block compression mode</b>, which is described below. Note that with release 2.7.0 both mode will coexist,
	  and you will still be able to read old archive and generate new one with the old single thread/streaming approach.
	<p/>
	<p>
	  The implementation is quite similar to the encryption approach, a set of worker proceed to the compression
	  or decompression of one block at a time and two helper threads are used, one feed on one side and the
	  other gather the results on the other side.
	<p/>
	<p>
	  The consequence is a slightly degraded compression ratio when the block size is chosen small. But with
	  larger blocks there is not noticeable compression ratio difference, though it has have an impact
	  on the memory usage. That's a balance to find between what memory mean you have
	  and want to use, how much time you have and want to spend for the compression and how much space you
	  have want to use to store the resulting archive.
	  Choosing a block size
	  of 102400 bytes (see extended syntax of -z option) with 4 threads (see -G option) will lead libdar to
	  allocated 2800 KiB (~ 2.7 GiB) in addition
	  to what libdar already uses for single thread approach (same formula as what is described above for
	  encryption). If you use both parallel encryption and parallel
	  compression at the same time you have to sum the additional memory requirements of each feature.
	</p>

	<h4>Command-line</h4>
	<p>
	  The thread number for both encryption and compression is driven by the -G option. See man page
	  for details. Note that, as described above, multi-threaded compression/decompression is only
	  possible with the new compression per block method, which means you have to specify a compression block
	  size (see -z option).
	</p>

	<h2><a name="block_compression_mode">Streaming versus block Compression mode</a></h2>

	<h3>Streaming compression</h3>

	<p>
	  Since its first day with version 1.0.0, dar/libdar provides streaming compression. It started with <b>gzip</b>
	  algorithm, has been extended later to <b>bzip2</b>, then <b>lzo</b>, <b>xz/lzma</b>, and since release 2.7.0
	  two new algorithms have been added: <b>zstd</b> and <b>lz4</b>.
	</p>
	<p>
	  Streaming compression let libdar provide new decompress data bytes and in return from time
	  to time get back some compressed data bytes from the compression engine. There is no limitation in size: a stream
	  of uncompressed data enters the compression engine from which exists another stream of compressed data.
	  You do not need very large memory buffers to feed and gather data to/from the engine,
	  and can obtain very good compression ratio. Though the process cannot be parallelized as
	  the stream of incoming data must be provided to this single engine in the expected order.
	  The process is similar for decompression, you give back to the engine the compressed data.
	  It does not matter if you feed the engine by smaller or larger groups of bytes than what
	  the engine provided when compressing, as long as you keep the byte sequence order, the resulting uncompressed stream will be the original
	  one you gave for compression.
	</p>
	<p>
	  Such streaming compression is natively provided by libz, libbz2, libxz/lzma and libzstd. So
	  dar store these compressed data as provided: an unstructured and arbitrarily long sequence
	  of bytes corresponding to the (compressed) data of a given file. Yes, dar compress file per
	  file: a new file, a new stream.
	</p>

	<h3>Block compression</h3>

	<p>
	  For lzo algorithm (added with 2.4.0 release) the liblzo2 library does not provide such streaming
	  interface, only a per buffer compression is available.
	  That is to say, you have to provide block of uncompressed
	  data from which you get a variable amount of compressed bytes. To decompress the data
	  you have to provide back this exact block to the decompression engine. If you sticked the
	  blocks together and do not provide them back at their original boundaries, the decompression
	  will fail.
	</p>
	<p>
	  For that reason dar use a structure inside the compressed file's data to record where
	  each block starts and ends, this way it is possible to feed the decompressing engine with coherent
	  compressed data and get back the original clear data.
	</p>
	<p>
	  The structure used is made of an arbitrary long sequence of blocks. Each block is constitued
	  of three fields:
	</p>
	<ul>
	  <li><b>a Type</b> which has two values, either <i><b>D</b>ata</i> or <i><b>E</b>nd of file</i></li>
	  <li>an <b>block size</b> which is zero for <i>end of file</i> block types</li>
	  <li><b>a sequence of compressed bytes</b> which length is given by the <i>block size</i> field</li>
	</ul>
	<code class=block>
	  +--------+-----------+---------------------+--------+------------+-----------------+-----+--------+----------------------+<br/>
	  | Type=D |block size | compressed data ... | Type=D | block size | compressed data | ... | type=E | block size = 0 (EOF) |<br/>
	  +--------+-----------+---------------------+--------+------------+-----------------+-----+--------+----------------------+<br/>
	</code>
	<p>
	  One thing you cannot guess from a given compressed data is how much uncompressed data it will
	  generate, while you have to prepare and allocate in memory a buffer large enough to receive it.
	  It has been decided to slice the file's data in uncompressed block <b>not larger than</b> the maximum
	  liblzo2 can handle, which is 240 kiB. So at decompression time, whatever is the size of the
	  compressed block we can provide a 240 kiB buffer to receive the decompressed data and repeat
	  the process as much time as necessary to process the whole compressed data of a given file.
	</p>
	<p>
	  With release 2.7.0 when adding lz4 compression algorithm, it has been found that the
	  "streaming API" of lz4 library was not really a streaming API as it requested to provide back
	  the exact compressed block and not a stream of bytes for the decompression to be possible. For that
	  reason, lz4 streaming is done the exact same way as what has been implemented for lzo streaming:
	  This is a block compression with 240 kiB block size.
	</p>
	<p>
	  The <b>block compression new feature</b> you have starting release 2.7.0 is thus just an extension of this method
	  by providing to the user the ability to select the maximum size of the uncompressed data block that
	  will be compressed at a time. The -z option has been extended to allow one to specify that block size.
	  If set to zero we stick to the streaming mode, else we use block compression.
	</p>
	<p>
	  But for a given algorithm (except lzo and lz4) streaming mode and block mode are not compatible,
	  the first has no structure (no block and cannot thus be processed in parallel) the second adds some
	  non-compressed bytes beside the compressed data (the block headers). For that reason the new archive
	  format contains the block size used at creation time in the archive header and trailer, which first
	  let libdar allocate buffer large enough to receive decompressed data and second use the correct
	  mode to retrive the
	  original data.
	</p>

	<h3>Multi threading</h3>

	<p>
	  As mentionned earlier, to provide multi-threading support for compression (added
	  at release 2.7.0) the use of per block compression is necessary. In the dar archive, this mode
	  of compression provides the same structured data of compressed blocks as described above for
	  lzo, but with the additional freedom to choose the block size (which stay fixed at 240 kiB for lzo and lz4
	  pseudo-streaming mode) and any compression algorithm supported by libdar.
	</p>
	<p>
	  But that is not enough to have per block structure in archive to have multi-threaded processing:
	  you have also to setup the code that read in advance, feed block in parallel to different threads,
	  having each their own compression/decompression engine set and ready for use, gather the
	  resulting work, reordering them if necessary before writing down to filesystem. The method used here is very similar to what was used
	  to add multi-threading suppor for encryption/decryption (see previous chapter)
	</p>
	<p>
	  Last, if multi-threading is not possible for real streaming compression, it is available in dar/libdar
	  for lz4 and lzo algorithm that are implemented as 240 KiB block compression. For other algorithm,
	  streaming compression use different API of the libraries libdar relies on from the block compression,
	  and is not possible to parallelize.
	</p>
      </div>
    </div>
  </body>
</html>