I am testing the retraining function of Z12 systems, and observe some strange behavior. I am using Zorro S 2.30 (Verbose = 3), the history data are from Zorro website (.t6 files from 20100101 to 20200229).

The end time of some of the Sample cycles seems wrong (1899-12-30). I am not sure if this indicates some bugs or not:
==========
WFO: Z12 2010..2020
Walk[18] Bar 10496..5285


Sample cycle 2
EUR/USD: 0..15782, 2010-01-04..2020-02-28

Sample cycle 3
EUR/USD: 0..22247, 2010-01-04..1899-12-30

Sample cycle 4
EUR/USD: 0..22247, 2010-01-04..1899-12-30
Loop[1][1] p1 step 1: 20.00 => 0.82 120/80
EUR/USD: 0..15782, 2010-01-04..2020-02-28

Sample cycle 2
EUR/USD: 0..15782, 2010-01-04..2020-02-28

Sample cycle 3
EUR/USD: 0..22247, 2010-01-04..1899-12-30

Sample cycle 4
EUR/USD: 0..22247, 2010-01-04..1899-12-30
Loop[1][1] p1 step 2: 26.60 => 0.88 112/78
EUR/USD: 0..15782, 2010-01-04..2020-02-28
...
...
==========