Quesitons f) and g Timestamp concurrency control without/with  the Thomas Write Rule

Does anybody ever reach to the point for that you have to use those rules? The conditions that I have in both cases(i.e. S1 and S2)  used the other options for validation. Just I want to make sure if my resolution is right or wrong? Thanks!

I initialize Timestamps for X, Y, Z first to Zero . Timestamps for T1, T2, T3 is set to 1,2,3 respectively. We have these sequences

Sequence S1: T1: R(X), T1: W(Y), T2:R(Z), T1:Commit, T3: R(X), T3:W(X), T3:Commit, T2: W(Y), T2:W(Z), T2: Commit

Sequence S2: T1:R(X), T3:R(Y), T3:W(X), T2:R(Z), T2:W(Y), T1:W(Z), T1:Commit, T3:Commit, T2:Commit

In Sequence S1 all Writes are being allowed, (condition TS(T) > WTS(O) is true)

In Sequence S2  We have these writes:

T3:W(x) , for which TS(T) > WTS(O) and TS(T) > RTS(O), so write will be allowed. (3>1 & 3>0)

T2:W(Y) is Aborted because: TS(T) < RTS(O)  (first condition), and TS(T) > WTS(O), so we never get to the second condition that Thomas Write Rule can be applied.

I checked my resolution many times and I think I did it write. With thanks.

-Gayaneh