This question has TWO parts.

a$.$ To prevent deadlocks in Maekawa$$s Mutual Exclusion algorithm someone proposes the following modification and calls it the Aunt$-$Mae$-$KawaSpiderman algorithm. In the $$enter$()$ function, a process pi instead of multicasting a request to processes in Vi$,$ instead orders Vi$$s processes in increasing order of ID$,$ and sends them Request messages sequentially, waiting for each Reply before sending the next Request message. That is$,$ pi sends a unicast Request to only the lowest ID process in Vi$,$ then waits for a Reply message from it$,$ and only then does it send a Request message to the second$-$lowest ID process in Vi$,$ waits for a Reply message from it$,$ and so on until the last voting set member. The rest of the protocol remains unchanged and same as Maekawa$$s algorithm. You can assume no failures happen in this system. Does this modified Aunt$-$Mae$-$Kawa$-$Spiderman algorithm: i$)$ still satisfy safety? ii$)$ prevent deadlocks? For each, give a proof or a counter example.

b$.$ For mutual exclusion, The Kingpin $($a frequent opponent of Spiderman$)$ has implemented the original Maekawa Mutual Exclusion algorithm, but to $$make it more efficient$($a dangerous phrase, if ever there was one!$)$ they have $$optimized$$ the voting set members similar to how Ricart$-$Agrawala works. In other words, when a voting set Vj$$s member pi receives a Release message from a process pj $($which is exiting the critical section$),$ the voting set member process pi now immediately multicasts a Reply message to all waiting processes in pi$$s queue, and empties the queue. Does this new Kingpin algorithm work $($i$.$e$.,$ guarantee safety$)?$ If yes, give a proof. If not, give a concrete scenario where this algorithm violates mutual exclusion $($in particular, safety$).$