Teamwork
Posted: Sun Apr 20, 2014 2:01 am
We need to partition the work load equally and fairly.
I suggest the following way to accomplish that.
There are m members belonging to the Center.
(1) Have the Chair assign a unique identification number (ID) to each member ranging from 0, ..., m-1.
(2) Form the members into a ring with member i having a left partner, (i-1) MOD m, and a
right partner, (i+1) MOD m.
(3) Those groups form (overlapping) teams to work on issues.
(4) Have the Chair create m lists (one for each link between members) Eg. With 4 members they form a square
with 4 sides so there would be 4 lists. The lists are of length n DIV m where n is the number of current unresolved
BlackBox issues. The left over n MOD m items are scattered randomly amongst the m lists.
(5) Each member has two lists (left and right) to work on. The ith member has a left partner who works on the
the left list with i and has a right partner who works on the right list with i. So each member i is working on two issues at a time with two different members.
(6) As agreement is reached on how to solve an issue the issue, with its solution, is given to the Chair. The Chair
then collates the returned issues and forms them into a candidate release.
At the beginning of the process and after step (6) is when Josef's and Ivan's comments about how to actually keep track of these issues in git, etc, become relevant and is not addressed here in this topic.
This topic deals only with how to partition the work load of the Center and how to make it run smoothly.
-Doug
I suggest the following way to accomplish that.
There are m members belonging to the Center.
(1) Have the Chair assign a unique identification number (ID) to each member ranging from 0, ..., m-1.
(2) Form the members into a ring with member i having a left partner, (i-1) MOD m, and a
right partner, (i+1) MOD m.
(3) Those groups form (overlapping) teams to work on issues.
(4) Have the Chair create m lists (one for each link between members) Eg. With 4 members they form a square
with 4 sides so there would be 4 lists. The lists are of length n DIV m where n is the number of current unresolved
BlackBox issues. The left over n MOD m items are scattered randomly amongst the m lists.
(5) Each member has two lists (left and right) to work on. The ith member has a left partner who works on the
the left list with i and has a right partner who works on the right list with i. So each member i is working on two issues at a time with two different members.
(6) As agreement is reached on how to solve an issue the issue, with its solution, is given to the Chair. The Chair
then collates the returned issues and forms them into a candidate release.
At the beginning of the process and after step (6) is when Josef's and Ivan's comments about how to actually keep track of these issues in git, etc, become relevant and is not addressed here in this topic.
This topic deals only with how to partition the work load of the Center and how to make it run smoothly.
-Doug