the following assumes that 3 or 4 shell scripts auto-creating issues etc is enough to tame Github for writing and reviewing papers.
Under the convention that authors known the names of their reviewers, then...
SOLUTION
- need to build links reviews <-> papers
_ SOLUTION_
- need to make the issues assigned to each reviewer seperate issues containing the review text
- create milestones for for 1jan16, 1feb16, 1mar16, etc
- assign due dates to those milestones (which can be reviewed first of each month
when "accepting" a paper, version N, for review, auto-create
- a directory for the corresponding author's last name
- a file in that directory called thatPaperTitleVN.md (camel case, no blanks)
- labelled "1.new"
- an issue report called lastname/thatPaperTitleVN/reviews
- with h1 title
- with h2 url
- with URL to lastname/thatPaperTitleVN
- with h2 entries for the review template
- that ends with a list of
- "must do" (if not done, cannot accept),
- "should do" (arguably, improve the paper),
- "might do" (optinal suggestions)
- editorial changes (typos)
- a file in that directory called thatPaperTileHistory.md
- with h1 title
- with h2 history
- with a list whose top item is a pointer to the issue lastname/thatPaperTitleVN/reviews
- with a list whose next item is a pointer to the file lastname/thatPaperTitleVN.md
then tell authors to write their paper into thatPaperTitleVN.md and relabelled in "2.submitted"
for all "submitted" papers, we must create
- one issue for each reviewer
- assigned to the milestone with its due date.
- assigned to that reviewer
- with a pointer to the vN file and the vN review issues
- NOTE: the actual reviewes are NOT written into this issue but into lastname/thatPaperTitleVN/reviews
- and the paper is labelled '3.underReview'