Re:The End of Gmane?
by dredmorbius on Thursday July 28, 2016 @10:56PM
(#12184790)
I've shot around estimates of the size of Usenet with some old Usenetters (of whom I was one, though only marginally active).
Gene "Spaff" Spafford thinks 50k - 500k users ~1988 - 1994 is probably ballpark accurate. Peter da Silva notes that in 1990 he was the most active user on Usenet ... and under another handle, the 3rd most active. Microsoft did some early 2000s research turning up low-single-digit million distinct user IDs IIRC (I've posted on this previously).
So, yeah, it was small.
Organising people into groups, even ad-hoc, from which exceptionally good messages might be shared to others, might be one way of handling scaling. Clustersizes ranging from a total size within Dunbar ranges, to a 1/9/90 split giving the 10% active users a Dunbar cap (about 300 - 3,000 users per cluster) might work.
I agree that 1990 Usenet in 2016 would be a fatal mistake.
Gene "Spaff" Spafford thinks 50k - 500k users ~1988 - 1994 is probably ballpark accurate. Peter da Silva notes that in 1990 he was the most active user on Usenet ... and under another handle, the 3rd most active. Microsoft did some early 2000s research turning up low-single-digit million distinct user IDs IIRC (I've posted on this previously).
So, yeah, it was small.
Organising people into groups, even ad-hoc, from which exceptionally good messages might be shared to others, might be one way of handling scaling. Clustersizes ranging from a total size within Dunbar ranges, to a 1/9/90 split giving the 10% active users a Dunbar cap (about 300 - 3,000 users per cluster) might work.
I agree that 1990 Usenet in 2016 would be a fatal mistake.