On 3 December 2014 at 17:15, Dave Crocker <dhc@xxxxxxxxxxxx> wrote:
So, after 25 years of effort, we've achieved 5% penetration. Wow.
Oh, I agree, it's pretty good, especially against a functioning incumbent.
Assuming you take the 1974 Cerf and Kahn paper as the startpoint for IPv4, which may not be entirely reasonable, then how long was it before one out of every twenty networked computers used it? 25 years would make that 1999. Hard to say what the percentages were, but I suspect that OSI was still significant, Netware had started the switchover from IPX/SPX only the year before, and Windows tended to use NetBIOS Frames (certainly the installed base did) - however overall, I seem to recall that IPv4 availability in businesses was fairly common at that point, so I'd be willing to bet it was higher than 5%.
However, I would guess it didn't pass 5% until toward the end of the '90's. Prior to Windows 95, getting IPv4 support on a typical PC required considerable technical ability, and prior to 1996 or so Win95 didn't install IPv4 by default, so I doubt the numbers were significant at all in percentage terms.
So if we pretend - probably incorrectly - that Win95 OSR2 caused an immediate rise above 5%, then it's 22 years for IPv4.
In any case, it's essentially in the same ballpark.
Dave.