[c-nsp] VLSM

[Pete Templin]

Stephen J. Wilcox wrote:

> you are correct tho, the semantics is perhaps unimportant, but generally use of 
> incorrect terminology also means their understanding of the actual networking is 
> incorrect too.

Bingo!

When people assume they can have a /24 (formerly known as class C) 
automatically with a T1 because that's how the Internet worked years 
(decade?) ago, it becomes a problem.  (Or when they think that address 
assignments are linear with bandwidth, etc.)

When people think they can only fit four AS5248s into a /24 (think "four 
/26") and they need another /24 for their fifth, it's a problem.

Caveat: I'm well aware that plenty of technical reasons can explain the 
need to stick with 4xAS5248 or similar per /24.  I'm just pointing out 
the fact of ISPs who don't think to use /27 plus /28 for 48 modems on 46 
B channels of two PRIs.

pt