Re: IP: Pac Bell says Net use may collapse phone system

["Fred R. Goldstein" <fgoldstein@bbn.com>]

  At 12:44 PM 10/28/96 -0500, Dave Reid wrote:
  
  >1 residential line consumes 3 CCS on average, while 1 ISP line
  >consumes 36 CCS on average. 
  
  Not correct.  36CCS is the theoretical limit.  In order to return <10% busy
  signals, the ISP generally has to stay below around 90% busy-hour fill (thus
  around 33 CCS), and in practice, since capacity is added in chunks, it's
  lower than that.
  
  >... Should an ISP pay ten times
  >as much for a line than a residential customer? 
  
  Here's the question:  Who pays for a call, the caller or the callee?  Are
  local calls SENT-PAID or COLLECT?  If the ISP is receiving the call, then
  isn't the norm SENT-PAID?  In that case the ISP should pay less, because the
  ISP is originating zero and the resi subscriber is responsible for the cost
  of the call!  If the ISP wants an 888 number, then they should pay!
  
  >Should a residential
  >customer pay more if they are having long holding times? Why should a
  >normal voice only residenital customer pay more to subsidize long
  >holding pattern callers? 
  
  You want long holding times, look at a teenager.  The only philosophical
  question you're legitemately raising is the one of flat-rate vs. measured
  residential usage in general.  I think the market, if it existed, would
  speak to that:  Users prefer flat rate.  It has to get damn expensive per
  month before users voluntarily bend over for measured.  Also, data users
  aren't all long hold times.  This debate began over ISDN, where it's much
  easier to do time-cutting (with fast call setup), yet telcos discourage it
  in favor of analog/modem, which means long hold times instead of time-cutting!
  
  >Cleary there needs to more emphazize on a user
  >pay structure to allow greater investment in switching plant to occur.
  
  So if you're losing money, raise your monthly rate!
  
  >ISP's that underprovision thier lines and oversell their accounts (no one
  >does that right?) cause another headache for the telco. Everytime a caller
  >gets busy, they call back again and again and again. This creates large volumes
  >of uncompleted calls all of which consume switch resources. When an ISP's
  >incoming line pool closely matches what their requirements really are their
  >impact on the switching fabric is reduced.
  
  True to some extent.  But it's often the telco's fault:  PRI installation
  lead times are measured in months, sometimes years.  We've had wicked busy
  incoming hunt groups sit months and months while waiting.
  
  >Seems to me we need more cooperation between the telco and ISP communitty. Both
  >need each other. The ISP's can't currently survive without the Telco's
  lines and
  >the Telco's need the extra renevue generated by the growth in local loop
  >utilization
  >that occurs both at the ISP end and the residential end. 
  
  Possibly true.  ISPs are somewhat dependent (some, very) on telcos.  Telcos
  however seem not to want the ISPs as subcribers, period.  Most don't want
  the revenue.  It ain't their old familiar voice business.
  ___
  Fred R. Goldstein      fgoldstein@bbn.com  
  BBN Corp.              Cambridge MA  USA    +1 617 873 3850