1997-07-31 - Re: non-censorous spam control (was Re: Spam is Information?)

Header Data

From: “William H. Geiger III” <whgiii@amaranth.com>
To: Kent Crispin <kent@songbird.com>
Message Hash: d65a1219b8ebd3aac55b106a495a24ed436708dcccf446ff820f189df7e729a7
Message ID: <199707311751.MAA21324@mailhub.amaranth.com>
Reply To: <19970731095523.04336@bywater.songbird.com>
UTC Datetime: 1997-07-31 18:04:55 UTC
Raw Date: Fri, 1 Aug 1997 02:04:55 +0800

Raw message

From: "William H. Geiger III" <whgiii@amaranth.com>
Date: Fri, 1 Aug 1997 02:04:55 +0800
To: Kent Crispin <kent@songbird.com>
Subject: Re: non-censorous spam control (was Re: Spam is Information?)
In-Reply-To: <19970731095523.04336@bywater.songbird.com>
Message-ID: <199707311751.MAA21324@mailhub.amaranth.com>
MIME-Version: 1.0
Content-Type: text/plain


In <19970731095523.04336@bywater.songbird.com>, on 07/31/97 
   at 09:55 AM, Kent Crispin <kent@songbird.com> said:

>>It should be noted that the Bandwith issue is a red-herring.

>However, I think your argument here is faulty, because bandwidth is in
>fact oversubscribed -- the whole system depends on each end subscriber
>not using all their bandwidth all the time.

>> The bandwith
>>of the USENET has been *PAID IN FULL* by every subscriber to an ISP. The
>>ISP customers pay for their connections to their ISP who in turn pay for
>>their connections to the Access providers who inturn pay for the Backbone.
>>The PIPE has been paid for what goes over it not an issue. If all I want
>>to do with my T1 connection is ship *.jpg files via ftp 24/7 that is no
>>ones busines but my own.

>Not really.  A T1 line, for example, can handle maybe 40-50 28.8 modems 
>going full blast, but a small ISP over a T1 might have 200 customers.  
>This goes right on up the line -- at every level bandwidth is 
>oversubscribed, and successful operation of the net depends on  certain
>statistical usage patterns.  So, while it isn't written down  in a
>contract anywhere, what you are really paying for is peak  bandwidth, not
>sustained bandwidth.

Well the ISP may have 200 customer for the T1 line but they woun't have
200 dial up lines per T1 (at least not one that wishes to stay in business
long). Now how many dial-up lines per T1 a ISP will have will depend on
the traffic analysis for his customer base. There is nothing wrong with
oversubscribing his bandwith because he knows that all his cutomers will
not be on-line all the time using 100% of their 28.8 dial-up bandwith.
What an ISP does have to provide for is enough bandwith to be able to
handle the amount of dial-ups he has available. If an ISP has 200 dial-up
lines then he best provide enough T1's to be able to support them.

The same is true for Access providers. If an access provider is servicing
20 T1's then he best have the bandwith to the backbone to provide the
bandwith that he has sold. He is collecting the $$$ to provide the service
he is obligated to provide it.

Now if an access provider does detailed analysis of his traffic and
determins that he needs only 4 T3's to provide service for 20 T1's and
therefore reduces his costs that's fine. But if one of his T1 customers
traffic increases he is obligated to add more bandwith on his end to
handle it.

This is what the whole bandwith issue comes down to. ISP & Access
providers atempting to maximise profits for given resources. This doesn't
nullify their obligations to their customers. If they sell T1 bandwith
24/7 to their customers then they are required to provide that service if
their customers demand it. The current movement to blaim users for using
the resources that they have been sold is wrong. It is no different that
if a car dealer sells you a 100,000 mile warrenty on a car then renigs on
the contract because he really didn't expect you to drive 100,000 miles
with it.

- -- 
- ---------------------------------------------------------------
William H. Geiger III  http://www.amaranth.com/~whgiii
Geiger Consulting    Cooking With Warp 4.0

Author of E-Secure - PGP Front End for MR/2 Ice
PGP & MR/2 the only way for secure e-mail.
OS/2 PGP 2.6.3a at: http://www.amaranth.com/~whgiii/pgpmr2.html                        
- ---------------------------------------------------------------

Version: 2.6.3a
Charset: cp850
Comment: Registered_User_E-Secure_v1.1b1_ES000000