@@28=================================================================
Date: Tue, 03 Oct 1995 14:59:31 Est
Reply-To: "Pimienta Daniel " <[email protected]>
From: "Pimienta Daniel " <[email protected]>
To: [email protected],
[email protected],
[email protected],
[email protected],
[email protected],
[email protected],
[email protected],
[email protected],
[email protected],
[email protected]
Subject: CUNET user figures
X-Mulbri: 00**0000000000

Dear colleagues,

We already had the traffic figures provided by Archie (related to the months 10/94 to 2/95) and published in Randy's gopher. We have now more information on the subject: the user figures compiled by Saul and published within the frame of our on-line discussion.

I am providing a synthesis of both data and an attempt to conciliate two figures which should show complementary and coherent aspects of the same reality. The question of the definition of "active user" remains at the center of the debate and is obviously related to the traffic.

The column marked SOURCE 1 is the average of the traffic data of the 5 months from Archie's figure. The column marked SOURCE 2 is directly made from Saul's last mail and expresses the number of active users. The column marked COMPUT1 is the result of the estimation I did from the source 1 and is also published and documented in Randy's gopher. The last line is relative to the REDID network and serves a comparison factor. The last column, marked ADBU, is the result of a simple computation of the Average Bytes per Day per User and pretend to serve as a measurable factor for the data consistency.
ADBU= (Total monthly traffic/30)/ total number of active users.

Please note that the ABDU figure includes the headers of the messages sent or received. This field varies from a minimum of 256 Bytes to more than 1024 Bytes with an average above 512.

                         SOURCE 1   COMPUT1   SOURCE 2      ADBU
                         MBytes/Month  users    users  Bytes/user/day      
Antigua and Bermuda    AG    1.06        2         20       1767 **
Bahamas                BS    4.15        8        140        988 **
Barbados               BB    8.59       17         30       9544 
Belize                 BZ   13.74       28         80       5725 *
Dominican Rep CUNET    DO    3.62        7         70       1724 **
Granada                GR     .63        1          ? 
Guyana                 GY     .10        1          ? 
St. Lucia              LC   11.72       24         82       4764 *
Vincent and Gren.      VC     .79        2        150        176 ***
Suriname               SR    2.29        5        350        218 ***
Trinidad/Tobago        TT   17.14       34         50      11427 
----------------------------------------------------------------
TOTAL               CUNET   63.83      129        972       2189 **

***Dominican Rep REDID DO   60.00      120                 16667 

This  simple   cross-checking  exercise  brings  questions  about
consistency for some of the nodes figures (very strong inconsi- stency for values below 1000 -***-, strong inconsistency for values below 3000 -**-, and moderate for values below 6000 -*-).

It is possible that the inconsistency is the result of comparing traffic data from 2/95 with user data from 10/95. Has the traffic pattern for VC, SR, BS, AG and DO increased dramatically in the last 8 months? Archie can help us on that point. If yes, can we have the last months figures? By the way, Archie, is it possible to have published the listing produced by the USERS command from Waffle for each location? That would allow more accuracy on the information, and, I am hoping that everybody will agree that the publication of the list of active users in the Caribbean Research Networks will help enhance intra-regional communication. I will have REDID publish his users listing in a future mail.

Saludos,
Daniel Pimienta
.