URLs: Mean time of survival?

[petworth at suba.com]

Sorry for posting this to the list, but this offlist reply to Dr Kuhn below
was returned to me.


Bill Thayer


   ----- The following addresses had permanent fatal errors -----
<Heinrich at library.berkeley.edu>

   ----- Transcript of session follows -----
... while talking to library.berkeley.edu.:
>>> RCPT To:<Heinrich at library.berkeley.edu>
<<< 550 <Heinrich at library.berkeley.edu>... Addressee unknown
550 <Heinrich at library.berkeley.edu>... User unknown

Final-Recipient: rfc822; Heinrich at library.berkeley.edu
Action: failed
Status: 5.1.1
Remote-MTA: dns; library.berkeley.edu
Diagnostic-Code: smtp; 550 <Heinrich at library.berkeley.edu>... Addressee unknown
Last-Attempt-Date: Mon, 28 Jul 1997 05:42:03 -0500 (CDT)

To: Heinrich at library.berkeley.edu
From: <petworth at suba01.suba.com> (Bill Thayer)





>    Does anyone know about any statistical information
> on the "mean time of survival" of URLs and electronic
> documents on the web in relation to a) type of ressource,
> b) status of author, c) subject?


Alas no, but I have a purely anecdotal opinion from nearly 2 years spent
running RomanSites-L (for info about RomanSites-L, see end of message),
which involves visiting maybe an average of 100 sites a day, i.e., maybe
500 or 600 pages=URLs a day:

1. Within the narrower field of "things relevant to Roman antiquity", I've
noticed almost *no* correlation with any of the 3 items above, with the
single exception of students and academic years or the end of their
university career. Even then, good sites get kept at the same URL -- by
someone -- after the student leaves, even if they are unmaintained.


2. OTOH, strongish anecdotal correlation with (a) the quality of the site;
(b) the position of the URL in the tree.

   Good sites stay longer; even when there is a change of server, the old
server -- often without the original owner's knowledge and sometimes even
without their consent (I have correspondence to back this up) -- keeps the
site where it is.

   The home page won't often change address; but pages several levels down
change frequently.



I'd be interested in any general conclusions you might find, BTW!


Best,

Bill Thayer
ListOwner, RomanSites-L

---------------------------------------------------------------------------
For RS, see for example:

Roman and Apicius Sites on Internet
http://www.bahnhof.se/~chimbis/infocentre/r-a-sites.htm
(see both head & foot of this small page)

Pedar Foss's ROMARCH at
http://www-personal.umich.edu/~pfoss/thayer.html
(sample of RS, selected & put up by Dr Foss, now very old)

Alessandro Cristofori's Rassegna at
http://ecn01.economia.unibo.it/dipartim/stoant/rassegna1/intro.html

Internet Ancient History Resource Guide  at
http://allserv.rug.ac.be/~kverbove/IAHRG/e-gettin.html

Jacques Poucet's Bibliotheca Classica Selecta at
http://www.fusl.ac.be/Files/General/BCS/GateSp.html

America OnLine's summary, FWIW, at
http://ifrit.web.aol.com/mld/production/yiaz0719.html

---------------------------------------------------------------------------
end