Subject: Re: About last night From: Milwnews@aol.com Date: Thu, 7 Nov 1996 13:43:31 -0500
How the Web Was Won
Subject: Re: About last night
From: Milwnews@aol.com
Date: Thu, 7 Nov 1996 13:43:31 -0500
Sender: owner-online-news@planetarynews.com
Precedence: bulk
Status: O
X-Status:
Ben, I guess I see it a little differently and think that the nytimes
headline you referred to in your other post may indeed have been the right
spin on election night.
I can understand CNN and other online services proclaiming the night as the
"dawning of a new age" in reader awareness/reliance on the Internet, and
calling it a great success, but part of that is the sort of spin we saw the
political parties putting on mixed election results.
Two days later there also seems a high level of disappointment and
frustration from a great many people who didn't get in. That may be what
they'll remember most from the experience, time will tell.
The use of numbers also disturbs me, to wit:
In a message dated 96-11-07 11:29:39 EST, bcompaine@usa.net (B. Compaine)
writes:
<< According to an AP report, CNN's site peaked at 5 million hits an hour
between 9 and 10 p.m. EST, Over the course of election night, they
recorded 50 million visits -- and that's only the people who got through.
``The record for CNN on the main site on a single day had been 18 million
hits, on the day of the Iraqi bombing,'' according to executive producer
Mike Riley.>
Hits are not visits and "50 million visits" -- I can't believe AP would write
that. Several thousand simultaneous visitors can still bring major sites to
their knees at times. I think anecdotally we can also figure from experience
that one person trying to access 10 times is still one person.
Another instant figure I saw was the three million persons were using the web
for election information Tuesday night, which does not sound unreasonable to
me but I don't know anyone who can demonstrate at this point that this
couldn't have been one million very aggressive, frustrated persons trying
site after site.
I do think a number of people, unhappy that their local TV turned to local
elections when they wanted to know some national figures, signed online to
look, and many found ways to succeed by avoiding some of the heavier traffic
sites. The on-demand advantage of the Net vs the instant results of TV work
in the Internet's favor -- you have to wait till TV decides it's important
enough to tell you, online you have the menu of choices.
But for many the the info was there, the services did a fine job, but you
couldn't get it. Moreover, remember these sites were busy uploading as well
as dispalying, which adds to the problem.
>Although the long-haul backbone reportedly held up well, there were
reported slow-downs at many of the most popular servers. CNN had beefed up
its servers but still had periods when its machines were all busy, leading
to requests to users to try later.>
Slowdown is a vast understatement from what we're finding out. Many users
didn't get to that request-- their connect clocks just went round and round.
>Users of MSNBC found it often took several minutes to load a page. It
recored more than 3 million hits during the evening, with many users
digging into data on state races -- as might be expected as these were
less likely to be reported in the network coverage.>>
>From what I can determine, MSNBC was down for several stretches of the night,
for all practical purposes. And few of the sites really simplified their
entry pages so that less hits and less loads would get you past the front
door, which I think was a major mistake.
>"The T-1's out there are rocking,'' reported MSNBC's James Kinsella.>
Well, that sounds techy but there were T-1s that could have handled far more
volume if people had gotten to them. I'm thinking of all the people who
didn't get to that far off dial-ins -- they couldn't get past the modem banks
on their IP or at AOL. A lot of phone/Internet /online service traffic is
predicated on people coming on and people getting off to handle volume --
let's not even think what would happen if everyone in the country tried to
make a simultaneous phone call. Internet access suffered from not having its
usual relay effect.
> My personal experience was that The New York Times' site yesterday was
uncharacteristcially slow, which I attribute (unless told otherwise) to a
high volume after morning after usage. On the other hand, Boston.Com, which
had excellent local coverage, was perhaps even a bit perkier than usual
yesterday.>
Boston did a nice job and despite the hackers I got into nytimes twice, but I
think I was lucky.
>Kinsella's assessment was that ``This election night feels quite frankly
like the validation that TV got from the Nixon-Kennedy debates." Or more
likely, I think, the earlier boost it got from carrying the 1952
Presidential conventions. Back then, the proportion of households with TV
sets was closer to the proportion with Internet access today.>
Your sense of history is better than his, but in 1952 there were I believe
far more multiple viewers per set than there are viewers per computer screen
today, which helped galvanize the power of television. And whatever the
technical glitches, they were usually one-way to the viewer. The
selectivity-side of the Internet really changes the equation. I guess what
some people got was the 1990s equivalent of the 1952 TV test pattern :-).
>>Anyway, it seemed to be a good night for news and information online. It
probably introduced many occassional or new users to the value of online
for specific information needs. The next hurdle is for all the news sites
to track whether their routine daily volume reflects any residual carryover
from election night visitors.>>
That's my worry. A lot of the newer users, who didn't know the alternatives
out there, who signed in because major TV anchors displayed web addresses
(apparently each time they did they crushed the servers) and didn't have
their own lengthy bookmarks of alternatives -- these were likely the people
who were the most frustrated and bamboozled in the course of the night. And
I'm concerned about the residual carryover of that.
Dominique Paul Noth
Milwnews@aol.com
Internet Column -- Dom's Domain: Media Sites and Strategies
http://www.arcfile.com/dom
E-mail version: subscribe domslist at listserv@arcfile.com
+---------------------------------------------------------------------------+
Posted to ONLINE-NEWS. Made possible by Nando.net - http://www.nando.net
From MBLink_Administrators@ap.org Thu Nov 7 17:07:48 1996
Received: from janus.ap.org (eos.ap.org [165.1.6.1]) by cnj.digex.net (8.6.12/8.6.12) with ESMTP id RAA28081 ; for ; Thu, 7 Nov 1996 17:07:47 -0500