User distribution by browser
Earl Young
eayoung at bna.com
Wed Jan 29 07:31:08 EST 1997
There was a joke circulating about fifteen years ago that the reason
God created the earth in only seven days was He didn't have to worry
about an installed base. The issue of how far back to anchor your
software has been going on for a long time.
The reach of the Internet exceeds anything those of us who hacked on
the old ARPA/DARPAnet ever imagined. Were browsers more expensive -
and they are nearly free these days - a case could be made that more
attention should be paid to users and systems that are still
character-based. But its hard to imagine that anyone doing serious
work doesn't need everything that we can provide. Consequently, I
skew toward building toward a Netscape 3.0 world. I normally leave
the glitz out - the ShockWave and such - because I've not seen
evidence that those items add value to reference content. They do
consume bandwidth, and bandwidth - more than browsers - is the choke
point on the Internet.
I choose to design toward Netscape instead of IE because (a) we have
lots of Unix systems, and IE doesn't run on Unix, and (b) anyone
competing with Microsoft is doing the user community a favor. I am
not an enemy of Microsoft - and I use some of their software. But
power corrupts, and all that, and they seem to do better work in the
markets where they still have to scramble.
One encouraging trend is the "high-width/low-width" capability on some
of the site creation tools. I build mostly intranet applications, and
bandwidth is as important inside the firm as it is outside. I prefer
giving our users the option to (a) see the cute stuff, or (b) get to
the data faster. This has the effect of providing access to older
browsers, but it doesn't require that I avoid trying new tricks just
because someone or some institution hasn't caught on to the
technology. I don't believe we should reduce the functionality of the
site to everyone just because some people don't find the information
important enough to buy the tools necessary to secure access. Thus I
design for bandwidth because it make sense - and allow older browsers
in as a side-effect - but otherwise the folks happy to live in the
late-80's never cross my mind.
Earl Young
______________________________ Reply Separator _________________________________
Subject: Re: User distribution by browser
Author: henkm at python.konbib.nl at INTERNET
Date: 1/29/97 5:55 AM
Hi
On Tue, 28 Jan 1997, John J. Burke wrote:
> really have any idea what level or type of browser the majority of Net users
> have? I often see comments to the effect that "well, no one uses Lynx
anymore"
> or "Now that everyone has moved to Netscape 2.0 or beyond . . . " I know that
> my institution is behind the times (we still are predominantly Lynx-only), but
> what about the Net as a whole? While any of us can construct local resources
> based on the local dominant browser, should we all feel the need to play to
the
> lowest common denominator? (My personal view is yes, but I'd like to know
> others').
Analyzing your logfiles with ANALOG
(http://www.statslab.cam.ac.uk/~sret1/analog/)
will get you some info on the types of browsers/spiders hitting your site
at the National Library of the Netherlands (about 65% of our hits are from
the .nl domain, the rest is from all over the planet)
Just to give an indication,
here's the distribution of total reguests for december 1996 of our
(bilangual) site (http:\\www.konbib.nl)
(Mozilla = Netscape)
#reqs: browser
------ -------
204260: Mozilla
41609: Mozilla (compatible)
29381: MOMspider
3784: NetCarta_WebMapper
2328: Ilse-robot (wiebe at il.ft.hse.nl)
1646: Mosaic
1266: WebSeer
1225: ArchitextSpider
1117: Scooter
1111: MuscatFerret
1083: Lynx
761: Microsoft Internet Explorer
686: NetAttache Light 1.1
489: MSProxy
380: IWENG
328: Lycos_Spider_(T-Rex)
322: Webinator-bow
289: IBM-WebExplorer-DLL
250: Gulliver
235: Cyberjack Web 7.00 (Windows)
It gives an indication on how important other browsers are beside
Netscape.
BTW, our stats are not publicly accessible.
Groet, Henk
_
----------------------------------- o ---------------- / \--- _/\_-------
/<_ /\ / \_/ \
_=\--\_ / \_/ \
Henk Matthezing_________________(_)/ /(_)__/ \
'A King, Riding on Random Water'
coordinating editor Virginia Woolf - The Waves (orig. Manuscript)
Redactie Infoservices
Koninklijke Bibliotheek/SURFnet B.V. Phone: +31 70 3140 615
P.O.B. 90407 Fax: +31 70 3140 501
2509 LK Den Haag E-mail: henkm at python.konbib.nl
--------------------------------------------------------------------------
More information about the Web4lib
mailing list