Re: Microsoft IE -- it just gets better and better

On Mon, 29 Jan 1996, Mike Meyer wrote:

> > On Mon, 29 Jan 1996, Mike Meyer wrote:
> > > Just specifying text/netscape-html probably works for 90% of the
> > I disagree, although I'd like to be able to agree.  If Netscape had
> > implemented this when they first added extensions to HTML, this would be
> > workable.
> 
> If NetScape had done that, we'd also have lots of real-world
> experience doing content negotiation.
> 
> > This is a stopgap solution, but it doesn't fill the 'gap'.  Old versions
> > of Netscape and other browsers would prevent this from working.
> 
> I agree completely. However, this argument applies to ANY stopgap
> solution. It also hinders a "complete" solution - content providers
> have to either deal with older browsers (including the NetScape 0.9
> I'm still getting hits from) or ignore them.

I personally would ignore 0.9. Out of 41,118 sessions in my current 
tracking database only 427 were from a version of Netscape 0.9.

0.9 is simply not even worth worrying about. 1.0 only marginally worth 
considering (*together* they are only about 4.3% of all sessions).

After despoofing MSIE, here is my top twenty (sorted only by 
manufacturer, not version). I don't rember what 'avberg@bgd.nl' is; 
possibly a broken bot that generated a bunch of bogus sessions a 
month or so ago while trying to do a depth first indexing.

Percent Browser                                              
 77.2%  Mozilla
 6.38%  Microsoft Internet Explorer
 4.67%  AOL
 2.22%  SPRY_Mosaic
 1.49%  PRODIGY-WB
 1.10%  NCSA Mosaic
 0.95%  avberg@bgd.nl
 0.92%  AIR_Mosaic(16bit)
 0.86%  IBM WebExplorer                       
 0.84%  Lynx
 0.82%  NetCruiser                   
 0.52%  NetManage Chameleon
 0.35%  GNNworks
 0.21%  Enhanced NCSA Mosaic
 0.14%  QuarterDeck Mosaic         
 0.11%  MacWeb
 0.10%  (Not reported)
 0.09%  Charlotte
 0.08%  eworldbrowser
 0.06%  Spyglass_Mosaic

The numbers were obtained via session tracking information on a site that 
counts only the 'first touch' - the first time a browser touched the 
site during a session. It also uses almost every trick I know to defeat 
caching.

-- 
Benjamin Franz

Received on Monday, 29 January 1996 18:47:55 UTC