Re: lynx-dev cookie incompatibility

[David Woolley]

> Browser_?  AFAIK, MSIE does not send anything like "MSIE ...," always
> "Mozilla/ ... MSIE," so it almost appears to me it is dependent on two
> conditions "Mozilla" AND "MSIE."

And it does this because of discrimination in favour of Netscape in the
early days.  The part in brackets was never intended to be anything
more than a comment.

Sites that optimise for different browsers server side generally use a
published browsercaps database, so to get them to correctly characterise
Lynx you need to track down the suppliers of these databases, and get
them updated, and, of course, convince the site operators that they need
to maintain a current database.

These database pattern match the whole string, because of Microsofts abuse of
user agents in response to content providers short sighted use of it.

"Modern" sites use other tactics, like Javascript probes, or test cookies,
possibly with META Refresh.

Also, many accusations of active bias against Lynx are really errors of
omission (e.g. not replacing the "you need frames" <noframes> section of
a pattern page, with a proper alternative).

A very few enlightened sites (and there are very few enlightened sites!)
write universal HTML.  Unfortunately broken implementations of CSS,
particularly in NS4, make this difficult to do whilst meeting their
marketing people's glossiness expectations.  Remember that most people
who actually code HTML are doing it under the direction of someone who
doesn't.  Some of these realise they are abusing it, but know that there
is no point in complaining.  Many of them just do it for a job, and aim
to please their employers.  Many of the latter don't really understand
HTML, let alone the philosophy behind it, they simply cut and paste
examples from other "cool" sites.


; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to address@hidden