Re: lynx-dev Re: Lynx interprets <a/> as </a>?

[Marco De la Cruz]

> Whether it's a bug or a feature/error-recovery, deponent knoweth
> not.  Definitely a bug if you're using lynx to validate markup,
> I guess, but that's probably not a good idea, anyway.

I don't use Lynx as a validator, but just as a reasonably
good indicator of markup. Although different browsers
handle the error differently, I thought that perhaps Lynx
would be consistent in its handling of closing tags
(admittedly I ignore the difficulties in determining this
type of behaviour, so I'm not about to criticize the
developers' decisions).

A few months ago the Deja News web-based email service
had a tremendous bug. Just before the list of emails
they had put some comment:

<!-- some comment --!>

and Lynx did _not_ assume the tag to be closed (it
still doesn't, AFAICT). The end result was that email
became irretrievable. This is of course no fault of
Lynx and I sent the people at Deja an email, asking
them if they could please delete *one* character (!)
from their buggy HTML. It was just one lousy character
in a comment, and I couldn't get them to do it. The
problem was finally "fixed" when they changed their
layout to their most recent monstruosity. The net
is sometimes a very sad (or funny, depending on how
you look at it) place indeed...

> For what it's worth, pro or con, IE4 doesnt seem to have any
> qualms about it, either.

I prefer to see the problem right away, even if it leads
to a mess as in Deja, but as people have correctly pointed
out this is what validators are for.

_________________________________
address@hidden
Gunnm: Broken Angel
http://128.100.80.13/marco/alita.html