Re: lynx-dev LYNX: need way to KILL a go-to while in-progress

[David Combs]

On Fri, Jul 09, 1999 at 03:20:15AM -0500, Klaus Weide wrote:
> On Thu, 8 Jul 1999, David Combs wrote:
> 
> > Like everyone on this list, I use a variety of keystroke-based tools:
> > mutt, emacs, trn, vi, ex, ..., lynx
> > 
> > 
> > Like in mutt, I hit "d" when finished reading an email.
> > 
> > 
> > So, here I am in lynx, and tired, and I hit "d", instead of "h"
> > (vikeys), and it starts asking do I accept cookies, and I curse,
> > hit ^G a few times, of course no effect, then I get the next
> > cookie question, ...
> > 
> > It would SURE be nice if there was a way to ABORT an entire
> > goto or d or whatever, to just ignore whatever internet-communication
> > protocol I've gotten myself in the middle of.
> 
> Did you try 'z'.  I few of those at the right time might have taken you
> out of it.  (but not at one of the cookie prompts.  You should have
> typed 'v' there.)
> 
> Of course you can change you key bindings as you like, if this happens
> often with 'd'.
> 
>    Klaus
> 

About that "v" for cookies:

You type in that, but then that link takes you to
another link, which asks the same question (this is
from my experience of a few months ago), and requires
a NEW 'v'.  And so on.

Now, you COULD just set something to ignore ALL cookies,
but as others have commented here, that might be overkill.

So, an idea.  Attach a vector of state-bit-words to parallel
the history list.  

Make one of the items "turn-off-refuse-cookies-when-you-get-back-here",
(or better, an unwind-protect), plus a new command to turn off
cookies for "here and deeper" (or maybe better, "here and deeper for
urls that have a prefix that matches this one").

You get the intent -- a level between totally-temporary and permanent.

Surely someone can take this idea further than I can, and
make a suggestion that can actually be implemented in the
existing lynx...

David