Re: LYNX-DEV patches (numbered forms + highlighting)

[Laura Eaves]

> Date: Wed, 7 May 1997 12:18:21 -0500 (CDT)
> From: Klaus Weide <address@hidden>
>...
> > Are there reasons I'm not aware of for not wanting to incorporate it?
> > The only little nit I see so far is that when you use the list command
> > in this mode, the numbers on the listed URLs may not match the link numbers
> > in the original document (if that document contains forms).  Is this a real
> > problem?  (There are klugy ways to fix this...  But I don't want to spend
> > time on it for nothing.)
>
> I think it should be fixed.

Well, "fixed" could be interpreted in different ways...  What do you think
it should do?  The list command is only intended to list links.
If a numbered form field is encountered, it could generate a link like
        * [123](form input field)
with an href back to the original doc.  You could even generate more info
about the field -- such as "text entry" or "checkbox".  (I don't think it
would be a good idea to copy the whole form to the list page so it could be
activated there.  For one thing, there's no context info.)
In any case, this would ensure that the link numbers are all the same as on
the original doc, so the user could go back to the original doc and type 123g
to see the context for that link/form input.

> I should also incorporate Fote's latest changes for numbering (but in a
> way that the old behavior is still there if one wants it), and that would
> likely interfere in some way with your patches.  So I prefer to look at
> your patches after that, and that at least for now you continue to
> maintain your patches separately (since you seem capable and willing to 
> do that.).

If you're referring to Fote's patch to remove and "unnumber" invisible links,
that does not affect my code at all (from what I gather from his posts).
I don't touch that part of the code.
Anyway, I'm against this patch not because it affects my code, but because
I think it's a bad idea.
I'm still running into pages with invisible links that appear to be the result
of parser actions.  (In fact, I'm getting quite different results on some
pages, such as different ISMAP/USEMAP/LINK info, and am not sure in some cases
which parser is doing the "right" thing.)
And even without the parsing issue, I still want to see numbers for invisible
links if I go to the trouble of turning on link numbering.  I consider that a
feature of link numbering.
--le

;
; To UNSUBSCRIBE:  Send a mail message to address@hidden
;                  with "unsubscribe lynx-dev" (without the
;                  quotation marks) on a line by itself.
;