Re: lynx-dev 2.8.2dev.21 page count broken?

[Bela Lubkin]

Johnny Tevessen wrote:

> Using the abovementioned version of Lynx and pointing it
> to http://www.snafu.de/~dbrueg/detebe.html#wanderer on a
> 100x37 console, lynx claims that the page is 4 screen
> pages long ("p1 of 4"). Well, scrolling down, it then
> says "p2 of 4" and then "p4 of 4". Hmm. :)
> 
> It seems like something is counted wrong.

On a 100x37 screen, the document is more than 3 and less than 4 pages
long, with the anchor "#wanderer" near the end of page 1.  Below is a
diagram of what you see with the fragmentless URL (detebe.html) and with
the fragment (#wanderer):

    detebe.html
   +------------+
   |     p1 of 4|
   |            |       detebe.html#wanderer
   |            |          +------------+
   | #wanderer  |          | #wanderer  |
   +------------+          |     p1 of 4|
   |     p2 of 4|          |            |
   |            |          |            |
   |            |          +------------+
   |            |          |     p2 of 4|
   +------------+          |            |
   |     p3 of 4|          |            |
   |            |          |            |
   |            |          +------------+
   |            |          |     p4 of 4|
   +------------+          |            |
   |     p4 of 4|          |            |
   |            |          |            |
   +------------+          +------------+

The full document has four pages on your display size.  Imagine "page-
end marks" at the end of each screen page of the full document.  The
*displayed* page number is the number of the last page-end which appears
on screen.  In this case, the third screen page contains the page-ends
of pages 3 *and* 4.

The display might be clearer if Lynx called the screen pages of the
fragment:

  "bottom of p1 + top of p2 (of 4)"
  "bottom of p2 + top of p3 (of 4)"
  "bottom of p3 + all of p4 (of 4)"

-- but this seems like unnecessary clutter.

>Bela<