lynx-dev LYNX: mode-line: say if was html or not.

[David Combs]

This also relates to changing the width of the displayed
text, a topic of a week or two ago.

To review, I like looking at www-docs wide, 132 wide.

But when I want to p(rint), to then download to my computer,
and print IN TWO COLUMNS, 9point, I want it narrow.

The question is how to get it narrow, eg 80 cols?

(1) At "getting" time, ^z out of lynx, set shell's cols 
to 80, fg back into lynx, and ^R, to redisplay, reparse
the html.

    Works like a charm -- underlining still there, done right,
etc.

    But HORRIBLE if the actual text is NOT html -- is actual
text, maybe 90 cols wide.  Because here setting shell's cols
to narrow, with a fg and ^R in lynx, seems to "wrap" the
text at that point, EVEN IN THE MIDDLE OF A WORD.

(2) run it through par, or fmt, or emacs M-Q, etc, to get
it re-"filled".  Nice when just a bunch of simple paragraphs,
but loses when there's lots of syntax, as in html.

---

Clearly, if html, you want to let lynx do the work; equally
clearly, if NOT html, you do NOT want lynx to do it -- let something
ELSE do it, something that knows more about paragraphs and filling
and words and maybe even hyphens.

One idea would be for lynx to ship along with par or to even just
use "fmt" (on unix, in mkstools on pc, or probably gettable for free
from somewhere, like gnu), and then it could do "the right thing" --
once it has parsed through the text, lynx KNOWS what type of doc
it is, REGARDLESS of the file extension.

---

Anyway, these are just ideas, to be modified by comments by
others.

David