[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: lynx-dev Re: t2thtml
From: |
Klaus Weide |
Subject: |
Re: lynx-dev Re: t2thtml |
Date: |
Sun, 28 Nov 1999 09:20:08 -0600 (CST) |
On Wed, 24 Nov 1999, Larry W. Virden wrote:
> What arguments are people finding of use when invoking t2thtml. I know
> that at one point this was suggested:
> DOWNLOADER:Preparse Tables:/path/to/bin/t2thtml maxwid=10 < %s >
> /tmp/foo.html; lynx /tmp/foo.html:TRUE
>
> But I see documented a number of other arguments and I thought I would
> ask what others have discovered.
I have mostly played with varying maxwid. Trying to make it narrow enough
to fit all columns within a screen width, but as wide as possible for
getting normal paragraph text displayed in a reasonable way. The optimum
varies by page, I haven't found the magic formula. Maybe t2thtml (and
t2tgrnest) could be made smart enought to figure out some of this without
help. (Or does t2tgrnest sometimes already do it? - I may have missed
it.)
nonest and vertstack are useful if too much display width would be
required because of nested tables, but I haven't tried them much.
I also used ucsgraphic (which I added in my copy), but that's more
a display nicety. (It only gives you something when using a PC code
page as Lynx display character set - althoguh in that case it is
quite nice. The original graphic setting is similar.)
I tried of course usetable (added by Hataguchi Takeshi) which relies
on TRST for post-processing and doesn't draw borders. It is more
sensitive to accidents (like table too wide, see maxwid), and when
the tabular rendering fails the results don't much sense.
Klaus