Re: [opensuse] RFR: Printing From Browsers

On Wednesday 20 August 2008 23:34, Brian K. White wrote:
----- Original Message -----
From: "Randall R Schulz" <rschulz@xxxxxxxxx>
To: <opensuse@xxxxxxxxxxxx>
Sent: Wednesday, August 20, 2008 3:53 PM
Subject: [opensuse] RFR: Printing From Browsers

Hi,

I am almost always frustrated by the results of printing HTML pages from
Firefox (I'm using version 2.0.x). The most common problem I encounter
is that text lines mean nothing to the pagination process and it is
common for a row of text to be sliced horizontally and split across a
page boundary.

Does anyone know of a browser that does well with printing?

I never heard of a browser that prints very well.
The obvious reason is html just isn't designed for that.
It's designed intentionally to flow to fill the shape of whatever
container it's in. (render as sensibly as possible in any size/shape
display device.) whereas documents intended to be printed are
written and formatted for a specific type of "display" (hard copy)
from the outset.

But that very statement about flowing to fit any window
makes me think that it should be no problem for a browser
or plug-in to simply re-render the page in an imaginary window
that has the same properties as a printed ***, (A4 or Letter
dimensions at 300 dpi, or maybe 100dpi for rendering the page
since thats what web designers assume screens are,
and then scale the result to 300 or just ask the printer
driver to do so)

In my own app I have to generate html specially for printing.
There are a few css attributes that can provide
the crudest of ability to make an html document that paginates
sort of ok. I have to have pre-formatted text (captured
print-to-file from an application) and then in html I specify
the font type and size, and wrap the text (or every x lines of
it) in pre tags, and apply css page-break-before to the pre tag.
Similarly, for scanned documents, I apply the same css
page-break-before attrib to an img tag and display the img
with 100% width and the height unspecified, which, as long as
as the image was a letter or a4 geometry, ends up printing
sensibly even though you can only see part of the image
actually in the browser window.

Since I have to both a) have content that I somehow
know is already pre-formatted well for printing,
and b) inject my own page breaks where I happen to
know they should go because of a). I don't see how
a random html that wasn't specifically written to
be printed can print very well.

<head>
<style type="text/css">
pre.ff {page-break-before: always}
img.ff {page-break-before: always}
</style>
</head>

<img src="pgtest01.jpg" width="100%" class=ff>

<pre class=ff>
[...]
</pre>

Years ago, there was a program for Windows 95 called Snap32. It would print
anything on screen, from a window, from a selected rectangle, from the whole
screen, and it would certainly print whatever kind of pdf or whatever.
(Just select the text on screen and print it.) It was pretty quick, too.

It came from a guy in Canada, who has long since removed the link and the
program, or maybe he died, I don't know. It was extremely simple and nice
to use. If anyone has this, and could reverse engineer it for Linux or
Windows, I would be happy to pay a reasonable fee for it. (For both
environments!) If anyone is willing to send me a copy (with password) for
Windows, I'd surely be interested. I lost mine in a disk crash.

--doug

Blessed are the peacemakers ... for they shall be shot at from both sides.
--A.M. Greeley
--
To unsubscribe, e-mail: opensuse+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: opensuse+help@xxxxxxxxxxxx