Re: flexfax: html to Postscript.....

[Peter Stamfest <peter.stamfest@eunet.at>]

On Mon, 8 May 2000, Andreas Kostyrka wrote:

> On Mon, May 08, 2000 at 09:22:20AM -0400, Aidan Van Dyk wrote:
> > On  8 May, Andreas Kostyrka wrote:
> > > On Mon, May 08, 2000 at 06:15:37AM +0200, Peter Stamfest wrote:
> > >> On Sun, 7 May 2000, Brendin Emslie wrote:
> > >> 
> > >> > Hi,
> > >> > 
> > >> > I have used html2ps and htmldoc to convert HTML to Postscript.  html2ps
> > >> > works better thatn htmldoc but neither work as good as loading the page in
> > >> > netscape and saving it as postscript.  Is it possible to save an html file
> > >> > as a postscript file using netscape on the command line?  Can anyone suggest
> > >> 
> > >> http://home.netscape.com/newsref/std/x-remote.php
> > >> 
> > >> But last time I checked, you needed to have netscape running in
> > >> interactive mode. The -remote commandline switch wants to contact an
> > >> already running netscape. 
> > > Well, one could write a queue system, that would reuse one netscape instance.
> > > Additionally the Framebuffer X server might be a nice idea.
> > > One problem with this that I see is, that you cannot be sure when the page is
> > > really loaded. (Perhaps I've overlooked something.) One way to monitor this
> > > might be writing a monitoring proxy server, so you can monitor if netscape is
> > > still activly loading something.
> > 
> > Hm..  Just  musing....
> > 
> > What about netscape running on a vnc server (X version of screen).  It
> > just runs there, and you can feed netscape the -remote commands.  If
> > you can "make" the html page you give the remote command to, you could
> > include a javascript onload that could doe stuff (load a new page in a
> > new window - whatever) to trigger the "completeness" of the page
> > loading...
> Well, right ;)
> But VNC is overkill. The Framebuffer X Server would make more sense, as
> it is designed for testing programs or things like this.
> > 
> > Netscape (at least navigator for Linux) does produce very nice
> > postscript code.  It's even readable...
> Well, but it produces in my experience rather poor Postscript :(
> (Background color for example is ignored, etc.)
> For b/w printers netscape might be ok, but since we replaced our printer
> with a color one it's quite frustrating :(
> (Netscape 6 has a much better output, but generates sometimes additional
>  blank pages)
> 


This thread is quite absurd! I sure hope you are all joking (some are
definitly).


Using netscape to do something like this is plain wrong - it would be a
VERY bad design, extremely fragile, heavy and in my experience the
netscape mechanism isn't robust either (I once hacked a quick script to
print out a collection of HTML pages - it did not work well).

Just imagine what happens if you send your script a page that requires a
brand new plug-in. Netscape will ask you if you want to download it - the
script (probably) hangs or doesn't produce PosScript for the right page..

I just mentioned the netscape URL for completeness, what it describes is
useful for some applications, but not for scripting.. 

I do think, however, that there should be a tool to produce postscript
from HTML. 

The best way to do this would probably be to modify the rendering engine
of some open sourcish browser to produce PostScript directly. Who knows,
maybe the mozilla people would like to do this as a spin off (if the
project isn't dead by now - even though they say Netscape 6 is based on
it). A command-line-only netscape version should work. Try to tell this to
the mozilla people, I'd say.

Such a solution would have the advantage of being maintained as new
techniques hit the web.

But the same difficulties apply as mentioned earlier: what to do about
that strange plug-in?

The conclusio probably is: HTML is only to be viewed/printed/faxed
interactively [at least for now].

Sorry, I sure don't like this.

--
peter


PS: I keep saying that faxing information is to destroy the information by
converting it to pixels. One could apply the same thought to (modern)
HTML: Converting a document to HTML actually destoys it - as stupid as it
is, sharing information only across the web makes it completly
unaccessible for anything but a presumably intelligent lifeform that scans
it via optical nerves.

PPS: Isn't this way off-topic?