Re: [LINUX5250] Re: scs2ps

[James Rich <james@xxxxxxxxxxx>]

On Wed, 13 Mar 2002, Scott Klement wrote:

> > I still prefer environment variable, though.  It seems to me that in the
> > future we would want the ability to change the behavior of scs2* without
> > restarting lp5250d.  An environment variable would allow that.
> >
>
> It seems to me that you'd still have to restart lp5250d..   unless you
> set your outputcommand= argument to a shell script -- and had that script
> run scs2pdf.  Then you could change the script, and lp5250d would call the
> new script with the next document.
>
> Of course, that would work for either a command-line argument, or for an
> environment var.
>
> So, I guess I don't follow you.  How does an env var allow us to change
> the behavior while it's running?

I guess my thinking is like this:  lp5250d is what actually talks to the
iSeries.  If lp5250d could somehow receive information about a specific
job, then it could change the environment variables accordingly.  I guess
it could manipulate command line arguments as well before starting the
appropriate program, but that seems hard.  Say for example that lp5250d
somehow knew that the spoolfile it was about to print was called MYSPOOLF.
Before invoking /usr/bin/scs2pdf it set the environment variable
TN5250_PDF=myspoolf.pdf.  Now the resulting PDF has the same name that the
spoolfile did.  Of course I have no idea how lp5250d would know any of
this since the spoolfile name is not in the print stream.  But there
might be some mechanism...

James Rich
james@eaerich.com