Re: Another fix for scs2pdf and more

On Mon, 11 Mar 2002, James Rich wrote:

> I posted a patch that fixes the page size problem.  The page size should
> now always be correct.  If it isn't let me know.

I found something strange -- I'll send you the SCS file, so you can
reproduce it...

>
> I also worked on one of the items on the TODO list which is to merge the
> library functions in scs2ascii.c, etc. into scs.c.  scs2ascii.c and
> scs2pdf.c now both #include scs.h which contains all the function
> definitions for things that they both use.  scs.c has been added to the
> build and to lib5250.la.  By doing this I discovered the solution to the
> page size problem.

Wow...  you're ambitious.  While you're at it, take a look at the
lp5250d-win.c file in the win32/ directory.   It implements scs2ascii
as a state machine because I found that the performance of using the
pipe in Windows was horrendous.

What we should do is come up with a general set of routines that we can
"attach" depending on the desired driver.   Sort of the way the different
stream objects work.  They could be used to create scs2ascii, scs2ps and
scs2pdf -- but also to attach directly into lp5250d under Windows.
Do you follow me?


>
> The diff is against CVS I just checked out (I noticed Scott checked in
> some of my changes just a few hours ago).  Get 'em here:
>
> http://www.eaerich.com/tn5250.diff
> http://www.eaerich.com/scs.c
>
> Scott, can you check these in for me?  I know you use scs2ascii a lot and
> I want to make sure everything is rock solid.
>

Okay, I've committed these into CVS.  They appear to work.  I should get
off my ass and finish up the Windows stuff so we can make a release.