SuSe 10 Linux Termcap for filePro

[Fairlight]

>From inside the gravity well of a singularity, Nancy Palmquist shouted:
> I got two excellent suggestions about this problem.  I tried Bob's 
> suggestion about the termcap.  It made no difference in the display.
> 
> I also got a suggestion from both Bob and Craig about checking the LANG= 
>   setting in Yast.  I found it and the LANG was set to English (US) but 
> on the detail page it had UTF-8 encoding checked.  Well, the yast 
> graphics displayed crappy also, so I unchecked that setting.

Gack...il8n is evil.  This started showing up in earnest about the era of
Red Hat 8.0.

> After yast did it's thing, the yast display's graphics were fixed.
> 
> However, the filePro graphics were still crappy.

Can we narrow down "crappy" to a definition a bit more conduscive to fixing
the problem, please?

> I tried the termcap suggestion in conjunction with the yast suggestion. 
>   No better.
> 
> xterm made worse graphic issues.  SCOANSI seemed to work the best - by 
> best I mean only one bad graphic character instead of two.

Which one/two are bad?

> Well something must still be missing.  I can't figure what.
> 
> Thanks for the great suggestions.  Got any more ideas?  ;-)

Don't use it from a SCO console is the first thing that comes to mind.  :)

Actually, you -do- have 'mapchan -n' set during this...right?

Further...  Don't use xterm.  SCO's console is not an xterm, and your
results will not improve.  Trying semi-random settings and praying they
work isn't likely to be fruitful.

You took Bob's suggestion and used the scoansi termcap in the main termcap.
You've not yet said (or if you did, I missed it) what corresponding PFTERM
you're using simultaneously.  What -are- you putting in that variable?  And
is the entry the same as on a system that does work for you?

Let's also take a step back a second.  Can you actually use fP from a SuSE
9.x box on the same exact SCO console without problems?  Are you -only-
having problems with SuSE 10, or is 9.x an issue as well?  If 9.x is fine,
then we could start a differential analysis of the relevant factors between
the two...something systematic.  There are only so many subsystems involved
here, which is to say "not many".

Another thought:  What version of fP?  5.6?  5.0.xx where xx >= 10?  

Ken?  If she has 5.6, is the linux version of that still compiled
dynamically, or have you folks reverted back to static linking?  I'm
thinking it's -possible- libtermcap may have been changed (although it
would surprise me at this late date).  Its potential relevance depends on
the linking.

mark->