24/32 bpp with 9libs vs. orig libXg

24/32 bpp with 9libs vs. orig libXg

[James A. Robinson]

A few people have noted problems running sam and 9term on 24 bpp or 32
bpp displays (we would get the error 'libg: i/o error' when we tried to
open a menu in 9term or simply bring up samterm). It appears that my
problem may have stemmed from a problem specific to the 9libs port of
the original libXg library.

The other day I installed limbo, and found that wm/wm couldn't handle
16bpp depth. 8bpp was unacceptable to me, and I had not been able to
run sam and 9term in 24bpp mode before.

Today I installed the original samterm from Rob Pike's release, available
at ftp://netlib.bell-labs.com/netlib/research/sam.shar.gz, and found the
samterm compiled from that source works perfectly at 24bpp. I recompiled
my 9term against this original libXg, and found that it also works at
24bpp without any visible problems. Since the samterm works with the
latest sam (editor only) released a few months ago, I'm happy. =)

My original goal was to get a static version of libXg compiled so that I
could debug it with gdb. I was having great difficulty getting the 9libs
version working with gdb (perhaps because it uses this .la/.lo format
that I've never really understood).  Imagine my surpise when compiling
the original libXg fixed the problem. =)

Jim