While working on wary-5.3-barebones I decided to have another go at bypassing all the boot steps in the older xorg 7.3 package. Here's what I did.
1.) I booted wary-5.3 in the normal way, choosing country (locale), keyboard, timezone, and Xvesa 640x480x16 (I figured this was as close to a universally acceptable resolution as I was going to get.
2.) Then I used edit-sfs to open up puppy_wary_5.3.sfs and copied in the following files from my running system:
- /usr/bin/
X (symlink to Xvesa)
/etc/
keymap
mousebuttons
mousedevice
videomode
/etc/X11/
xorg.conf
xorg.conf.
/usr/lib/locale/en_US
LC_ADDRESS
LC_COLLATE
LC_CTYPE
LC_IDENTIFICATION
LC_MEASUREMENT
LC_MESSAGES
LC_MONETARY
LC_NAME
LC_NUMERIC
LC_PAPER
LC_TELEPHONE
LC_TIME
Code: Select all
#!/bin/sh
if [ ! -f /etc/Xvesa_done ]
then
exec video-wizard &
touch /etc/Xvesa_done
fi
It also creates a file in /etc called Xvesa_done which keeps the script from being repeated when X is restarted or when the machine is rebooted with a warysave file.
The user can also change to Xorg by going to the Setup menu and choosing Xorgwizard. Likewise they can change locale by going to the Desktop menu and choosing Chooselocale. I'm sure a Quickstart screen could be built as in the more modern Puppies.
I'll be glad to upload an ISO if anyone is interested, either barebones or full Wary. I also built one with all the possible xorg modules stripped out, it worked but only saved 4MB so didn't really seem worthwhile.
Anyway I had fun doing this and am using it as a print server. It does a complete boot headless, that is without monitor, so I don't have to have a monitor setup all the time on my print server.
Cheers, J