I don’t use Gnome, KDE, etc.
My window manager is WMX with some patches.
Without all those fancy window environments, you get rid of all magic auto configurations, and wizards at the same time, and you can configure your system like real man do: using the command line. ;-)
I bought a HP OfficeJet Pro 8500 A910 last week.
Here’s how I set up and use the scanner of this device using the command line:
SETUP the scanner:
> hp-mkuri -i <ip_of_your_printer>
(hp-mkuri works also for devices on routed networks: no broadcast magic needed)
Store this “hpaio:…” string in the environment variable SANE_DEFAULT_DEVICE (.bashrc)
USE the scanner:
> scanimage -y 297 -x 210 –resolution 300 | pnmtopng > scan.png
Scans an A4 page with 300dpi.
Package used for this:
hplip 3.10.6-2 HP Linux Printing and Imaging System
sane-utils 1.0.21-9 API library for scanners — utilities
If you look into a harddisk from an HP server with smart array configured for raid1, you will recognize that the partition table of this disk has only one partition:
Device Boot Start End Blocks Id System
<device>p3 1 1 8001 12 Compaq diagnostics
The real partitions are invisible? No. I found the start of the real data and the real partition table at block 1088 (*512Byte). 1088 = 17 * 63 ?? Where does this number come from ?
You can get the real disk start with:
dd if=disk bs=512 skip=1088 count=100 of=/tmp/head
If you want to mount a partition from this disk directly, you can do this with losetup:
losetup -f -o $(( ( 1088 + <partition_start_block_from_fdisk> ) * 512 )) disk
mount /dev/loop0 /mnt/target
I don’t know if this number 1088 is correct for all disks but it was correct for all I checked.
A small mysqldump hint:
You can use mysqldump –skip-opt to get proper insert statements in the SQL dump.
But –skip-opt disables –quick which means mysqldump buffers full tables before writing -> Out of memory.
In this case I use: mysqldump –skip-opt –quick <DBNAME>