Is it possible to run my Red Hat 8.0 system without a network card?
Sure, it's possible, and you don't have to do anything. Any normal
Linux system configures lo (the loopback interface) automatically.
It's even possible to run without that (as embedded Linux systems
sometimes do). However, you shouldn't have to do anything special.
Simply install your distribution and refrain from entering any
I recently used Konqueror 2.2.2 to visit a credit-card Web site (URL
withheld to protect the guilty) to check my account. The site said I
needed to upgrade to at least Netscape 4.0 or Microsoft Internet Explorer
4.0 in order to use the 128-bit-strong encryption their site used.
I know that Konqueror 2.2.2 is a much later version than 4.0 of either of
the browsers they mentioned. I checked the various browser settings,
and sure enough, SSL2 and 3 both were enabled, with encryptions going
as high as 168 bits. Almost every encryption standard was enabled
in my settings; the exceptions being FZA-FZA-CBC-SHA, FZA-NULL-SHA,
NULL-MD5 and NULL-SHA—they all say “0 of 0 bits”. Finally, I changed the user agent setting to broadcast that it was
Netscape or MSIE, and magically, I stopped getting the error messages.
Now when I fill in the forms, such as the login screen, my input is
and cookies. Nothing works. Am I doing something wrong or is the site nonstandard? If
I look in my settings under certificates, it says I have a
certificate from the Web site. I can verify that certificate, but it
does not say what type of encryption they use or give me any other useful
I'll bet the Web site uses the user agent string to help identify what it
thinks is a valid browser version.
Mozilla. If that doesn't work complain to the Web site.
Too many Web developers code to a particular
implementation (target platform) rather than to
the standard protocols and APIs (application
programming interfaces). This is exacerbated
several different Web browsers is daunting.
I encourage all Web site developers to start
with the simplest implementation of the
core requirements. Add bells and whistles in
core functions without it.
My suggestions: complain to your bank and try using Mozilla 1.x or
There is a silver lining, however. You may not
be aware of this, but the default Apple browser,
Safari, is based on the KHTML rendering engine
from the KDE Project. This, obviously, is the same
engine that Konqueror uses. Also, now that Microsoft discontinued
development on the Macintosh version of Internet
Explorer, Web developers who
code for Windows and Mac only will be coding
for you too.
I am using an ASUS motherboard. Under Microsoft Windows 98, a program called
asusprobe reports that the CPU temperature is about 47°C/116°F and the
motherboard temperature is about 31°C/87°F. These may change slightly but
are pretty steady. How do I find these temperatures under Linux?
Most commodity motherboards that provide this information are built
around the LM78 series of chips that communicate over the i2c 2-wire
bus. The SMBus is a particular implementation of i2c. The drivers and
utilities for accessing this information under Linux are in the
lm-sensors package. You can learn more about that project from
The i2c and lm-sensors drivers are included with mainstream kernel sources
and are compiled into all mainstream Linux distributions like Red Hat,
Debian and SuSE. Perusing the FAQ reveals that different motherboards report differing
numbers for temperature and voltage; you can adjust those settings
by editing the /etc/sensors.conf file. Debian installed a sample
sensors.conf file that's about 20 pages long.
Also, the FAQ specifically mentions ASUS P2B
motherboards in relation to odd temperature readings—if that's
your motherboard, read the FAQ at the above URL.
Incidentally, a number of packages use this lm-sensors
interface, dæmons that store histories of readings for statistical
analysis and graphing, GUI widgets that run in KDE, GNOME or Window
Maker panels and so on. At the very least you probably should use the sensors
command that will read the settings from your /etc/sensors.conf to
adjust the raw readings it gets from the drivers.
Suppose there's a site called www.foo.org/technical/pics.
How can I download only the pictures—let's say
the only extension is .jpg—from a Web site using
Here's an example:
wget -r -l1 --no-parent -A "*.jpg" \ http://www.server.com/dir/
This recursively (-r) downloads all the *.jpg files from the dir
directory on the www.server.com Web server up to one level depth.
Do a man wget for more of this great utility's
Felipe Barousse Boué