[an error occurred while processing this directive]

  Linux Letter 08

The Linux Letter for July 19, 1999

I spend a lot of time talking and writing about how wonderful Linux is for network connectivity. I've praised the fact that since networking has been an integral part of all UNIX operating systems, Linux has an advantage over other operating systems that have seen networking as an extra feature added over time.

But I also have to admit that I'm a little spoiled when it comes to networking. I do my work at nospin.org using a local area network, which is connected to a high speed DSL network. That means that I haven't had to use a modem to connect to the Internet for well over a year. So, that tends to make me unaware of the problems that occur when connecting a Linux computer to the Internet using a modem.

Fortunately, plenty of people have alerted me to my insensitivity. The most common questions that I receive are about problems connecting to the Internet with a modem. So let's take a look at how a Linux and a modem work to get you on the Internet.

Without delving into the arcane of the Internet, you should know that there are several ways to establish an internet connection. You've probably heard of TCP/IP, the underlying set of protocols that makes the Internet what it is. TCP/IP was designed to work with network devices, such as network interface cards. But since we can't all connect our computers to a local area network to get Internet access, we have to find other ways of getting around the virtual world. Since modems are a very common and relatively inexpensive way to move personal-sized chunks of data around the world, it seems only logical to use them to connect computers to the Internet.

Up until a few years ago, the only protocol choice for modems was the serial line interface protocol, or SLIP. It was designed to work with serial devices like modems, but wasn't the most efficient protocol to use because it was designed with relatively high speeds in mind. Fortunately, something better came along.

Point to Point Protocol, PPP, was created specifically with modems in mind. It was optimized to support the kind of data that a relatively slow Internet connection transmits. And since a huge fraction of Internet connections are made by modems, PPP has become virtually the only way to make an Internet connection with a modem.

All of this is a roundabout way to discuss the way that Linux supports modem connections. As you might expect, Linux supports PPP and supports any modem that does not require some soft of software support to function. In other words, if it's not a Winmodem, then Linux will probably support it.

So how do you make it all work? If you're using a standard Linux installation, you've already got the support built in. You must be sure that Linux knows which port your modem is configured to. If you don't know, a quick study of the Linux Modem HOWTO will speed you on your way.

Once you've configured your modem, you'll need to configure PPP. You'll need to get some information from your Internet service provider before you can proceed:

  • Dialup telephone number
  • Name server
  • Static IP address if one is assigned to you
  • Once you've collected this information, you can get to work

It seems safe to assume that if you are reading this, you have access to the Internet. Since that is probably the case, the best place for guidance on connecting your computer to the Internet is the PPP-HOWTO. Once you have configured your system, you can probably expect that things will not quite work correctly on the first try. Fear not! Simply turn to the PPP FAQ for more guidance. It contains a very detailed troubleshooting guide.

And if you continue to experience troubles, some of the most helpful people in the Linux community use the Internet. See the newsgroups in the comp.os.linux hierarchy. A search of deja.com is usually very helpful.

PPP-HOWTO: http://metalab.unc.edu/LDP/HOWTO/PPP-HOWTO.html

Linux PPP FAQ: http://metalab.unc.edu/LDP/FAQ/PPP-FAQ.html


Finally, let me take the opportunity to point out The NOSPIN Group's distribution of RedHat Linux 6.0. This distribution is different from those available in the stores and through mail order. Because we mirror RedHat's Linux directories daily, we always have the very latest files available directly from RedHat. So, instead of purchasing a stale, months old copy of the operating system, you can receive the very latest edition of RedHat Linux from us. And as a little extra, we'll even throw in a copy of WordPerfect for Linux 8.0 and the latest available Linux kernel.

Take a look at:


Hot Tip of the Week

If you're like me, you probably have experienced a core dump or two over time. Usually you can catch the resulting core file and deal with it appropriately. But sometimes you just miss them. Here's a method that will get rid of any lingering core files that are over five days old:

Put this line in a file in your cron.daily directory and make it executable:

find / -name core -atime +5 -exec rm -f {} ';'

Happy computing!

Drew Dunn


Get your free email account...  TODAY!!!


The Power


[an error occurred while processing this directive]