So you've finally managed to setup your network connection, now what? How do you know that it's working? How do you know that you set it up correctly? And just what do you do now that it's setup? Well this chapter is for you.

Slackware Linux includes a great many networking tools for troubleshooting and diagnosing network connection troubles, or just for seeing what's out there on the network. Most of these tools are command-line tools, so you can run them from a virtual terminal or in a console window on your graphical desktop. A few of them even have graphical front-ends, but we're going to deal almost exclusively with command-line tools for now.

ping(8) is a handy tool for determining if a computer is operational on your network or on the Internet at large. You can think of as a type of sonar for computers. By using it, you send out a "ping" and listen for an echo to determine if another computer or network device is listening. By default, ping checks for the remote computer once per second indefinitely, but you can change the interval between checks and the total number of checks easily, just check the man page. You can terminate the application at any time with CTRL-c. When ping is finished, it displays a handy summary of its activity. ping is very useful for determining if a computer on your network or the Internet is available, but some systems block the packets ping sends, so sometimes a system may be functioning properly, but still not send replies. darkstar:~# ping -c 3 www.slackware.com 64 bytes from slackware.com (64.57.102.34): icmp_seq=1 ttl=47 time=87.1 ms 64 bytes from slackware.com (64.57.102.34): icmp_seq=2 ttl=47 time=86.2 ms 64 bytes from slackware.com (64.57.102.34): icmp_seq=3 ttl=47 time=86.7 ms --- slackware.com ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 2004ms rtt min/avg/max/mdev = 86.282/86.718/87.127/0.345 ms

traceroute(8) is a handy tool for determining what route your packets take to reach some other computer. It's mainly of use for determining which computers are "near" or "far" from you. This distance isn't strictly geographical, as your Internet Service Provider may route traffic from your computer in strange ways. traceroute shows you each router between your computer and any other machine you wish to connect to. Unfortunately, many providers, firewalls, and routers will block traceroute so you might not get a complete picture when using it. Still, it remains a handy tool for network troubleshooting. darkstar:~# traceroute www.slackware.com traceroute to slackware.com (64.57.102.34), 30 hops max, 46 byte packets 1 gw.ctsmacon.com (192.168.1.254) 1.468 ms 2.045 ms 1.387 ms 2 10.0.0.1 (10.0.0.1) 7.642 ms 8.019 ms 6.006 ms 3 68.1.8.49 (68.1.8.49) 10.446 ms 9.739 ms 7.003 ms 4 68.1.8.69 (68.1.8.69) 11.564 ms 6.235 ms 7.971 ms 5 dalsbbrj01-ae0.r2.dl.cox.net (68.1.0.142) 43.859 ms 43.287 ms 44.125 ms 6 dpr1-ge-2-0-0.dallasequinix.savvis.net (204.70.204.146) 41.927 ms 58.247 ms 44.989 ms 7 cr2-tengige0-7-5-0.dallas.savvis.net (204.70.196.29) 42.577 ms 46.110 ms 43.977 ms 8 cr1-pos-0-3-3-0.losangeles.savvis.net (204.70.194.53) 78.070 ms 76.735 ms 76.145 ms 9 bpr1-ge-3-0-0.LosAngeles.savvis.net (204.70.192.222) 77.533 ms 108.328 ms 120.096 ms 10 wiltel-communications-group-inc.LosAngeles.savvis.net (208.173.55.186) 79.607 ms 76.847 ms 75.998 ms 11 tg9-4.cr01.lsancarc.integra.net (209.63.113.57) 84.789 ms 85.436 ms 85.575 ms 12 tg13-1.cr01.sntdcabl.integra.net (209.63.113.106) 87.608 ms 84.278 ms 86.922 ms 13 tg13-4.cr02.sntdcabl.integra.net (209.63.113.134) 87.284 ms 85.924 ms 86.102 ms 14 tg13-1.cr02.rcrdcauu.integra.net (209.63.114.169) 85.578 ms 85.285 ms 84.148 ms 15 209.63.99.166 (209.63.99.166) 84.515 ms 85.424 ms 85.956 ms 16 208.186.199.158 (208.186.199.158) 86.557 ms 85.822 ms 86.072 ms 17 sac-main.cwo.com (209.210.78.20) 88.105 ms 87.467 ms 87.526 ms 18 slackware.com (64.57.102.34) 85.682 ms 86.322 ms 85.594 ms

Once upon a time, telnet(1) was the greatest thing since sliced bread. Basically, telnet opens an unencrypted network connection between two computers and hands control of the session to the user rather than some other application. Using telnet, people could connect to shells on other computers and execute commands as if they were physically present. Due to its unencrypted nature this is no longer recommended; however, telnet is still used for this purpose by many devices. Today, telnet is put to better use as a network diagnostic tool. Because it passes control of the session directly to the user, it can be used for a great variety of testing purposes. As long as you know what ASCII commands to send to the receiving computer, you can do any number of activies, such as read web pages or check your e-mail. Simply inform telnet what network port to use, and you're all set. darkstar:~$ telnet www.slackware.com 80 Trying 64.57.102.34... Connected to www.slackware.com. Escape character is '^]'. HEAD / HTTP/1.1 Host: www.slackware.com HTTP/1.1 200 OK Date: Thu, 04 Feb 2010 18:01:35 GMT Server: Apache/1.3.27 (Unix) PHP/4.3.1 Last-Modified: Fri, 28 Aug 2009 01:30:27 GMT ETag: "61dc2-5374-4a973333" Accept-Ranges: bytes Content-Length: 21364 Content-Type: text/html

As we mentioned, telnet may be useful as a diagnostic tool, but its unencrypted nature makes it a security concern for shell access. Thankfully, there's the secure shell protocol. Nearly every Linux, UNIX, and BSD distribution today makes use of OpenSSH, or ssh(1) for short. It is one of the most commonly used network tools today and makes use of the strongest cryptographic techniques. ssh has many features, configuration options, and neat hacks, enough to fill its own book, so we'll only go into the basics here. Simply run ssh with the user name and the host and you'll be connected to it quickly and safely. If this is the first time you are connecting to this computer, ssh will ask you to confirm your desire, and make a local copy of the encryption key to use. Should this key later change, ssh will warn you and refuse to connect because it is possible that some one is attempting to hijack the connection using what is known as a man-in-the-middle attack. darkstar:~# ssh alan@slackware.com alan@slackware.com's password: secret alan@slackware.com:~$ The user and hostname are in the same form used by e-mail addresses. If you leave off the username part, ssh will use your current username when establishing the connection.

So far all the tools we've looked at have focused on making connections to other computers, but now we're going to look at the traffic itself. tcpdump(1) (which must be run as root) allows us to view all or part of the network traffic originating or received by our computer. tcpdump displays the raw data packets in a variety of ways with all the network headers intact. Don't be alarmed if you don't understand everything it displays, tcpdump is a tool for professional network engineers and system administrators. By default, it probes the first network card it finds, but if you have multiple interfaces, simply use the -i argument to specify which one you're interested in. You can also limit the data displayed using expressions and change the manner in which it is displayed, but that is best explained by the man page and other reference material. darkstar:~# tcpdump -i wlan0 tcpdump: verbose output suppressed, use -v or -vv for full protocol decode listening on wlan0, link-type EN10MB (Ethernet), capture size 96 bytes 13:22:28.221985 IP gw.ctsmacon.com.microsoft-ds > 192.168.1.198.59387: Flags [P.], ack 838190560, win 3079, options [nop,nop,TS val 1382697489 ecr 339048583], length 164WARNING: Short packet. Try increasing the snap length by 140 SMB PACKET: SMBtrans2 (REPLY) 13:22:28.222392 IP 192.168.1.198.59387 > gw.ctsmacon.com.microsoft-ds: Flags [P.], ack 164, win 775, options [nop,nop,TS val 339048667 ecr 1382697489], length 134WARNING: Short packet. Try increasing the snap length by 110 SMB PACKET: SMBtrans2 (REQUEST)

Suppose you need to know what network services are running on a machine, or multiple machines, or you wish to determine if multiple machines are responsive? You could ping each one individually, telnet to each port you're interested in, and note every detail, but that's very tedious and time consuming. A much easier alternative is to use a port scanner, and nmap(1) is just the tool for the job. nmap is capable of scanning TCP and UDP ports, determining the operating system of a network device, probing each located service to determine its specific type, and much much more. Perhaps the simplist way to use nmap is to "ping" multiple computers at once. You can use network address notation (CIDR) or specify a range of addresses and nmap will scan every one and return the results to you when it's finished. You can even specify host names as you like. In order to "ping" hosts, you'll have to use the -sP argument. The following command instructs nmap to "ping" www.slackware.com and the 16 IP addresses starting at 72.168.24.0 and ending at 72.168.24.15. darkstar:~# nmap -sP www.slackware.com 72.168.24.0/28 Should you need to perform a port scan, nmap has many options for doing just that. When run without any arguments, nmap performs a standard TCP port scan on all hosts specified. There are also options to make nmap more or less aggressive with its scanning to return results quicker or fool intrusion detection services. For a full discussion, you should refer to the rather exhaustive man page. The following three commands perform a regular port scan, a SYN scan, and a "Christmas tree" scan. darkstar:~# nmap www.example.com darkstar:~# nmap -sS www.example.com darkstar:~# nmap -sX www.example.com Be warned! Some Internet Service Providers frown heavily on port scanning and may take measures to prevent you from doing it. nmap and applications like it are best used on your own systems for maintenance and security purposes, not as general purpose Internet scanners.

Often network problems stem from a failure of DNS (Domain Name Service) which maps domain names to IP addresses. An easy way to perform quick DNS lookups is the host(1) command. When this is run, your computer will perform a few common DNS lookups and return the results. darkstar:~# host www.slackware.com www.slackware.com is an alias for slackware.com. slackware.com has address 64.57.102.34 slackware.com mail is handled by 1 mail-mx.cwo.com.

More complex DNS lookups can be manually performed with the dig(1) tool. dig is "the meanest dog in the pound" when it comes to troubleshooting DNS issues. With this tool, you can perform virtually any DNS lookup from reverse lookups to A, CNAME, MX, SP, TXT records and more. There are far too many command-line options and lookup types to go into depth here, but the man page lists all the common use cases. darkstar:~# dig @207.69.188.185 www.slackware.com a ; <<>> DiG 9.4.3-P4 <<>> @207.69.188.185 www.slackware.com a ; (1 server found) ;; global options: printcmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 57965 ;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 2, ADDITIONAL: 0 ;; QUESTION SECTION: ;www.slackware.com. IN A ;; ANSWER SECTION: www.slackware.com. 86400 IN CNAME slackware.com. slackware.com. 86400 IN A 64.57.102.34 ;; AUTHORITY SECTION: slackware.com. 86400 IN NS ns2.cwo.com. slackware.com. 86400 IN NS ns1.cwo.com. ;; Query time: 348 msec ;; SERVER: 207.69.188.185#53(207.69.188.185) ;; WHEN: Sat Jul 3 16:25:10 2010 ;; MSG SIZE rcvd: 105 Let's take a took at the command-line options used above. The @207.69.188.185 argument tells dig what DNS server to query. If it is not specified, dig will simply use whatever servers are listed in /etc/resolv.conf. The a argument at the end is the type of DNS record to lookup. In this case we looked for an "A" record which returned an IPv4 address.

finger(1) isn't exactly a network diagnostic tool as much as it is a network-user diagnostic tool. Using finger, you can gather a handful of useful information about users on servers running the fingerd(8) daemon. Today very few servers still offer fingerd, but for those that do it can be a useful tool for keeping track of your friends and co-workers. darkstar:~# finger alan@cardinal.lizella.net [cardinal.lizella.net] Login: alan Name: Alan Hicks Directory: /home/alan Shell: /bin/bash Office: 478 808 9919, 478 935 8133 On since Wed Apr 13 17:43 (UTC) on pts/9 from 75-150-12-113-atlanta.hfc.comcastbusiness.net 32 minutes 24 seconds idle (messages off) On since Wed Apr 13 17:45 (UTC) on pts/10 from :pts/9:S.0 48 minutes 56 seconds idle Mail forwarded to alan@lizella.net No mail. No Plan.

Slackware includes a variety of web browsers. If you're using a graphical desktop, you'll find Firefox, Seamonkey, and others you may already be familiar with, but what about console access? Fortunately, there are a number of capable web browsers here as well.

The oldest console-based web browser included with Slackware is definitely lynx(1), a very capable if somewhat limited web browser. lynx does not support frames, javascript, or pictures; it is strictly a text web browser. Navigation is performed using your keyboard's arrow keys and optionally, a mouse. While it lacks many features that other browsers support, lynx is one of the fastest web browsers you'll ever use for gathering information. For example, the -dump argument sends the formatted web page directly to the console, which can then be piped to other programs.

A more feature-rich alternative is the popular links(1), a console-based web browser that supports frames and has better table rendering than lynx. Like its predecessor, links is navigable with the arrow keys, and the use of a mouse is supported. Unlike lynx, it also includes a handy menu (simply click on the top line with your mouse to activate) and generally formats web pages better.

Unlike the other browsers we've looked at, wget(1) is non-interactive. Rather than display HTTP content, wget downloads it. This takes the "browsing" out of the web browser. Unlike the dump modes of other browsers, wget does not format its downloads; rather it copies the content in its exact form on the web server with all tags and binary data in place. It also supports several recursive options that can effectively mirror online content to your local computer. wget need not operate exclusively on HTTP content; it also supports FTP and several other protocols. darkstar:~# wget ftp://ftp.osuosl.org/pub/slackware/slackware-current/ChangeLog.txt --2010-05-01 13:51:19-- ftp://ftp.osuosl.org/pub/slackware/slackware-current/ChangeLog.txt => `ChangeLog.txt' Resolving ftp.osuosl.org... 64.50.236.52 Connecting to ftp.osuosl.org|64.50.236.52|:21... connected. Logging in as anonymous ... Logged in! ==> SYST ... done. ==> PWD ... done. ==> TYPE I ... done. ==> CWD /pub/slackware/slackware-current ... done. ==> SIZE ChangeLog.txt ... 75306 ==> PASV ... done. ==> RETR ChangeLog.txt ... done. Length: 75306 (74K) 100%[======================================>] 75,306 110K/s in 0.7s 2010-05-01 13:51:22 (110 KB/s) - `ChangeLog.txt' saved [75306]

Slackware also includes a variety of email clients. If you're using a graphical desktop, you'll find Thunderbird, Kmail, sylpheed and others. As with web browsers, there are also applications that function within the shell. Once you start using an email client in the console, you may find yourself not wanting to use anything else; the flexibility and configurability can be addicting.

pine is one of the oldest command-line interface mail clients still in existance and remains one of the most user-friendly. pine was created by the University of Washington and carries with it both a trademark and a copyright license that are difficult to work with. Thankfully back in 2005, the university saw fit to re-write it without the trademark and with a more open license, so alpine(1), the pine-clone distributed with Slackware, was born. To start using alpine, simply type pine at the command line. Using it is very simple due to its menu-driven system as well as the command reference neatly located at the bottom of the screen. See for yourself: ALPINE 2.00 MAIN MENU Folder: INBOX No Messages ? HELP - Get help using Alpine C COMPOSE MESSAGE - Compose and send a message I MESSAGE INDEX - View messages in current folder L FOLDER LIST - Select a folder to view A ADDRESS BOOK - Update address book S SETUP - Configure Alpine Options Q QUIT - Leave the Alpine program Copyright 2006-2008 University of Washington [Folder "INBOX" opened with 0 messages] ? Help P PrevCmd R RelNotes O OTHER CMDS > [ListFldrs] N NextCmd K KBLock Before configuring any mail client, you should check the documentation of your mail server to gather all of the pertinent information about what protocols and security measures your mail service uses. This will help you configure pine correctly. To start configuring pine, you should be able to enter the [S]ettings menu, and then the [C]onfig screen. There are about a hundred different options, but the most common and most important ones would be your Personal Name, Inbox Path, and SMTP Server. Personal Name The name you wish people to see in the "From" field of your emails. Inbox Path Your mail server. The server itself is often something like mail.example.com (although it may not be; check with your host to be sure), but if that server requires login information then you may need to set it to something like bob@mail.example.com or similar. SMTP Server The server through which you will send your email; often this is smtp.example.com or outgoing.example.com or some variation on that theme. It often requires authentication, but you should be prompted for this in pine when sending mail. A sample configuration might look a little something like this: Personal Name Alan Hicks SMTP Server smtp.lizella.net Inbox Path {imap.lizella.net/user=alan@lizella.net}inbox Warn if Blank Subject X Sort Key: Arrival X And so on. Once you have it configured according to your mail server's settings, you will be able to navigate the application from the main menu. It's quite intuitive, with a persistent list of keyboard shortcuts at the bottom of the screen.

Some people don't like pine. Some people want more control. Some people want a fully-configurable mail client with plugin support and a no-nonsense attitude. Those people use mutt(1). mutt isn't as user friendly as pine, but makes up for it with power. You won't find the user-friendly command reference at the bottom of the screen, mutt uses every last inch of real-estate for mail processing duty. It's feature support is extensive: threaded displays are no problem for the mighty mixed-breed! You can configure mutt with a .muttrc file in your home directory. With all the many different possible configuration options, there's even a man page for that, muttrc(5). You might want to read up on it. 3172 N Jan 17 Thomas Morper (2.8K) │ └─>Re: [Slackbuilds-users] Exim Sbo 3173 N Jan 17 TuxaneMedia (2.5K) └─>Re: [Slackbuilds-users] Exim Sbo 3174 N Jan 06 Uli Sch?fer (4.6K) [Slackbuilds-users] mrtg-2.15.2 breaks wi 3175 N Jan 06 Willy Sudiarto (4.0K) └─> 3176 N Jan 06 Audrius Ka??uka (3.9K) └─> 3177 N Jan 06 Niels Horn (4.4K) └─> 3178 N * Jan 09 King Beowulf (6.5K) [Slackbuilds-users] libreoffice (3.3.0rc 3179 N * Jan 09 Niels Horn (4.8K) └─>Re: [Slackbuilds-users] libreoffice (3 3180 * Jan 09 Niels Horn ( 73) └─> 3181 N * Jan 09 King Beowulf (5.9K) └─> 3182 N Jan 10 Robby Workman (7.8K) [Slackbuilds-users] Updates - 20110110 3183 N Jan 10 B Watson (3.6K) [Slackbuilds-users] Fwd: nted slackbuild 3184 N Jan 11 SlackBuilds@cat (2.5K) [Slackbuilds-users] /usr/share/ package s 3185 N Jan 11 Robby Workman (3.4K) └─> 3186 N Jan 12 SlackBuilds@cat (3.4K) └─> 3187 N Jan 11 Robby Workman (4.5K) └─> 3188 D Jan 12 Klaatu (3.5K) [Slackbuilds-users] LiVES updates 3189 N Jan 12 Robby Workman (2.3K) [Slackbuilds-users] Unavailability of mai 3190 N Jan 13 Mr. B-o-B (3.7K) └─> 3191 N Jan 13 Josiah Boothby (3.2K) [Slackbuilds-users] Ted's Word Processor 3192 N Jan 14 Adam Swift (7.4K) [Slackbuilds-users] Web submission failur 3193 N Jan 14 Rob McGee (3.5K) └─> Using mutt is unique because it is by nature a Mail User Agent (MUA), meaning its true purpose is to read and sort email. This was its only job originally, although some additional features such as retrieving mail via POP3 and even very basic transfering messages via SMTP have snuck into the application. As is so often the case with robust console-based applications, the configuration options are myriad, and there is no "right" or "wrong" way of using mutt as long as it does what you want it to do. Here is a simple example of a system utilizing mutt, with a toolchain consisting of: fetchmail to pull mail from your mail server via POP3 procmail to sort the mail as needed. mutt to read, sort, and send email. Optionally, formail to re-format your mail if needed (depending on your mail server settings) fetchmail is a simple application that exists only to pull mail from mail servers and copy those email messages to your local hard drive. The configuration is fairly simple: darkstar:~$ cat ~/.fetchmailrc poll pop.example.com with proto pop3 user "bob@example.com" there has password "secretpassword" is bdobbs here and wants mds "/usr/bin/procmail" poll Defines the mail server to check with proto Defines the protocol used; in this case, POP3. user Your email username. Sometimes this is just the string in front of the @ at symbol, othertimes it is your whole email address. Again, see your provider's documentation to find out what you should be using. there has password Your email password. This is entered as plaintext into your config file, but since this is in your home directory, it is only readable by you. For good measure though, it is safe to chmod 0600 ~/.fetchmailrc is username here Defines which user on your local computer this email account actually belongs to. For instance, if my email account is bob@example.com, but my username on my Slackware box is bdobbs, then obviously there would be no way for fetchmail to know who bob@example.com is. This defines it. and wants mda "/usr/bin/procmail" Tells fetchmail what application will be sorting the email once it is pulled from the mail server. Procmail ships with Slackware, so this will probably not change for you and is not dependent upon your mail server whatsoever. Now that fetchmail is configured, it will successfully pull your email from the server. But what will your computer do with all of that mail? This is the job of procmail. The settings for procmail can be stored in ~/.procmailrc and is also fairly simple; for example: PATH=/usr/bin MAILDIR=$HOME/Mail DEFAULT=$MAILDIR/Inbox LOGFILE=$MAILDIR/logs Now procmail is configured, so your mail will be sent to the right place. There is much more you can do with procmail so feel free to investigate its many filtering functions (or "recipes") later, once you have everything working. At this point, your email will be pulled from the server by fetchmail and sorted on your local system by way of procmail. All that's left to do is to configure mutt so that it knows where to find your email messages so that you can view them. Because mutt is such an extensible program, it's a good idea to make a folder for its configuration file. Your initial configuration will be frightfully simple, but if you ever look for new configurations of mutt you might be happy to have a whole folder to play around in. darkstar:~$ cat ~/.mutt/.muttrc set realname="Bob Dobbs" set from="bob@example.com" set spoolfile=/home/bdobbs/Mail/Inbox set mbox=/home/bdobbs/Mail/mbox set record=/home/bdobbs/Mail/sent set postponed=/home/bdobbs/Mail/postponed set sort=reverse-date set editor=emacs set smtp_url="smtp://bob:secretpassword@smtp.example.com:465" In this sample configuration, the spoolfile defines the incoming messages that you have not read yet (you might recognize it as the end destination for procmail), while mbox defines where mail that you have read and responded to will be stored. All other settings are self-explanatory, and all have been drawn directly from the muttrc man page; refer to it for more options. If your mail server delivers your email in a format that mutt cannot understand, use the formail mailbox converter, a part of the procmail distribution. For example: darkstar:~$ formail -ds < ~/Mail/Inbox >> ~/Mail/Muttbox Rather than performing this conversion every time you check your email, you might want to script it to happen automatically after procmail is finished its job. The commands used to navigate around in mutt are highly customizable but the defaults can be listed by typing ?.

If your mail server supports SSL then you can and should use SSL for better security. Most graphical clients have common SSL certificates built into them; mutt does not but it's trivial to configure. Download an SSL trust certificate (the most common set is from Thawte) and place it in your Mutt or Mail folder, and list it as a certificate_file in your .muttrc file. If your mail server is not using Thawte, then they may be using either another SSL certificate authority or they may be their own certificate authority, but if an unknown certificate is encountered, mutt will prompt you to accept it (or not). If you do accept it, the certificate can be saved into this certificate file. To obtain the Thawte certificate: bash$ wget http://www.thawte.com/roots/thawte_Premium_Server_CA.pem -P ~/Mail/

So those are great and everything, but what if you just want a mail client that isn't menu-driven? Thankfully mailx is here to save you. mailx is based on the Berkeley Mail application, with a mail command appearing as early as Version 1 of AT&T's UNIX. It can be used either interactively or non-interactively. To launch mailx, type mail at a shell prompt. mailx reads mail from your computer's mail spool and displays the usual combination of sender, subject, status, and size in a list, leaving the user at an interactive prompt. In fact, it might look familiar to you if you bothered checking your mail immediately after installing Slackware and read Pat Volkerding's greeting. root@darkstar:/home/bdobbs# mail Heirloom mailx version 12.4 7/29/08. Type ? for help. "/var/spool/mail/root": 2 messages 2 new >N 1 To root Thu Mar 10 23:33 52/1902 Register with the Linux counter project N 2 To root Thu Mar 10 23:35 321/15417 Welcome to Linux (Slackware 13.37)! ? To read a message, enter the number of the message at the prompt. This displays the message using more, so use the RETURN key to view the next page. Once the end of the message has been reached, press q to return to the list view, or RETURN to continue to the next message. To see a list of available commands, type ? at the mail prompt; using the commands provided, you can view the headers of mail in the spool, reply, delete, save, and many other common email tasks. mailx is most powerful when used in scripting. For all of the options available for mailx, view its man page. A simple way to send an email to someone requires only the command itself and the destination address: darkstar:~$ mail alan@lizella.net After the command has been issued, an interactive prompt appears for a subject line, the message body, and the end character (a single dot on an otherwise empty line). mailx can be used entirely without human intervention, however. Generally, it's safe to assume that any attribute you can define in the interactive shell for mailx can also be defined while scripting it or using it as one non-interactive command: darkstar:~$ mail -n -s "Test message" alan@lizella.net < ~/message.txt In this example, the contents of the file message.txt would be sent as the message body to the specified recipient. No interaction from the user is required. Within one's own computer (localhost) or one's own network, sending email in this manner is entirely possible. But over the internet a few more steps are usually required along the way. Of course, most notably there is usually an smtp server handling the delivery of your email. This, too, can be specified as part of your mail command: darkstar:~$ env MAILRC=/dev/null from="bob@example.com (Bob Dobbs)" smtp=relay.example.com mail -n -s "Test message" connie@example.com < ~/message.txt In this case, the MAILRC variable is set to /dev/null in order to override any system defaults, and the smtp server as well as the FROM: line are defined. The rest of the command is the same as using mailx internally within one's own computer or network. Over all, mailx is usually viewed as a mail client with the bare-minimum features; this is largely true, but when you need to be able to script sending notification emails or important update messages, it quickly becomes a lot more valuable than a fully interactive application like pine or mutt.

Lots of data is stored on FTP servers the world over. In fact, Slackware Linux was first publically offered via FTP and continues to be distributed in this fashion today. Most open source software can be downloaded in source code or binary form via FTP, so knowing how to retrieve this information is a handy skill.

The simplest FTP client included with Slackware is named simply, ftp(1) and is a reliable if somewhat simple means of sending and retrieving data. ftp connects to an FTP server, asks for your username and password, and then allows you to put or get data to and from that server. ftp has fallen out of favor with more experienced users do to a lack of features, but remains a handy tool, and much of the documentation you see online will refer you to it. Once an FTP session has been initialized, you'll be placed at a prompt somewhat like a shell. From here you can change and list directories using the "cd" and "ls" commands, just like a shell. Additionally, you may issue the "put" command to send a file to the server, or a "get" command to retrieve data from the server. If you're connecting to a public FTP server, you'll want to use the "anonymous" username and simply enter your e-mail address (or a fake one) for the password. darkstar:~# ftp ftp.osuosl.org Name (ftp.osuosl.org:alan): anonymous 331 Please specify the password. Password: secret 230 Login successful. Remote system type is UNIX. Using binary mode to transfer files. ftp> cd pub/slackware/slackware-current/ 250 Directory successfully changed. ftp> get ChangeLog.txt local: ChangeLog.txt remote: ChangeLog.txt 200 PORT command successful. Consider using PASV. 150 Opening BINARY mode data connection for ChangeLog.txt (33967 bytes). 226 File send OK. 33967 bytes received in 0.351 secs (94 Kbytes/sec) ftp> bye 221 Goodbye.

ncftp(1) (pronounced nick-f-t-p), is a more feature rich successor to ftp, supporting tab completion and recursive retrieval. It automatically connects to a server as the anonymous user, unless you specify a different username on the commandline with the -u argument. The primary advantage over ftp is the ability to send and retrieve multiple files at once with the "mput" and "mget" commands. If you pass the -R argument to either of them, they will recursively put or get data from directories. darkstar:~# ncftp ftp.osuosl.org Logging in... Login successful. Logged in to ftp.osuosl.org. ncftp / > cd pub/slackware/slackware-current Directory successfully changed. ncftp ...ware/slackware-current > mget -R isolinux isolinux/README.TXT: 4.63 kB 16.77 kB/s isolinux/README_SPLIT.TXT: 788.00 B 5.43 kB/s isolinux/f2.txt: 793.00 B 5.68 kB/s isolinux/initrd.img: 13.75 MB 837.91 kB/s isolinux/iso.sort: 50.00 B 354.50 B/s isolinux/isolinux.bin: 14.00 kB 33.99 kB/s isolinux/isolinux.cfg: 487.00 B 3.30 kB/s isolinux/message.txt: 760.00 B 5.32 kB/s isolinux/setpkg: 2.76 kB 19.11 kB/s ncftp ...ware/slackware-current > bye

The last client we're going to look at is lftp(1). Like ncftp, it supports tab completion and recursive activity, but has a more friendly license. Rather than user "mget" and "mput", all recursive operations are handled with the "mirror" command. "mirror" has many different options available, so I'll have to refer you to the man page and the built-in "help" command for complete details. darkstar:~# lftp ftp.osuosl.org lftp ftp.osuosl.org:~> cd /pub/slackware/slackware-current cd ok, cwd=/pub/slackware/slackware-current lftp ftp.osuosl.org:/pub/slackware/slackware-current> mirror isolinux Total: 2 directories, 16 files, 1 symlink New: 16 files, 1 symlink 14636789 bytes transferred in 20 seconds (703.7K/s) lftp ftp.osuosl.org:/pub/slackware/slackware-current> bye

Ready to see something cool? Have you ever found yourself needing just a handful of files from a large directory, but you're not entirely sure which files you already have and which ones you need? You can download the entire directory again, but that's duplicating a lot of work. You can pick and chose, manually check everything, but that's very tedious. Perhaps you've downloaded a large file such as an ISO, but something went wrong with the download? It doesn't make sense that you should have to pull down the entire file again if only a few bits have been corrupted. Enter rsync(1), a fast and versatile copying tool for local and remote files. rsync uses a handful of simple, but very effective techniques to determine what needs to be changed. By checking file size and time stamps, it can determine if two files are different. If something has changed, it can determine what bytes are different, and simply download that handfull of data rather than an entire file. It is truly a marvel of modern technology. In its simplist form, rsync connects to an rsync protocol server and downloads a list of files and directories, along with their sizes, timestamps, and other information. It then compares this to the local files (if any) to determine what it needs to transfer. Only files that are different will be synced. Additionally, it breaks up large files into smaller chunks and compares those chunks using a quick and simple hash function. Any chunks that match are not transferred, so the amount of data that must be copied can be dramatically reduced. rsync also supports compression, verbose output, file deletion, permission handling, and many other options. For a complete list, you'll need to refer to the man page, but I've included a small table of some of the more common options. ArgumentExplaination -v Increased verbosity -c Checksum all files rather than relying on file size and timestamp -a Archive mode (equivilant to -rlptgoD) -e Specify a remote shell to use -r Recursive mode -u Update - skip files that are newer on the receiving end -p Preserve permissions -n Dry-run - perform a trial run without making any changes -z Compress - handy for slow network connections Due to the power and versatility of rsync, it can be invoked in a number of ways. The following two examples connect to an rsync protocol server to retrieve some information and to another server via ssh to encrypt the transmission. darkstar:~# rsync -avz rsync://ftp.osuosl.org/pub/slackware/slackware-current/ \ /src/slackware-current/ darkstar:~# rsync -e ssh ftp.slackware.com:/home/alan/foo /tmp/foo