For assignments and projects, we will work on a unix (actually linux)
machine called capricorn.informatics.indiana.edu.
Accounts will be or have been created on capricorn for registered students. To log
into capricorn, use your IU username (aka network ID) and password. For
example if your email is janedoe@indiana.edu, then your
username is janedoe.
To log into capricorn, you must use ssh. IU has various ssh clients depending on
what machine you are working from. From a unix machine or terminal
(eg, the Terminal in Mac OS X) you would simply type ssh
capricorn.informatics.indiana.edu, then enter your username and
password as prompted. Note: if your client unix username (eg, your Mac
OS X username) does not match your capricorn username (IU network ID), then
specify your username by typing, say, ssh -l janedoe
capricorn.informatics.indiana.edu. At the end of your session you
would type exit to logout. From a windows machine, you will
enter the capricorn hostname in the GUI of an ssh client; everything
else is the same.
Once logged into capricorn, you use the unix shell to enter commands and edit, debug, and run your script files. We will give a brief primer of unix in class but unless you are already familiar with unix you are strongly encouraged to consult one of many excellent online tutorials, an IT Training self-study, or books.
In unix, there are several editors you can use. The easiest, but least
powerful is nano, a clone of
pico.
More powerful alternatives are vi and emacs. You are strongly
encouraged to learn either vi or emacs, they are worth the steep
learning curve. A much less efficient alternative is to use a good code
editor with which you are familiar on your client machine, and then
transfer your file(s) to capricorn using scp or sftp. Debugging is very
inconvenient with this approach.
Sure. Edit an HTML file named index.html in the
public_html subdirectory of your home directory.
Make sure the file is world-readable. You can then access this page at
http://capricorn.informatics.indiana.edu/~myusername/.
In class we will briefly discuss how to write, compile, debug, test and run your perl scripts. Unless you have some prior programming experience with perl, you are strongly encouraged to attend one of the STEPS workshops on perl offered by IT Training & Education especially for this course. Because they are partially funded by the Student Technology Fee, these workshops are free for enrolled IU students (present your IU student ID). You will also receive a set of workshop materials at no additional charge. You can obtain the materials for free even if you cannot attend the workshop, by presenting your IU student ID.
All the perl modules we are likely to need for the class assignments and project are already installed on capricorn. If you think you need an additional module, post a message on the Oncourse Discussion. There is likely to be a simpler alternative, otherwise I will discuss the module installation with our sysadmin.
Name the CGI script
using the extension cgi, for example
myscript.cgi, and place it in the
/var/www/cgi-bin/myusername directory. Make sure
that the script has the appropriate permissions: it should
be world-readable and world-executable (755).
Test your script locally (from the shell) first, to make sure it
compiles and runs without errors before it is run over the Web. Finally,
point a browser to
http://capricorn.informatics.indiana.edu/cgi-bin/myusername/myscript.cgi.
You can also place static HTML pages in /var/www/html/myusername,
for example pages containing forms whose actions are CGI scripts.
Such a page would then be accessible at a URL like
http://capricorn.informatics.indiana.edu/myusername/mypage.html.
Finally, if your script needs to write and read data files, these files should be
located in /var/www/data/myusername. Make all of your files
group-readable and group-writable so that Fil can help debugging if needed.
It may happen that you launch a script that is either very slow, or
never terminates due to a bug. If you run such a script from the
browser (see above advice about testing locally first), the script
will be executed by the Web server (apache) and so you cannot directly
kill its process like you would normall kill processes that you
run from the shell. This is because apache would own the process, not you.
To allow you to kill such a script, the sysadmin placed a kill_proc
script in https://capricorn.informatics.indiana.edu/cgi-bin/kill_proc/kill_proc.pl.
Find the process ID (PID) of the apache-spawned process you want to kill by running ps or
top from the shell. Then run the kill_proc script from your browser,
entering the PID. This script requires authentication, and we log its use; any abuse
(eg, killing somebody else's process) will not be tolerated.
For unix commands, use man (type man
command at the prompt). For perl, use perldoc module
or perldoc -f function (perldoc is also online). Google can also be very helpul, of
course. Finally, in case it is not obvious from this brief introduction,
the IU Knowledge Base is a
fantastic resource. Use it!