Difference between revisions of "Content:CS Dept Site Setup"

From Earlham CS Department
Jump to navigation Jump to search
(→‎Checking Out)
 
(11 intermediate revisions by 2 users not shown)
Line 1: Line 1:
−
== Locations ==
+
Both versions of the department's website run off [[Servers:Quark|Quark]].
−
* The department's website runs off [[Servers:Quark|Quark]].
 
−
* The live copy is located at <code>/clients/www</code>
 
−
** The html itself and the root for the website is located at <code>/clients/www/html</code>
 
−
* This is automatically copied over from CVS:
 
−
** CVS for the website is located at <code>/clients/users/cvs/html</code>
 
−
** Committing to CVS will automatically cause the changes to be copied to <code>/clients/www/html</code>
 
  
−
== Using CVS ==
+
=== Staging ===
−
''Note:'' When running any of these commands, you'll need to specify <code>-d /clients/users/cvs</code> to specify which cvsroot you're talking about. 
+
The staging server can be accessed at staging.cs.earlham.edu.  This contains a copy of everything on the website and everything that's currently being worked on.  We'll use this as a place to make changes before we commit them to going live.
 +
'''Note:''' This is the only place that you should make changes to the website!!
  
−
Otherwise you can also put a line in your .bashrc file like this: <code>export CVSROOT='/clients/users/cvs'</code>
+
The files for the staging site are located on quark at <code>/clients/www/staging/html</code> and <code>/clients/www/staging/cgi-bin</code>.
  
−
=== Checking Out ===
+
To access staging from a web browser, go to http://staging.cs.earlham.edu/ You'll be prompted for the group's username and password.  If you've forgotten these or haven't gotten them yet, talk to one of the other content group members.
−
When you're checking out a part of the branch with CVS, be sure to specify that part.  It should always start with 'html' if you're working with the website content. Some of the important directories are here:
 
−
* <code>pages</code>
 
−
* <code>projects</code>
 
−
* <code>css</code>
 
−
* <code>images</code>
 
−
* <code>contentadmin</code>
 
−
* <code>news</code> - this might still be in use
 
  
−
These ones appear to be old and currently unused, but I might be wrong:
+
=== Live ===
−
* <code>Attic</code> - is empty
+
After we've reviewed changes to a page or to a project, we can copy them over to the live website by using rsync.  This can be either just files or entire directories.  Immediately after, commit the changes to the CVS tree by running
−
* <code>MoL</code> - this currently redirects to http://molweb.org/
+
 
−
* <code>about</code> - about the CS department - not sure this is still being used
+
:<code>cvs commit</code>
−
* <code>applied-groups</code> and <code>appliedgroups</code> - not sure if these are being used or not
+
 
−
* <code>barn</code> - has some old job descriptions and other strange things
+
from inside the <code>html</code> directory.
−
** This actually is probably the barndoor project, which we should probably fix, used to turn e-mail addresses to png until clicked on by an oncampus IP
+
 
−
* <code>contentadmin</code> - I think only the index.html file is being used here, the other bits look out-of-date
+
The files for the live website are located on quark at <code>/clients/www/html</code> and <code>/clients/www/html/cgi-bin/</code>.
−
* <code>courses</code> - this is old and uses the old CSS (but still available online)
+
 
−
* <code>elections</code> - not available online
+
=== Cruft ===
−
* <code>email</code> - available online and broken
+
 
−
* <code>esslli07mts</code> - old workshop by Jim, need to see if this still is important
+
There's a lot of currently unused stuff in <code>/clients/www</code> and also in the <code>html</code> directory.  Some of this is currently part of CVS.  Eventually we should move over to SVN and only put in there stuff that's actually used.
−
* <code>little-fe</code> - redirects to littlefe.net
+
 
−
* <code>mol</code> - empty
+
* [[Content CVS Directories]] - a breakdown of the directories and what's in them
−
* <code>mt</code> - Movable Type install, not working
+
 
−
* <code>projects</code> - this is part of the old website
+
=== Accessing Quark from Home ===
−
** <code>projects/barndoor</code> - the broken mail to pngs stuff
+
==== Shell Access ====
−
** <code>projects/robotics</code> - a senior seminar from awhile ago
+
You can get remote access to a command line on quark remotely using SSH.  If you're running Mac OS X or Linux, you already have a built-in shell that you can use.  For Windows, you'll need to download Putty:
−
* <code>search</code> - outdated application for 2008-2009 faculty member
+
 
 +
http://www.putty.org/
 +
 
 +
You don't need to install it: it's entirely a self-contained .exe file.  Enter <code>quark.cs.earlham.edu</code> for the hostname. Use your ACL username and password.
 +
 
 +
==== Transferring Files ====
 +
You can transfer files back and forth from Quark using SCP (a file transfer protocol that runs over SSH).  Download WinSCP:
 +
 
 +
http://winscp.net/eng/download.php
 +
 
 +
=== Changing the News Section ===
 +
You update the news blog --- just add a new entry and mark it for
 +
publication when you save it --- and then rebuild the blog.  The second
 +
part is the part that is most easily missed.  You have to select rebuild
 +
from the menu on the left side of the blog page and then something like
 +
"rebuild all" from the pop-up.  There is a cron script that runs every
 +
five minutes or so which checks to see if the rss file, which will be
 +
updated when you rebuild the blog, has been modified more recently than
 +
the html include file and regenerates the html if it has.  So there will
 +
generally be a delay between the rebuild of the blog and the update
 +
showing up in the web.
 +
 
 +
The jobs blog works the same way.
 +
 
 +
The cron job is
 +
:<code>/clients/www/html/cgi-bin/feed2page.pl -f /clients/www/html/cgi-bin/feeds.list</code>
 +
 
 +
feeds.list has the format, one per line:
 +
 
 +
:<code><filename.rdf>|<filename.html></code>
 +
 
 +
where <filename.rdf> is the pathname of the the rss file and
 +
<filename.html> is the pathname of the html include file.

Latest revision as of 11:50, 10 February 2009

Both versions of the department's website run off Quark.

Staging

The staging server can be accessed at staging.cs.earlham.edu. This contains a copy of everything on the website and everything that's currently being worked on. We'll use this as a place to make changes before we commit them to going live. Note: This is the only place that you should make changes to the website!!

The files for the staging site are located on quark at /clients/www/staging/html and /clients/www/staging/cgi-bin.

To access staging from a web browser, go to http://staging.cs.earlham.edu/ You'll be prompted for the group's username and password. If you've forgotten these or haven't gotten them yet, talk to one of the other content group members.

Live

After we've reviewed changes to a page or to a project, we can copy them over to the live website by using rsync. This can be either just files or entire directories. Immediately after, commit the changes to the CVS tree by running

cvs commit

from inside the html directory.

The files for the live website are located on quark at /clients/www/html and /clients/www/html/cgi-bin/.

Cruft

There's a lot of currently unused stuff in /clients/www and also in the html directory. Some of this is currently part of CVS. Eventually we should move over to SVN and only put in there stuff that's actually used.

Accessing Quark from Home

Shell Access

You can get remote access to a command line on quark remotely using SSH. If you're running Mac OS X or Linux, you already have a built-in shell that you can use. For Windows, you'll need to download Putty:

http://www.putty.org/

You don't need to install it: it's entirely a self-contained .exe file. Enter quark.cs.earlham.edu for the hostname. Use your ACL username and password.

Transferring Files

You can transfer files back and forth from Quark using SCP (a file transfer protocol that runs over SSH). Download WinSCP:

http://winscp.net/eng/download.php

Changing the News Section

You update the news blog --- just add a new entry and mark it for publication when you save it --- and then rebuild the blog. The second part is the part that is most easily missed. You have to select rebuild from the menu on the left side of the blog page and then something like "rebuild all" from the pop-up. There is a cron script that runs every five minutes or so which checks to see if the rss file, which will be updated when you rebuild the blog, has been modified more recently than the html include file and regenerates the html if it has. So there will generally be a delay between the rebuild of the blog and the update showing up in the web.

The jobs blog works the same way.

The cron job is

/clients/www/html/cgi-bin/feed2page.pl -f /clients/www/html/cgi-bin/feeds.list

feeds.list has the format, one per line:

<filename.rdf>|<filename.html>

where <filename.rdf> is the pathname of the the rss file and <filename.html> is the pathname of the html include file.

Retrieved from "https://wiki.cs.earlham.edu/index.php?title=Content:CS_Dept_Site_Setup&oldid=7453"