Department of Mathematics - University of Utah

HomeComputingCourse SchedulesCSMECurrent PositionsFAQ (Computing)FormsGraduateHigh SchoolLecture VideosMailbox AccessMath BiologyMath EducationNewsletterPeopleResearchRTG GrantsSeminars

Web access FAQ

Last update(s): Sat Sep 24 06:47:07 2005     Wed Oct 26 15:55:18 2005     Sat Sep 23 05:16:49 2006     Tue Aug 18 15:53:36 2009     Thu Mar 23 14:07:35 2017                Valid HTML 4.0!

Table of contents

  1. What GUI Web browsers are available?
  2. What text Web browsers are available?
  3. How can I download a Web page outside a browser?
  4. What Web browser versions are available?
  5. Where is the campus Web page?
  6. Where is the departmental Web page?
  7. How do I report problems about departmental Web pages?
  8. How to I create my own Web pages?
  9. What should I know about writing Web pages?
  10. Where are the departmental Web pages stored?
  11. My browser fails to start, complaining about a lock. Why?
  12. Why can the firefox browser not access FTP sites?
  13. Why can I not reach University of Utah library databases from off campus?
  14. How do I get a Java console in firefox-3?

Questions and answers

  1.   What GUI Web browsers are available?

    For use on a workstation with graphical user interface (GUI), we recommend the firefox browser, or its still-experimental, but reasonably stable, second generation, firefox2. The newer version has the convenient feature that, after a crash, it can restart with all previous windows intact. These browsers are the ones under most active development, with new releases appearing every two to three months.

    Other GUI browsers include amaya, arena, chimera, grail, hotjava, iexplorer, konqueror, mozilla, netscape, opera, seamonkey, xchimera, and xmosaic. Most of these are no longer actively developed, and the older ones support neither Java nor JavaScript, nor newer features of HTML, such as the style sheets used in this FAQ for colored highlighting, so they are mainly of historical interest. Nevertheless, they still useful for testing portability of Web page formatting.

    Of these browsers, all but konqueror are available on Sun Solaris SPARC systems. That browser is available in GNU/Linux on AMD64, IA-32, and IA-64 platforms.

    mozilla, netscape, opera, and seamonkey can also function as e-mail clients, as described in the electronic mail FAQ .

    Apple Mac OS X systems have the vendor-supported safari browser, but it is tied to the native window system of that platform, and does not work in an X11 Window System environment, or on other operating systems.

    Microsoft Windows systems have the vendor-supported iexplorer, but because it continues to be a source of very serious security holes that lead to viruses, worms, and filesystem corruption, we strongly recommend that you avoid it for anything other than running the Windows update procedure from the Tools menu. Our local Microsoft Windows systems also generally have at least firefox, mozilla, netscape, opera, and seamonkey installed.

    In Fall 2006, Microsoft announced a major new release of iexplorer, version 7, that they claim addresses the security problems of its predecessor. You can recognize it by its new icon, the blue letter e surrounded by a Saturn-like ring from lower left to upper right. It remains to be seen whether the vendor is successful in improving browser security.

    On all systems, it is important to have multiple browsers, because there are Web sites that work correctly with only certain browsers, and rarely, one finds a site that refuses to even display a page if the client browser is not iexplorer. The opera browser has the convenient feature of being able to masquerade as several other browsers, foiling such poorly designed Web sites.

  2.   What text Web browsers are available?

    Most of our systems have elinks, links, lynx, netrik, and w3m text browsers.

    While less convenient to use than GUI browsers, because they work in an ordinary terminal window, text browsers can be used when a window system is not available, and they cannot display images, sound, or popups. This makes them ideal for viewing Web pages at sites that are riddled with advertising and other unwanted distractions. Because they do not support client-side program execution with Java, JavaScript and others languages, text browsers are much safer than GUI browsers.

    In addition, most can dump the contents of a Web page in plain text, with or without links to URLs. This makes them very useful for batch processing, such as in shell scripts.

    Despite working in a terminal window, elinks and links recognize the mouse for link selection.

    lynx and w3m make extensive use of color to making browsing easier.

  3.   How can I download a Web page outside a browser?

    The wget utility can fetch a single Web page with a URL (uniform resource locator) prefixed with either http: or ftp:. It has options to recursively follow, and fetch, links to a specified depth, making it possible to download entire Web trees into a local filesystem.

    The ncftpget utility can fetch Web pages with URLs prefixed with ftp:. It supports filename wildcards, so you can, for example, fetch a software distribution and its detached digital signature file like this:

    % ncftpget ftp://ftp.gnu.org/gnu/which/which-2.16.tar.gz\*
    

    Notice the required protecting backslash preceding the asterisk, preventing its interpretation by the Unix shell. You can use wget that way as well.

    ncftpget can also recursively fetch a file tree:

    % ncftpget -R ftp://ftp.gnu.org/gnu/which/
    ./which-2.11.tar.gz:         102.34 kB  160.78 kB/s
    ./which-2.12.tar.gz:         102.38 kB  163.67 kB/s
    ./which-2.13.tar.gz:         103.78 kB  166.22 kB/s
    ./which-2.14.tar.gz:         109.58 kB  176.06 kB/s
    ./which-2.16.tar.gz:         122.75 kB  197.49 kB/s
    ./which-2.16.tar.gz.asc:     307.00 B     2.75 kB/s
    

    On Berkeley Unix systems (FreeBSD, NetBSD, and OpenBSD), the ftp program can fetch URLs prefixed with either http: or ftp:. On OpenBSD, ftp also handles filename wildcards.

  4.   What Web browser versions are available?

    You can find the filesystem location of any command with the which command, or the locally written mwhich and pathfind commands:

    % which firefox
    /usr/local/bin/X11/firefox
    
    % mwhich firefox
    981281 lrwxrwxrwx 1 jones staff 15 Sep 22 15:29 /usr/local/bin/X11/firefox -> firefox-1.5.0.7
    1196120 -rwxr-xr-x 1 root root 5159 Sep 11 14:13 /usr/bin/firefox
    
    % pathfind PATH firefox
    /usr/local/bin/X11/firefox
    
    % pathfind -a PATH firefox
    /usr/local/bin/X11/firefox
    /usr/bin/firefox
    

    When multiple pathnames are shown, they are in the order found in the PATH variable: the first one found is the one that is used when you invoke a command by its bare filename.

    Many of the browser commands are symbolic links to version-specific names, which you can see either with the mwhich command, as in the example, or else like either of these:

    % file /usr/local/bin/X11/firefox
    /usr/local/bin/X11/firefox: symbolic link to `firefox-1.5.0.7'
    
    % ls -l /usr/local/bin/X11/firefox
    lrwxrwxrwx 1 jones staff 15 Sep 22 15:29 /usr/local/bin/X11/firefox -> firefox-1.5.0.7
    

    Having located the directory in which the browser resides, you can now use wildcards to see if other versions are available with identifying suffixes:

    % ls /usr/local/bin/X11/firefox*
    /usr/local/bin/X11/firefox          /usr/local/bin/X11/firefox-1.5.0.2
    /usr/local/bin/X11/firefox-1.0.4    /usr/local/bin/X11/firefox-1.5.0.4
    /usr/local/bin/X11/firefox-1.0.6    /usr/local/bin/X11/firefox-1.5.0.7
    /usr/local/bin/X11/firefox-1.0.7    /usr/local/bin/X11/firefox-2.0b1
    /usr/local/bin/X11/firefox-1.5      /usr/local/bin/X11/firefox-2.0b2
    /usr/local/bin/X11/firefox-1.5.0.1  /usr/local/bin/X11/firefox2
    

    You can run a specific one by filename, or by complete pathname. Each new version generally contains important security patches, so you should normally run the latest release. However, we preserve earlier versions so that they can be used if problems show up in a new release.

    With the bash, tcsh, and zsh shells, you can also request completion of a command by typing one or two tab characters:

    % fire<TAB>fox<TAB><TAB>
    firefox             firefox-1.5.0.1     firefox-1.5.0.7
    firefox-1.0.5       firefox-1.5.0.2     firefox2
    firefox-1.0.6       firefox-1.5.0.3     firefox-2.0-a3_1.1
    firefox-1.0.7       firefox-1.5.0.4     firefox-2.0b1
    firefox-1.5         firefox-1.5.0.5     firefox-2.0b2
    
    % firefox-2<TAB>.0<TAB><TAB>
    firefox-2.0-a3_1.1  firefox-2.0b1       firefox-2.0b2
    

    The cursor is left at the end of the line, allowing you to type more characters. The command is not executed until you finally type a carriage return.

    Most GUI Web browsers have a menu path Help -> About that pops up a panel showing the version number and other information. Some of the modern ones also have a Help -> Check for updates menu path that you can use to see if a newer version is available, although you can install such a version only on a personally owned machine for which you have administrative privileges.

    Our local systems staff generally ensure that new versions are installed within a day or two of their announcement, so you can count on having up-to-date browsers.

  5.   Where is the campus Web page?

    http://www.utah.edu/

  6.   Where is the departmental Web page?

    http://www.math.utah.edu/

  7.   How do I report problems about departmental Web pages?

    Send e-mail to webmaster.

  8.   How to I create my own Web pages?

    1. Learn something about HTML (HyperText Markup Language); there are plenty of books and online tutorials on the subject. Although there are scores of HTML markup commands, and hundreds of HTML entities for special characters, you can get by nicely with a dozen or so commands, such as those used on this page. While you can use special HTML editors to prepare your input, the sad reality is that most produce nonconforming, and horridly formatted, HTML files. Using HTML markup directly is a useful discipline that helps you structure both your thoughts, and your documents.
    2. If you have no Web directory yet, run the local command websetup. It will create your top-level public_html directory under which all of your personal Web pages reside, set the permissions of that directory and your login directory to allow access from the Web, and create a template index.html file that is the index to the visible part of your Web tree.
    3. To ensure maximal readability by others, check your Web pages for strict conformance to one of several HTML grammar levels. The html-pretty utility can fix up many problems, and make your HTML input easier to read and edit, and the html-check and html-ncheck tools can do strict grammar checks.
    4. The W3C HTML 4.0 check image on the timestamp line under the title of this page illustrates how you can indicate conformance to an HTML grammar level, and also get remote validation of a Web page with a single button press on the image. Follow the View -> Page Source menu path in your browser to see the actual HTML code.
    5. Remember that you have (almost) no control over the final appearance of HTML markup at your readers' sites: if you want your Web pages to be useful to others, do not make assumptions about screen size, font size, or color. The simplest Web page format is likely to be the most widely usable.
    6. If you really need to control the remote appearance of Web material (e.g., for mathematics) then post PDF documents instead of HTML ones.
    7. Avoid posting documents in data formats that are not universally available on all platforms (e.g., desktop office software proprietary formats for word-processor documents, spreadsheets, and slides). It is usually possible to save such documents in HTML or PDF form instead, ensuring wider availability to your readers.
    8. Faculty, staff, and graduate students may on request have their Web page indexed from the People page linked from the departmental Web page. If you want this feature, send e-mail to webmaster with your request.
  9.   What should I know about writing Web pages?

    As you create new files in your Web tree, remember five important things:

    1. You are entirely responsible for their content, and for adhering to University of Utah policies about Web content. Your personal and scientific reputation may be affected by your Web content and quality. See the University World Wide Web Resources Policy.
    2. Only files underneath your $HOME/public_html directory are visible on the Web. Files elsewhere in your login directory tree are hidden from the Web, and visible to local users only according to the file- and directory-permission settings that you choose.
    3. Web files must have read access for everyone: use the chmod a+r file(s) command to ensure that. In addition, every directory in the URL path must have execute access: use chmod a+x dir(s) to set that permission.
    4. Web search engines will index your Web tree, usually at intervals of every few weeks, so any Web page in your tree that is reachable from a URL listed anywhere in the world will be ultimately be indexed. This is generally useful, but some users have been surprised by it. Once a document has been indexed, it cannot be recalled or hidden: you have just achieved long-term world-wide royalty-free publication.
    5. As long as a world-readable index.html file exists, the Web server will not permit directory listing by a remote Web client. This lets you further control visibility of files in your Web tree. Files whose URLs are not mentioned in at least one index.html file will not be visible on the Web, but could still be readable if their URL is known. You can use this feature to exchange (semi)private documents with a limited circle of friends and colleagues. However, if any of them reveal the URL to others, the documents are now public, and can suddenly become part of a search-engine database.
  10.   Where are the departmental Web pages stored?

    They reside under /home/www/htdocs/, a very large directory tree that also includes the entire FTP tree with hundreds of thousands of files.

  11.   My browser fails to start, complaining about a lock. Why?

    Most Web browsers (and most e-mail clients) record session information in files stored in a hidden directory tree underneath your login directory, such as .microsoft, .mozilla, .mozilla/firefox, .netscape, and .opera. If multiple instances of a browser run at the same time, file corruption from inconsistent file updates is very likely.

    To prevent this disaster, the browsers create an empty file with a known name, such as lock, somewhere in that directory tree. An attempt to start a subsequent instance of the browser will fail with an error report when the lock is found. Unfortunately, browsers are complex programs that, sadly, crash all too frequently, often leaving locks in place. For example, the firefox browser creates a lock file that you can find and diagnose like this:

    % find ~/.mozilla -name '*lock*'
    /u/ma/jones/.mozilla/firefox/r5z7z6jo.tmp1/lock
    
    % file /u/ma/jones/.mozilla/firefox/r5z7z6jo.tmp1/lock
    /u/ma/jones/.mozilla/firefox/r5z7z6jo.tmp1/lock: broken symbolic link to `155.101.96.134:15034'
    
    % grep  55.101.96.134 /etc/hosts
    155.101.96.134  sunfire.math.utah.edu sunfire
    

    Here, the lock file is a symbolic link to a nonexistent file named with the numeric Internet address and the process number of the browser. You can find out the hostname as indicated, and then use the ps command on that system to see whether the browser process is still running:

    % ps -e | grep 15034
    15034 pts/4   12:43 firefox-
    

    If the process is not running (i.e., no output from the ps command) then you can safely remove the lock, and then restart your browser:

    % rm /u/ma/jones/.mozilla/firefox/r5z7z6jo.tmp1/lock
    
    % firefox &
    

    However, if the browser is still running, you should either reuse it on your desktop, or if it is attached to another session, kill it gracefully like this:

    % kill -HUP 15034
    

    Browser termination and lock cleanup is clearly a messy and error-prone task, so a locally written program can often do the job for you, after which, you can restart the browser:

    % cleanupnetscapelocks
    
    % firefox &
    
  12.   Why can the firefox browser not access FTP sites?

    Sites that provide anonymous FTP access typically require a password that is a valid e-mail address, and firefox instead provides an invalid one. Worse, it gives you no way to set a correct one via the normal Edit -> Preferences menu path. This seems to be a bug!

    The error message that you see is in a pop-up window that reports The response `mozilla@example.com' is not valid ... Please use your e-mail address as your password.

    With some care, you can fix the firefox preferences file, either by editing it directly, or by setting it in the GUI, as described later. To edit preferences manually, first, exit from the browser (otherwise, it will just destroy the preference changes that you are about to make). Then find its file named prefs.js with the Unix command find $HOME/.mozilla -name prefs.js. Visit the reported file in a text editor and add these two lines at the end, substituting your own username for USERNAME in the e-mail address:

    user_pref("advanced.mailftp", true);
    user_pref("network.ftp.anonymous_password", "USERNAME@math.utah.edu");
    

    Save your changes to that file, and restart firefox. Check your changes with a URL that points to an FTP site, such as our own ftp://ftp.math.utah.edu/pub site.

    You should also be able to see your changes listed in the preferences that are displayed when you visit the fake (and undocumented) URL about:config.

    You can change those preferences in the output window, though doing so is a bit tricky and nonobvious:

    1. Position the cursor on a preference line to select it for changing and click the right mouse button to get a small pop-up with menu items Copy Name, Copy Value, New, Modify (or Toggle, if the type is Boolean), and Reset.
    2. Select the Modify button to get a pop-up text-entry box, or the Toggle button to alternate between true and false.
    3. Enter the new value, or toggle the existing Boolean value.

    Sometime before you exit the browser, your prefs.js file will be updated with the changes that you just made.

    You can also use the New submenu to enter a preference name and a value in two successive pop-up boxes. You may find that easier than editing prefs.js directly, and it avoids the need to stop and restart firefox.

  13.   Why can I not reach University of Utah library databases from off campus?

    The University of Utah Marriott Library provides Web access to a substantial number of commercial document databases, but must pay for those services. Consequently, Web access is restricted to members of the University of Utah, either requiring a special username and password, or now more typically, simply by checking that the client Web browser is running on a machine with a University of Utah Internet address (and thus, having a hostname ending in .utah.edu).

    It is, however, possible to obtain database access from remote sites by going through a campus intermediate connection between your computer and the database provider. This is called a Virtual Private Network, and details for setting up such a service are documented at the Campus VPN Service Web site.

    A second solution is a Web browser proxy. See the Marriott Library remote access Web page for instructions. This is, however, only useful if the database has a Web interface. A few instead use the Citrix system to create a window on your remote computer.

    A third solution is to run the Web browser on a campus Unix system, but direct the display to the remote computer connected via SSH (secure shell), as described in the login FAQ In addition, that computer must be running the X11 Window System, either natively on Unix (including GNU/Linux and Apple Mac OS X), or on Microsoft Windows with one of the Unix-like environments, as described in the software FAQ.

  14.   How do I get a Java console in firefox-3?

    The Tools -> Java console menu path in firefox-1 and firefox-2 is not supported in firefox-3. You can create a Java console window either by running an older version of firefox (run the command ls /usr/local/bin/X11/firefox* to see what versions are available) after first exiting any newer version of the browser, or on Sun Solaris systems, by running the command ControlPanel & in a terminal window.

    A Web search for firefox-3 +"java console" should find many Web pages discussing this change, which was made by software developers outside this University.


Dept Info Outreach College of Science Newsletter

Department of Mathematics
University of Utah
155 South 1400 East, JWB 233
Salt Lake City, Utah 84112-0090
Tel: 801 581 6851, Fax: 801 581 4148
Webmaster


Entire Web                     Only http://www.math.utah.edu/