Copyright (c) 2004 Information Today
|In today’s world, a library’s presence on the Web ranks only slightly behind its building in shaping its user’s impressions. Library Web sites do an impressive job of representing the library in a positive and effective way. Breeding offers some of his observations and tips on issues that strike as essential elements of a library Web site.|
Over the course of the last 5 years or so, I've visited the Web sites of at least 10,000 libraries. One of my long-standing projects has been the lib-webcats online directory of libraries. This data-base-driven resource provides a way for researchers to find the sites of libraries and their online catalogs on the Web. I started this database in 1997, and released it to the public in May 1999. For the general public, lib-web-cats serves as a finding aid for libraries. But for my personal research, it works as a rich data source for library automation trends. Part of the information tracked in the database includes the current library automation system and any previous systems each library has used. Although some of the Web sites in lib-web-cats have been self-contributed by libraries, I personally review each entry and view each Web site referenced.
At my primary job at Vanderbilt, I participate in the Web Task Force, a group responsible for ongoing development and maintenance of our library system's Web site. For that group, we regularly scout the Web sites of other large academic libraries, using what others have done to inform our decisions as we approach a given issue or problem.
In today's world, a library's presence on the Web ranks only slightly behind its building in shaping its users' impressions. In the course of my excursions through multitudes of library Web sites, I've seen that the vast majority of them do an impressive job of representing the library in a positive and effective way. For a small minority of these sites, however, I've found it hard to find key bits of information, or I've experienced problems with basic site navigation. If I have these difficulties, I worry that their own library users are also not optimally served. This month, I offer some of my observations and tips on issues that strike me as essential elements of a library Web site.
URL Persistence. Help us find you on the Web! Create an easily remembered URL and stick with it. Changing your library's Web address should be done with the same level of care and frequency as that of its street address. The URL should stay the same even if the library changes physical Web servers, hosting services, Internet service providers, or page delivery applications.
I've seen lots of smaller libraries that use Web hosting services and take the URLs that come with those services. These URLs might not necessarily give the library a memorable Web address. Moreover, if the library changes hosting services, it is forced to find a new identity.
Today, registering a domain is cheap and easy, allowing a library to craft its own URL. Once the library has registered a domain name, it can be used as its online identity regardless of whether it hosts its own Web site, relies on its parent organization, or depends on a commercial Web hosting service. Libraries must be careful, however, to maintain registration of their domain names. Once registration of a domain has lapsed, it may be difficult to get the name back, again forcing an unwanted change of address.
One of the trends that I have seen involves public libraries selecting domains such as www.clevelandlibrary .org rather than those reflecting geographic conventions, like http://www .lib.Cleveland.tn.us. I find the name-oriented domains to be much easier to remember and type than the geographic ones.
A library home page, for example, might reside in a file called index.html. For my library, it would be possible to advertise this URL: http://www.library .vanderbilt.edu/index.shtml. But, with the proper configuration, the simpler http://www.library.vanderbilt.edu stands as our URL, completely independent of the actual file names involved. This principle applies to both the root directory of the Web server and subdirectories. Thus, I can advertise my personal Web page, which lives in a subdirectory of our staff Web server, as http://staffweb.library.vanderbilt.edu/ breeding, even though the actual page resides in a file called "index.html."
To this end, it's important to configure the Web server to deliver the correct Web page when no file name is provided. Without this configuration detail, many Web servers will list the files in the directory rather than deliver Web pages. Though the mechanics of how to implement this feature vary, all Web servers have the ability to deliver a default Web page within any directory if no particular page is specified. This approach allows the library to make changes, such as moving from static HTML pages to an environment that uses a scripting language like ASP or Perl to deliver its pages. Changing the underlying technologies might mean that the actual page would change from "index.html" to "index .asp." If the library is able to avoid requiring page-name specification by relying on the Web server to deliver the default page, these changes in technologies do not have an impact on the library's identity.
Once the Web server has been configured to use the default Web page, it's important to advertise and link to the simpler form of the URL. I've seen many libraries advertise their URLs in the page-specific format, even though the simpler address is fully enabled and could be used instead. Again, by using the longer form on the URL, the library adds external link repair issues when a file name changes.
Contacting the Library. It is important to provide the means for site visitors to send e-mail queries to library staff. Two basic options prevail. One involves offering a mail-to link, which consists of a clickable e-mail address that is automatically pasted into the visitor's mail client. The other approach uses a Web form that allows one to send a message to the library directly from the Web site. In most cases, the former approach is more convenient to site visitors because it allows them to use their own mail clients. The latter approach, however, is becoming more common because libraries are reluctant to expose an e-mail address for (the quite valid) fear of becoming the recipient of a barrage of spam.
If the library provides a mail-to link for general inquiries, it should use a generic e-mail address not directly associated with a particular staff member, such as email@example.com. It seems ill-advised to advertise the e-mail address of a particular staff member as a general point of contact, since the actual person who responds may change day-to-day (though having a directory of all staff e-mail addresses is well-appreciated). A generic address will not have to be changed as the library experiences staff turnover. I've observed that when I send e-mail to a mail-to link that is obviously associated with a library staff member, the chance of that mail bouncing is high. I'm also unpleasantly surprised by how often a message I send to an address provided for general inquiries goes unacknowledged. If a library posts an email address for inquiries, it should ensure that it works correctly and is regularly monitored.
Don't Overlook the Basics. It is important to present the basic facts about your library in an obvious location on your Web site. It's very frustrating to have to look through many layers of a library's site just to find its address! I have encountered some Web sites of large libraries that have hundreds of pages, yet omit this key piece of information. The majority of libraries, fortunately, include their full address and telephone number in the footer that appears on each Web page. I've seen some sites that provide maps and driving directions to the library, but do not give the mailing address.
Another frustration is including information in a graphic, but not as text on the Web page. If a patron wants to send mail to the library-perhaps to pay a fine-it's convenient for him or her to be able to copy and paste the mailing address into a word processor. Presenting the address, or other information, graphically in the banner of the Web site makes it impossible to use standard copy-and-paste techniques to grab that information. Sometimes that also makes it hard to read.
These are some of the basic elements that I appreciate seeing on the Web site for even the smallest of libraries:
Larger libraries typically focus considerable effort on providing access to and assistance with their collections of electronic resources. Features I expect on a site for such a library expand significantly:
I find that, increasingly, most large libraries go well beyond these basic elements, and that they are evolving into feature-rich Web portals that offer their users both information resources and library services.
Avoid Unnecessary Frills. While any Web site should do all that it can to be attractive and interesting, some features interfere with finding information quickly and easily. Here are some of the features I occasionally encounter on library Web sites that I find especially problematic:
When traveling, you always seem to remember the things that went wrong, even when the overall trip was an overwhelmingly positive experience. I think that it's much the same when browsing the Web. My general impression is that libraries put more thought, creative energy, and effort into their Web sites than do other types of organizations. Only a very small minority of Web sites lack some of what I consider to be essential elements. I know that readers of Computers in Libraries are the least likely to commit such sins of omission!
Please visit http://www.librarytechnology.org/libwebcats to be sure that your library is listed and that its information is complete and current. While visiting, feel free to send me a note. Comments on any of my columns are appreciated, as are suggestions for future topics.
|Type of Material:||Article|
Computers in Libraries|
|Volume 24 Number 2|
|Systems Librarian Column|
|Last Update:||2012-12-29 14:06:47|
|Date Created:||0000-00-00 00:00:00|