The Internet for Anthropologists (http://www.lib.ncsu.edu/staff/morgan/eric-talks-to-anca.html) Eric Lease Morgan, NCSU Libraries

This text is a written version of the end-user In short, client/server computing provides a presentation given to the Association of 8. Repeat mechanism for disparate computers to North Carolina Anthropologists (ANCA) cooperating on a single computing task. at the North Carolina State University This client/server interaction is a lot like Libraries April 22, 1995. Its purpose is to going to a French restaurant. At the Common Internet provide an overview of the Internet restaurant, you (the end-user) are presented (specifically the World Wide Web) and what a menu of choices by the waiter (the protocols it can mean for anthropologists. client). After making your selections, the waiter takes note of your choices, translates Most of the software developments taking place on the Internet have done so only in Client/Server Computing them into French, and presents them to the French chef (the server) in the kitchen. the past few years. Namely, those being wide area information servers (WAIS), To truly understand how much of Internet After the chef prepares your meal, the waiter returns with your diner (the results). gopher, and the hypertext transfer protocol works, you must understand the concept of (HTTP, or more commonly called the client/server computing. The client/server Hopefully, the waiter returns with the items you selected, but not always; World Wide Web or WWW for short). model is a form of distributed computing Before these newer protocols were where one program, the client, sometimes things get "lost in the translation." developed, the standard telnet, file transfer communicates with another program, the protocol (FTP), and electronic mail server, for the purpose of exchanging User-interface development is the most protocols were tools used most often.The information. The first piece of software is majority of these protocols are based on the client. Its responsibility is usually to: obvious advantage in client/server computing. Within this model it is client/server computing and therefore possible to create an interface to data require at least client and server software to • Handle the end-user interface make them useful. • Translate the end-user's requests into independent of the computing environment hosting the data. Therefore, the user the desired protocol Telnet is a way of logging on to a remote • Send the requests to the server interface of a client/server application can be written on a Macintosh and the server computer and using your computer as if it • Wait for the server's response were attached to the remote computer. • Translate the response into "human can be written on a mainframe. At the same time, clients could be written for When you telnet to a remote computer, readable" results your local computer is acting like a • Present the results to the end-user DOS- or Unix-based computers and access the same data from the same mainframe. "dumb" terminal. It is a lot like using your modem from home to dial into your The sever's functions include: Since the user interface is now the responsibility of the client, the server has campus computing centers. This tool was more computing resources to spend on originally used to share computing • Listening for client queries resources of a central site (like super • Processing those queries analyzing any queries and disseminating information. Here lies another advantage of computer resources) with researchers at • Returning the results back to the remote distances. Nowadays it is mostly client client/server computing; it tends to use the strengths of divergent computing platforms used to connect to the electronic card catalogs of libraries. Atypical client/server interaction goes like to create a more powerful applications. this: There is no reason why a Macintosh may not be used as a server except its FTP provides rudimentary file computing and storage capabilities are manipulation services on remote 1. End-user runs client software to computers. This includes the ability to create a query dwarfed by the mainframe's. The client/server model also provides the copy files from (get) remote computers as 2. Client connects to the server well as copying files to (put) remote 3. Client sends the query to the server opportunity to store information in a central location and disseminate that computers. FTP also provides the ability 4. Server analyzes the query to create and manipulate remote directory 5. Server computes the results of the information regardless of the remote computer. structures. Thus you can create and delete query directories on remote computers. 6. Server sends the results to the client 7. Client presents the results to the

______page 1 of 4 Wide Area Information Servers (WAIS) are For example, scholarly articles usually based VMS or Unix computers, and an indexing/searching/document- include multiple footnotes. If hypertext is computers running X Windows. distribution technology. The indexing part applied to a journal article, then the reader of WAIS creates pointers to specific could select footnotes from the article and The real power of these WWW clients collections of data just as Sociological be "transported" to the footnote. The (usually referred to as "browsers") is their Abstracts points readers to specific journal footnote, in turn could contain links and ability to understand multiple Internet articles. Unlike Sociological Abstracts, the process could go on indefinitely. protocols. Specifically, each of the WAIS also provides the means of retrieving Within this same system the reader has the browsers described here understand how to items from the collection using the opportunity to return to where they FTP files, act as gopher clients, as well as client/server model. Thus, after a collection originated. Since its inception, the read and interpret the output of HTTP of files (text files, software, data files, hypertext concept as embodied by HTTP servers. Additionally, each of these pieces graphics, et cetera ) has been indexed, a has to include descriptions of college and of software understand "forms", an HTML client program can be used to search the university departments, collections of extension allowing the end-user to index. The user's search strategy is then Internet resources, newspapers, other items complete electronic forms similar to sent to the server who processes the query of a non-scholarly nature, and just about gopher+ ASK blocks. While none of these and returns the results. The client then has anything else you can conceive. clients directly understand the telnet the opportunity to select items from the protocol, each one of these browsers can be results and retrieve the actual files. It wasn't until early 1993 when Bob configured to load and run your telnet McCool and Marc Andreessen, then of the software. Since WWW browsers take The gopher protocol, developed by a National Center for Super Computing URLs as input, and since URLs uniquely computing center at the University of Applications (NCSA) wrote both HTTP describe the location of various Internet Minnesota was the first Internet service to client and server applications that HTTP resources, then WWW browsers are the put a user-friendly front end on the really started to become more popular. tools that really turn the World Wide Web protocols outlined above. Using the Since the server application (httpd) was into a "world wide web". client/server model of computing, the available for many flavors of Unix, not gopher protocol allows a server just NextStep, the server application was As the name implies, MacWeb administrator to construct a menu of easily put to use by many people. Since (http://galaxy.einet.net/EINet/MacWeb/Mac Internet services including most of the the client application (Mosaic for X WebHome.html) is a WWW browser for items listed above as well a few others. Windows) supported graphics, as well as the Macintosh. Written at Microelectronics Furthermore, it allows the administrator to the WAIS, gopher, and FTP protocols, the and Computer Technology Corporation present these services in such a way that client application was head and shoulders (MCC), MacWeb is distributed via the the user simply selects items from a menu over the original CERN client application Enterprise Integration Network (EINet). and the server does the rest. Previous to in terms of aesthetic appeal as well as MacWeb requires System 7 and at least gopher, to telnet, FTP, read USENET functionality. Later, a more functional MacTCP version 2.0.2. Just about news, or search a WAIS index the user had terminal-based client (Lynx) was developed anybody using a Macintosh is using have client software for each of these by Lou Montulli then of the University of System 7. MacTCP is a operating system services. With the advent of the gopher Kansas and made HTTP accessible to the extension available from Apple Computer protocol and a well organized gopher server, lowest common denominator, vt100 based enabling your Macintosh to understand the these services were reduced to selecting terminals. Lastly, since NCSAlater Transport Control Protocol (TCP) options from a menu. released Macintosh and Windows versions necessary for Internet communications. A of Mosaic, HTTP became even more very important piece of software called The WWW began in 1989 as the brain- popular because an even wider audience "StuffIt Expander", is strongly child of Tim Berners-Lee, and first realized now had access to the 'Web. Since then recommended when using MacWeb as well in 1991 while working at CERN, a particle other client and server HTTP applications as MacMosaic. StuffIt Expander is a utility physics laboratory in Geneva, Switzerland. have been developed, but the real program used to translate and uncompress The WWW protocol, formally called the momentum was created by the developers files usually retrieved via FTP archives. hypertext transfer protocol (HTTP) was at NCSA. first intended as a means to share Besides its speed, elegant and easily information between members of the high Client Software customizable interface, the automatic energy physics community. The operative creation of HTML documents from its word describing HTTP is "hypertext" as Five examples of WWW client software are hotlists, MacWeb indirectly supports the originally described by Vannevar Bushand described here: MacWeb, Mosaic for WAIS protocol by launching MCC's WAIS coined by Theodor H. Nelson. In this Microsoft Windows, Lynx, Mosaic for X client, MacWAIS. system, text is presented to a reader with Windows, and Netscape. These particular "links" to other texts intended to provide pieces of software are described because Mosaic for Microsoft Windows more explanation of the original text. they presently represent the best clients for (http://www.ncsa.uiuc.edu/SDG/Software/ Scholarly journal articles represent an the most common operating systems: WinMosaic/HomePage.html) is bound to excellent application of this technology. Macintosh, Microsoft Windows, terminal- be one of the more popular WWW ______page 2 of 4 browsers since most people have or will server (httpd), really gave the WWW the protocol://host/path/file have Microsoft Windows-based computers. momentum and visibility it has today. [11] Mosaic for Microsoft Windows This browser supports copy and paste from "Protocol" denotes the type of Internet requires a winsock DLL. Like MacTCP, a the display. Direct WAIS support, and resource. The most common are: gopher, winsock DLL is software allowing your therefore URLs like wais, ftp, telnet, http (WWW), file, and computer to understand TCP. Common wais://wais.lib.ncsu.edu/alawon?nren are mailto (electronic mail). "Host" denotes the winsock packages include LANWorkplace valid. At the present time, just about the name or Internet Protocol number of the and Trumpet Winsock. Additionally, only thing it doesn't support is the mailto remote computer. Examples include: Mosaic for Microsoft Windows requires URL. One of the nicest features of Mosaic www.lib.ncsu.edu or 152.1.39.42. "Path" Windows extensions (Win32s) in order to for X Windows is its ability to directly is a directory or subdirectories on the take advantage of 32-bit applications. Since deliver a displayed document to somebody remote computer. "File" is the name of the almost all computers running Windows via email.The downside of Mosaic for X file you want to access. contain '386 microprocessors, and since the Windows is its requirement of a relatively Win32 software is available via FTP from high powered computer. While a Macintosh Using variations of this general form, you NCSA, then you will be able to use equipped with MacX or a Windows-based can use URLs and your WWW browsers to Mosaic for Microsoft Windows if you are computer with HummingBird can run X access just about any Internet resource. willing to install the necessary software. Windows terminal sessions, Mosaic for X Simple examples of URL's include: Windows really requires direct access to a One of the nicest features of Mosaic for Unix or VMS machine running the X 1. ftp://ftp.lib.ncsu.edu/pub/stacks/ Microsoft Windows is its menubar Windows software. alawon/alawon-v1n04 customizability. By editing the mosaic.ini 2. http://www.lib.ncsu.edu/stacks/ file, you can delete or add menu items to Netscape (http://www.mcom.com/) is the alawon-index.html the menu bar. Consequently you can most full-featured WWW browser available configure Mosaic and have it display to date. It smoothly integrates the FTP, The first example illustrates how your commonly used Internet resources for you gopher, and usenet news protocols into one WWW browser can be used to copy a file and your clientele. application. It has implemented many from a remote FTP server. It translates to: proposed extensions to the hypertext Lynx markup language (HTML). This means it 1. FTP to ftp.lib.ncsu.edu (ftp://ftp2.cc.ukans.edu/pub/WWW/lynx/) understands markup tags like 2. Log on as anonymous is a character-cell WWW browser, meaning

as well as table-creation 3. Change directories to it is intended to be used on DOS computers tags. Unlike some of the other browsers /pub/stacks/alawon/ or "dumb" terminals running with the Unix mentioned here, Netscape opens up 4. Get the file alawon-v1n04 or VMS operating systems. multiple connections to remote servers at a single time. This makes the Netscape Since WWW browsers understand and The flavors of Lynx are ideal clients in two browser seem faster than other browsers. implement the File Transfer Protocol cases. First, they are wonderful when your Netscape also implements a security (FTP), you do not have to remember all only Internet connection is located on a standard known as the secure sockets layer the commands necessary to do FTP. All remote computer, ie. most dial-in access. (SSL). By using this standard, the use of you have to remember is how to create a Secondly, these clients are excellent when Netscape, in conjunction with an SSL FTP-style URL. you need to provide a lowest common compliant server application, and send and denominator interface, ie. vt100 terminals. receive confidential information over the Similarly, the second example opens up a Internet with no possible breach in HTTP connection to www.lib.ncsu.edu, You won't get pictures with Lynx clients. security. changes directories to stacks, and retrieves Nor will you hear sounds. But the Lynx the file alawon-index.html. clients do support the "mailto" URL. Uniform Resource Mailto URLs are URLs specifying the URLs are more complicated than the Simple Mail Transfer Protocol (SMTP) or Locators general form illustrated above; URLs can Internet mail. Thus, when an end-user also provide the means to present the logon using a Lynx client selects a mailto URL, Uniform Resource Locators (URL) are a name for telnet connections, a then the end-user will be presented with a fundamental part of the WWW. They are communications port, an index/search "form" to complete and the resulting text used to concisely describe and identify the query, and/or an HTML anchor. from the form will be delivered via Internet protocol and location of Internet resources. mail to the person or computer specified in Presently, the most definitive document For example, a complete URL identifying the URL. describing URLs is called "WWW Names a telnet session would look something and Addresses, URIs, URLs, URNs". like: Mosaic for X Windows (http://www.ncsa.uiuc.edu:80/SDG/Softwa In general, a URL has the following form: telnet://[email protected]:23/ re/XMosaic/), coupled with NCSA's HTTP ______page 3 of 4 where "library" denotes the logon name and gopher://gopher.lib.ncsu.edu/00/library/abo then it is much safer to first visit the 23 denotes the communications port. (Port ut resource in question and copy resulting 23 is the standard telnet communications URL from your client's display. port.) Thus you can use your WWW The 0's denote text files. browser to initiate a telnet session. In summary, URLs unambiguously Specifically, the above example opens up a Anasty wrench gets thrown into the describe the location of Internet resources. telnet connection to "library.ncsu.edu" and business when the path and/or file names Using URLs as a standard, Internet-client depending on your browser, you may be of the Internet resources contain special programs like WWW browsers can reminded to log on as "library". This URL characters like spaces or colons. In these interpret URLs and retrieve the desired does not use the "path" nor "file" cases, escape codes must be used to denote information. URLs describe the protocols parameters because they are meaningless the special characters. For example, and locations of Internet resources for telnet sessions. independently of Internet-capable client gopher://gopher.lib.ncsu.edu/0ftp%3amrcne software. On the other hand, if you wanted to xt.cso.uiuc.edu%40/pub/etext/etext91/aeso manually query the Geographic Name p11.txt Server, then the URL would be: This long URL first specifies a gopher telnet://martini.eecs.umich.edu:3000/ server (gopher.lib.ncsu.edu) to FTP a file from mrcnext.cso.uiuc.edu, and get Since the Geographic Name Server requires aesop11.txt. Notice the "%3a" and "%40" no password, no password is specified, but in the URL. They are used to denote a since the Geographic Name Server "listens" colon and at-sign (@), respectfully. on port 3000, a non-standard port number Furthermore, notice the zero proceeding the must be specified. "ftp". Here again, this is used to identify the remote file as a text file. In short, WAIS searches can be specified using gopher URLs are particularly difficult to URLs. Unfortunately, at the present time, decipher. If you must communicate the only Mosaic for X Windows directly location of a gopher resource via a URL, implements the WAIS protocol. WAIS URLs have the following form: Anthropology-related Internet resources wais://host:port/database?query There are quite a number of anthropology-related Internet resources available. They where "port" is assumed to be 210, the run the gamut of FTP sites to WWW servers. Some of the more extensive standard WAIS/Z39.50 port, "database" is collections include: the source file to search, "?" delimits the database from the query, and "query" is the • Anthropology - http://www.lib.ncsu.edu/disciplines/anthropology.html your search strategy. For example: • Anthropology and Culture - gopher://riceinfo.rice.edu:70/11/Subject/Ant • Anthropology on the Internet - wais://vega.lib.ncsu.edu/alawon.src?nren http://www.umanitoba.ca:80/anthropology/aaa-revue.html • Institute of Social and Cultural Anthropology - Gopher servers and files can be specified http://www.rsl.ox.ac.uk/isca/ with URLs as well. Since gopher resource • World Wide Web Virtual Library: Anthropology - specifications require "Type" identifiers, http://www.usc.edu/dept/v-lib/anthropology.html and since paths to gopher resources often include spaces, gopher URLs usually Sometimes these pages may not contain all the sorts of information you need and deviate from the norm. For example, here therefore you may have to go searching for it. Try these resources as examples: is a URL describing the location of a subdirectory in Gopher at the NCSU • Lycos - http://lycos.cs.cmu.edu/lycos-form.html Libraries: • Searching the Web - http://www.yahoo.com/Reference/Searching_the_Web/ • WebCrawler - gopher://gopher.lib.ncsu.edu/11/library/ http://webcrawler.cs.washington.edu/WebCrawler/WebQuery.html • Yahoo - http://www.yahoo.com/ Notice the pair of 1's after the Internet name of the computer. These 1's specify Finally, there was recently published an article entitled "Internet resources for the resource as a directory. On the other anthropology" by Anita Cohen-Williams and Julia A. Henderson in College and hand, the following URL specifies a Research Libraries News volume 56, number 2, February 1995 pages 87-90, 113. specific text file within that directory: This article lists many Internet resources for anthropologists including more than a few discussion groups. page 4 of 4