Planning

part of Software Engineering for Internet Applications by Eve Andersson, Philip Greenspun, and Andrew Grumet
If you're reading this chapter, we assume that you've completed the "Basics" problem set and are going to stay with the course for the rest of the semester. Welcome. Now it is time to plan your work during the core of the course.

Everyone in this course will be building an online learning community, a site where users teach each other. The work may be done alone or in groups of two or three students. Ideally, you or your instructors will find a real client for you, someone who wants to publish and administer the community on an ongoing basis. A good client would be a non-profit organization that wants to educate people about the subject surrounding its mission. A good client would be a medium-sized company that wants a knowledge-sharing system for employees. A good client would be a student group at your university. If you can't find a client, pick something that you're passionate about. It could be Islamic architecture. It could be African Cichlids (a family of freshwater fishes, living mostly in the rift lakes of East Africa; see www.cichlid.org). It could be cryptography. Pick something where you think that you can easily get or generate magnet content, some tutorial information that will attract users to your service.

You are building the same type of project as everyone else in the class. Thus it will be easy for you to compare approaches to, for example, user registration or content management.

Before you starting writing code, however, we'd like you to do some planning and competitive analysis. Fundamentally you need to answer the questions "Who is going to teach what to whom?" and "What alternatives are currently available for this kind of learning?"

User Classes

Start by dividing your users into classes. Two users should fall into the same class if you expect them to want substantially the same experience with your service. It is almost always useful to think about different levels of administrative privileges as you are dividing the users into classes. It is almost never useful to think about teachers versus learners; the whole point of an online community is that each user learning some of the time and each user is teaching some of the time.

Example user class decomposition on photo.net

To give you an idea of what a user class decomposition might look like, we'll walk through one for the photo.net service.

First, consider the overall objective of photo.net: A place where a person can go and get the answer to any question about photography.

Second, consider levels of administrative privilege. There are site-wide administrators, who are free to edit or delete any content on the site. These administrators also have the power to adjust the authority of other users. We have moderators who have authority to approve or delete postings in particular discussion forums. Finally there are regular users who can read, post, and edit their own contributions. A less popular service could probably get away with only two levels of admin privilege.

A different way of dividing the users is by purpose in visiting the service:

A final way of dividing users that may be useful is by how they connect. In the case of photo.net, it is easy to envision the Web browser user. This user is uploading and downloading photos, participating in discussions, reading tutorials, shopping for equipment, etc. The same person may connect via a mobile phone, in which case he or she becomes a mobile user. If the mobile user is in the middle of a photographic project, we want to provide information about nearby camera shops, processing labs, repair shops, time of sunset, good locations, and other useful data. If the mobile user is connecting for social purposes, we need to think about what are practical ways for a person on a mobile phone to participate in an online community. Our engineering challenge is similar for the telephone user.

Usage Scenarios

For each class of user, you should write down a rough idea of what a person in this class would get from your new service. You may want to hint at page flow.

Example: novice photographer shopper at photo.net

The novice should start by reading a bunch of carefully authored camera-buying advice articles and then reviews of specific cameras. Much of the best shopping advice is contained in question-and-answer exchanges within the discussion forums so editors will need a way to pick out and point to the best threads in the forum archives. After our user has read all of this stuff, it would be ideal if he or she could be directed into a Q&A forum where "here's what I've decided to buy; what do you think?" questions are welcomed. That could be implemented as an explicitly social shopping system with one column for responses from other readers and an adjacent column for bids from camera shops.

Example: site-wide administrator at photo.net

The site-wide administrator should log in and see a page that gives the pulse of the community with statistics on the number of new users registered, the quantity of photos uploaded into the photo sharing system, the activity in the discussion forums, the relative efforts of the moderators (volunteers from the community). If there are unbanned users who have been responsible for an onerous amount of moderator work in deleting off-topic postings, etc., these should be listed with a summary of their problematic activities and an option to ban.

Exercise 1a

Answer the following questions: Make sure that your answers to this and all subsequent exercises are Web-accessible. It is a good idea to get into the discipline of ensuring that all documents relevant to your project are available on the project server itself, perhaps in the /doc/ directory.

Exercise 1b: Build User Profile Pages

When building an application it would be ideal to have the potential users in the room with you at all times. Their needs would then be concrete and manifest. This isn't very practical, however, and therefore as an aid to making the people for whom you're building the application concrete, you should build two or three profile pages. A profile page contains the following information: (a) a picture of the user, (b) the user's name, age, occupation, marital status, housing situation, and income, (c) the user's short-term and long-term goals relevant to the online community that you're building, (d) the immediate questions that this user will bring to the site, (e) the kind of computer equipment and connection in this person's house, and (f) any other information that will help to humanize this fictitious person.

To assist you in this task we've created a couple of examples for an online learning community in the area of general aviation:

If you don't have a good photo library of your own, lift photos (with credit) from photo.net and other online sources.

Don't spend more than one hour on this exercise; plenty of truly awful software has been written with fancy user profiles on the programmers' desks. There is no substitute for launching a service to real users and then watching their behavior attentively.

Exercise 1c

For each class of user identified in Exercise 1a, produce a textual or graphical usage scenario for how that user will experience your service.

Evaluating Alternatives: Offline

Nick Gittes and Alex. 1998. Whenever Nick Gittes (1902–2000; grandfather of one of the authors) would see a computer advertised on television, he'd say "A home computer. Why would anyone want a thing like that?"

The dotcom boom is over. You ought to have a good reason for building an information system. If a curmudgeon wants to know why you need all these fancy computers instead of a book, some chalk, and pencil and paper, it would be nice to have a convincing answer.

There are good reasons to look at the best elements of offline resources and systems. After several millenia, many of these systems are exquisitely refined and extremely effective. Your online community and technology-aided learning environment can be much improved by careful study of the best offline alternatives.

Example: Popular Photography magazine

The largest circulation offline publication in the U.S. world of photography is the sixty-five-year-old magazine Popular Photography. It is extremely effective at answering the following questions: What is the price of a Nikon 50/1.4 lens? What are the latest cameras available? How does the new Canon Elan 7 body perform on a test bench?

The magazine is ineffective for start-to-finish learning. It is impossible for Popular Photography to sell enough ads to run a twenty-page tutorial article much less a complete beginner's textbook on photography. The magazine is ineffective for finding the average price for a used or obscure item. (Shutterbug is the place for used and classic camera dealer ads.)

The magazine is ineffective as a means of getting answers to arbitrary questions. There is a "Too Hot to Handle" section in every issue that promises "Honest, forthright answers to your most probing questions." Unfortunately, only four questions are answered per issue. Presumably these were submitted at least a couple of months previously. And the answers come only from the editors of the magazine; they don't tap the reserves of knowledge that may reside among the hundreds of thousands of subscribers.

The magazine is ineffective as a means of exhibiting reader work. The "Your Best Shot" column displays five or six reader-contributed photos in every issue, but being selected is a slow and uncertain process.

Example: face-to-face course

The main strength of a face-to-face course in photography is guaranteed mentoring. The instructor keeps track of each student's progress and questions. Another strength of a face-to-face course is the availability of critiques from other students and the instructor.

Face-to-face courses require more travel, time, effort, and commitment than many people are able to give. Once the course is over, the student is unable to avail him or herself of mentoring and critiques from the instructor and other students.

Exercise 2

Write down the best features of offline alternatives for learning the subject matter of the service that you're building. Indicate those features that you think can be translated into an online community and, if so, how. Write a three-sentence justification for why your online learning community will be an improvement over offline alternatives for at least some group of people.

Evaluating Alternatives: Online

In performing your analysis of online competitors you should look at any service or resources for learning in your subject area. But pay special attention to online learning communities. Before setting off to surf, you might find it useful to think about the following elements of sustainable online community:
  1. magnet content authored by experts
  2. means of collaboration
  3. powerful facilities for browsing and searching both magnet content and contributed content
  4. means of delegation of moderation
  5. means of identifying members who are imposing an undue burden on the community and ways of changing their behavior and/or excluding them from the community without them realizing it
  6. means of software extension by community members themselves
When you visit a site ask yourself "Did they author or license a substantial body of tutorial content?" Look at the ratio between publisher-authored and user-authored content. If more than half of the content is authored by users, the site is heading away from publishing and toward community. Note the number of different ways in which a user can post some information that then becomes available to other users ("means of collaboration"). See whether the default search utility on the site returns results from things like discussion forum postings. As an unprivileged user, it will be hard for you to determine whether or not the site has provisions for distributing the content moderation burden or excluding difficult users. However, you can sometimes make inferences about Element 6, whether or not the software can be extended by people who are regular users and administrators of the community but aren't expert programmers. If the site is commercial, check the job listings to see what skills are required. If the site hasn't adopted the abstract URL religion (see "Basics" chapter) the file extensions may give you a clue as to implementation technology. A ".pl", ".asp", ".php", ".adp", or ".tcl" indicates the use of a scripting language amenable to program extensions by novices. A ".jsp" or the term "servlet" in the URL indicates the use of Java, a language that is intended only for professional programmers.

Exercise 3

Find the best existing online communities in your subject area. Note how closely they conform to the six elements of sustainability listed above. Also write down anything strikingly good or bad about the registration process and the mechanisms of collaboration, e.g., in discussion forums, comments on articles, and chat rooms. Look for voice and mobile interfaces. If present, try them out. (The "Adding Mobile Users To Your Community" chapter provides a list of desktop browser-based phone emulators so that you won't have to use your mobile phone; alternatively type "WAP emulator" or "Mobile browser emulator" into a public search engine.) Look for evidence of personalization and direct controls over preferences.

Does the World Need More than One Online Community?

Suppose that in Exercise 3 you found 100 online communities in your subject area. Should you be discouraged? A search at amazon.com for "aquarium" yields 679 books on the subject of keeping fish at home. The National Center for Educational Statistics (nces.ed.gov) found that there were 3913 colleges and universities in the United States in 1998. The computing world historically has tended toward concentration. There is only room for one or two desktop user interfaces, one or two word processors, one or two spreadsheets, a handful of enterprise accounting systems, etc. Two factors prevent similar concentration in the world of online learning communities. One is the idiosyncracy of authorship. A person may believe that Anna Karenina is the best novel ever written and simultaneously be interested in reading C. D. Payne's Youth in Revolt. A moviegoer may believe that Gone with the Wind will never be excelled and yet still invest two hours in Charade. The range of idiosyncracy is narrower in explicitly tutorial works. Nonetheless one can admit the towering accomplishment of Stroebel's Basic Photographic Materials and Processes and still profit from working through Mastering Black-And-White Photography (Bernhard Seuss 1995).

The second force that prevents concentration in the world of online learning communities is the nature of community itself. Christopher Alexander, Sara Ishikawa, and Murray Silverstein argue in A Pattern Language (Oxford University Press 1977) against countries of more than a few million people:

"It is not hard to see why the government of a region becomes less and less manageable with size. In a population of N persons, there are of the order of N^2 person-to-person links needed to keep channels of communication open. Naturally, when N goes beyond a certain limit, the channels of communication needed for democracy and justice and information are simply too clogged, and too complex; bureaucracy overwhelms human process. ...

"We believe the limits are reached when the population of a region reaches some 2 to 10 million. Beyond this size, people become remote from the large-scale processes of government. Our estimate may seem extraordinary in the light of modern history: the nation-states have grown mightily and their governments hold power over tens of millions, sometimes hundreds of millions, of people. But these huge powers cannot claim to have a natural size. They cannot claim to have struck the balance between the needs of towns and communities, and the needs of the world community as a whole. Indeed, their tendency has been to override local needs and repress local culture, and at the same time aggrandize themselves to the point where they are out of reach, their power barely conceivable to the average citizen."

If it were possible for everyone to pile into a single community and have a great learning experience, America Online would long ago have subsumed all the smaller communities on the Internet. One of the later chapters of this book is devoted to the topic of growing an online community gracefully to a large size. But, for now, rest assured that it is a hard problem that nobody has solved. Given sufficiently high quality magnet content and an initial group of people dedicated to teaching, there will always be room for a new learning community.

Exercise 4

Identify sources of magnet content for your community this semester. If some of this content is going to come from other people, write to them and ask for permission. Even if you're only using their work experimentally, one concern that an author or publisher might have is that your site will get indexed by search engines and readers will be misdirected to your site instead of theirs. In practice, this is not a problem if your server isn't accessible from the public Internet or if you include a robots.txt file that will instruct search engines to exclude certain content. You may get a friendlier response from copyright holders if you agree to provide a hyperlinked credit and to ensure that their content does not become multiply indexed.

If you have a client who is supplying all the magnet content, write down a summary of what is going to be available and when. Next to each class of documents note the person responsible for assembling and delivering them. As an engineer, it isn't your job to assemble and develop content, but it is your job to identify risks to a project, such as "not enough magnet content" or "nobody has thought about magnet content".

Domain Name System

The Domain Name System (DNS) translates human-readable hostnames, e.g., www.google.com, into machine-readable and network-routable IP addresses, e.g., 216.239.57.100. DNS is a distributed application in that there is no single computer that holds translations for all possible hostnames. A domain registrar, e.g., register.com, records that the domain servers for the google.com domain are at particular IP addresses. A user's local name server will query the name servers for google.com to find the translation for the hostname www.google.com. Note that there is nothing magic about "www"; it is merely a conventional name for a computer that runs a Web server. The procedure for translating a hostname such as froogle.google.com is the same as that applied for translating www..

Exercise 5: Settle on a Hostname

Working with your client, pick a hostname for the application that you'll be building this semester. If you're building something within MIT, for example, you'll probably want to pick something.mit.edu. You and your client will need to navigate the IT bureaucracy to register that hostname and map it to the IP address of your server. If you're building a service for a client who does not have an Internet domain, encourage them to come up with a good name and register it. The authors have had good experience with register.com, a service that includes bundled use of their DNS servers; the domain owner can edit the hostname-to-IP-address translations with a Web browser interface.

Exercise 6: Negotiate Intellectual Property Rights

One of the thing that distinguishes a professional software engineer is success in negotiating intellectual property rights. If you give away all rights to everything that you produce as a "work for hire", you won't have a personal toolkit of software that you can reuse on new projects. If you don't give away any rights, nobody will be able to run your software, which probably means that you won't be able to solve social or organizational problems. A good negotiator gives away things that are valuable to the other side, but that aren't valuable to his or her side.

During this course, for example, you will ideally want to retain ownership of all software that you produce. You will therefore be free to reuse the code in any way, shape, or form. The client, however, is going to be putting in a lot of time and effort working with you over a period of months and is thus entitled to some benefit. Your university tuition payments have probably drained away all of the cash in your bank account and therefore you won't be giving the client money as compensation for his or her time. What you can do is give the client a license to use your software. This obviously benefits the client but it also benefits you. The more people that are out there happily running your software the better your professional resume looks.

Should you try to limit what the client can do with your software? Generally this isn't worthwhile. Any organization that comes to you for programming assistance is probably not an organization that will want to hang out a shingle and offer to develop software for others. If they do decide that it would make sense to adapt your software to another application within the company, it is very likely that they will call you first to offer a consulting fee in exchange for your assistance.

How about limiting your liability? Oftentimes software engineers are called upon to write programs whose failure would have catastrophic results. Suppose that you are offered $100,000 to write a trading program for an investment bank. That may seem like a great deal until the bank sues you for $100 million, alleging that a bug in your program cost them $100 million in lost profits. In the biomedical field a bug can be much more serious. There is the famous case of the Therac-25 radiation treatment machine, bugs in whose control software cost lives (see http://sunnyday.mit.edu/therac-25.html).

Disclaiming liability is difficult, even for trained lawyers, and hence this is best left to professionals. Nearly every commercial software license includes a disclaimer of warranty. Here's a snippet from the Microsoft End User License Agreement (EULA):

19. DISCLAIMER OF WARRANTIES. TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW, MICROSOFT AND ITS SUPPLIERS PROVIDE THE SOFTWARE AND SUPPORT SERVICES (IF ANY) AS IS AND WITH ALL FAULTS, AND HEREBY DISCLAIM ALL OTHER WARRANTIES AND CONDITIONS, WHETHER EXPRESS, IMPLIED, OR STATUTORY, INCLUDING, BUT NOT LIMITED TO, ANY (IF ANY) IMPLIED WARRANTIES, DUTIES OR CONDITIONS OF MERCHANTABILITY, OF FITNESS FOR A PARTICULAR PURPOSE, OF RELIABILITY OR AVAILABILITY, OF ACCURACY OR COMPLETENESS OF RESPONSES, OF RESULTS, OF WORKMANLIKE EFFORT, OF LACK OF VIRUSES, AND OF LACK OF NEGLIGENCE, ALL WITH REGARD TO THE SOFTWARE, AND THE PROVISION OF OR FAILURE TO PROVIDE SUPPORT OR OTHER SERVICES, INFORMATION, SOFTWARE, AND RELATED CONTENT THROUGH THE SOFTWARE OR OTHERWISE ARISING OUT OF THE USE OF THE SOFTWARE. ALSO, THERE IS NO WARRANTY OR CONDITION OF TITLE, QUIET ENJOYMENT, QUIET POSSESSION, CORRESPONDENCE TO DESCRIPTION, OR NON-INFRINGEMENT WITH REGARD TO THE SOFTWARE.

20. EXCLUSION OF INCIDENTAL, CONSEQUENTIAL, AND CERTAIN OTHER DAMAGES. TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW, IN NO EVENT SHALL MICROSOFT OR ITS SUPPLIERS BE LIABLE FOR ANY SPECIAL, INCIDENTAL, PUNITIVE, INDIRECT, OR CONSEQUENTIAL DAMAGES WHATSOEVER (INCLUDING, BUT NOT LIMITED TO, DAMAGES FOR LOSS OF PROFITS OR CONFIDENTIAL OR OTHER INFORMATION, FOR BUSINESS INTERRUPTION, FOR PERSONAL INJURY, FOR LOSS OF PRIVACY, FOR FAILURE TO MEET ANY DUTY INCLUDING OF GOOD FAITH OR OF REASONABLE CARE, FOR NEGLIGENCE, AND FOR ANY OTHER PECUNIARY OR OTHER LOSS WHATSOEVER) ARISING OUT OF OR IN ANY WAY RELATED TO THE USE OF OR INABILITY TO USE THE SOFTWARE, THE PROVISION OF OR FAILURE TO PROVIDE SUPPORT OR OTHER SERVICES, INFORMATION, SOFTWARE, AND RELATED CONTENT THROUGH THE SOFTWARE OR OTHERWISE ARISING OUT OF THE USE OF THE SOFTWARE, OR OTHERWISE UNDER OR IN CONNECTION WITH ANY PROVISION OF THIS EULA, EVEN IN THE EVENT OF THE FAULT, TORT (INCLUDING NEGLIGENCE), MISREPRESENTATION, STRICT LIABILITY, BREACH OF CONTRACT, OR BREACH OF WARRANTY OF MICROSOFT OR ANY SUPPLIER, AND EVEN IF MICROSOFT OR ANY SUPPLIER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.

21. LIMITATION OF LIABILITY AND REMEDIES. NOTWITHSTANDING ANY DAMAGES THAT YOU MIGHT INCUR FOR ANY REASON WHATSOEVER (INCLUDING, WITHOUT LIMITATION, ALL DAMAGES REFERENCED HEREIN AND ALL DIRECT OR GENERAL DAMAGES IN CONTRACT OR ANYTHING ELSE), THE ENTIRE LIABILITY OF MICROSOFT AND ANY OF ITS SUPPLIERS UNDER ANY PROVISION OF THIS EULA AND YOUR EXCLUSIVE REMEDY HEREUNDER (EXCEPT FOR ANY REMEDY OF REPAIR OR REPLACEMENT ELECTED BY MICROSOFT WITH RESPECT TO ANY BREACH OF THE LIMITED WARRANTY) SHALL BE LIMITED TO THE GREATER OF THE ACTUAL DAMAGES YOU INCUR IN REASONABLE RELIANCE ON THE SOFTWARE UP TO THE AMOUNT ACTUALLY PAID BY YOU FOR THE SOFTWARE OR U.S.$5.00. THE FOREGOING LIMITATIONS, EXCLUSIONS, AND DISCLAIMERS SHALL APPLY TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW, EVEN IF ANY REMEDY FAILS ITS ESSENTIAL PURPOSE.
This is so important to Microsoft that it is the only part of a twelve-page agreement that is printed in boldface and the only part that is presented in a French translation for Canadian customers as well.

If you don't want to cut and paste Microsoft's verbiage, which might expose you to a copyright infringement action from Redmond, consider employing a standard free software or open-source license, of which the GNU General Public License is the best-known example. Note that using a free software license doesn't mean that your software is now free to the world. You may have licensed one client under the GNU GPL but whether or not you decide to offer anyone else a license is a decision for the future.

If you wish, you can use the sample contract at the end of this book as a starting point in negotiating rights with your client. And remember that old bromide of business: You don't get what you deserve; you get what you negotiate.

More

To the Instructor

It is helpful during the second meeting of the class to bring clients on campus to give three-minute presentations pitching their projects. Here is a suggested outline to the client for the presentation:
  1. introduce the speaker and the organization he or she represents (15 seconds)
  2. explain who the users are and why they need to interact via an Internet application, i.e., what problem is this online community solving (1.5 minutes)
  3. describe how users will be attracted to the site initially, e.g., is there a collection of magnet content that these people need that isn't available anywhere else? (30 seconds)
  4. after the site has been up and running for a few months, what will a typical interaction look like for a new user? (30 seconds)
  5. what will happen after the semester is over, how will the system be funded and sustained? (15 seconds)
The client should be prepared to answer questions for a minute or two after the presentation.

Time and Motion

This problem set involves a two-hour meeting with the client, perhaps two hours of discussion and Web surfing among the students, an hour to build the user-profile pages, two hours of writing. That's seven hours total, most of which requires the entire team to work together.
Return to Table of Contents

[email protected], [email protected], [email protected]