Thoughts on the Intranet Project

Version 1.1, Saturday, January 25, 1997

I'm simultaneously wildly enthusiastic and concerned about the project. My enthusiasm comes from realizing that the project has the potential of accomplishing all at once many of the things those of us that use computers have hoped for and dreamed of. My concern is that its implementation will die the horrible death of so many good initiatives here at Cal Poly (as I often say, "They could heat the whole University on the wasted enthusiasm of the faculty", to which I would certainly included staff, and no small number of administrators).

In detail, then:

What does it do?

The Intranet is a Distributed Computing Environment (hereafter DCE; we have to have acronyms :-). The name is not exactly descriptive, except to people who work with computer networks: the computing environment is distributed over a number of machines, but to the user it seems rather monolithic. In fact, as far as the user is concerned, it might as well be running on a single machine.

The two parts of the DCE that are the most important to users are the authentication system and the Distributed File System (DFS). The authentication system handles user names and passwords. That by itself isn't very spectacular, but there are two features that make it exceptional. First, you will only have one user name and password for all your Cal Poly computing. If you have accounts on more than one machine, you know how hard it can be to keep track of them all. Second, you will have permission to use computer resources based on who you are, not on what machine you're using. Computer privileges will be allocated according to your membership in groups. Some of these groups will be automatic: all students, all faculty, all the students currently enrolled in BIO 190 sect. 5, all the faculty in the Biological Sciences Department.. Other groups can be custom-made: the faculty working on a collaborative project, the students in the Tri-Beta Honor Society.

One of the important consequences of groups is that resources can be easily targeted to the people for whom they are intended. Currently, if you put up a web page, all the world can see it. Sometimes you might want that, but other times you might prefer that only the students currently in a class have access. Indeed, in the case of information such as student records, that can be made more easily accessible through the Intranet than through Banner, there are legal restrictions that make this ability to restrict access crucial. The strength of the authentication system is that all this is handled without great effort on the part of the user.

From the user's standpoint, the Distributed File System gives you a new hard disk. Not a physical drive in your desktop machine, but rather a "virtual disk", consisting of actual storage on who knows how many machines (hence "distributed"). Its physical location is unimportant; the important thing is that on your Mac or Windows desktop it acts just like another (very large and somewhat slow) hard disk. It will contain most if not all the files of the entire university--all the files that are currently spread among the Academic and Administrative clusters, college and departmental machines, the campus webserver, and many individual desktops. Every member of the campus community will have access to the DFS, through the authentication system. So instead of logging onto several different computers for several different reasons, it will all be right there on your desktop. You will have access to every file that is permitted to you by the authentication system. You will control access to your personal files, and to files of groups that you create, so that you can give or restrict access to individuals or groups.

Many of the tools for creating and maintaining groups and managing files are accessible through the World Wide Web (and there will be many other campus services available through the web, when the authentication system is available campus-wide), but the Intranet Project isn't just about the Web, it's a system to handle all the computing resources on campus.

Advantages

It will be voluntary (why anyone wouldn't want to take advantage of it, though, is beyond me). If you choose to keep files on it,
  1. They will be available to you from any machine anywhere in the world that is connected to the Internet.
  2. If you have a computer at home, they will be available to you, on your desktop (through a PPP connection), just as they are available at Cal Poly.
  3. You can selectively make them available to any other users you would like to, or you can keep them entirely to yourself with a security mechanism that is stronger than that of most existing campus computers.
  4. You will only need to remember one username and one password for everything.
  5. All your files will be backed up automatically.
In addition, you can access site-licensed applications from the same system, either directly, running them off the DFS, or installing them on your own local hard disk from installation files on the DFS.

Here are some scenarios of things it will allow a faculty member to do:

  1. You are teaching a class. From a simple Web interface, you can
  2. By putting your Eudora files on the DFS, you can check mail from work or home without worrying where the messages end up.
  3. By putting gradebooks on the DFS, you can check and enter grades from work or home without worrying about inadvertently creating parallel copies of gradebooks.
  4. Even if you keep local copies on your hard disk, you can syncronize with the DFS using such things as the Windows 95 Briefcase.
  5. You're on sabbatical half a world away. If you can get hold of a machine with a net connection, the DFS is right there waiting for you.
  6. You've just created a dozen 4 megabyte graphics files at New Media Studio. No worry about putting them on floppies, just drag them to your DFS space. And you didn't have to put that Powerpoint presentation on floppies, either; just shoot the slides from the copy on DFS.
Here are some scenarios for students:
  1. You have an hour between classes. You can go to any computer on campus and work on your term paper, right where you left it on the DFS.
  2. You and your lab partner are collaborating on a lab report. She has permitted the file to you, so you can work on your part at home through your dial-up connection.
  3. You're applying for a job, so you order a transcript through a simple Web interface.
  4. You find out your grades, or even review your entire academic record, through the Web, without worrying about unauthorized people seeing it.

Things that seem like disadvantages, but aren't

If you forget your password, you're locked out until you can have a new one issued. But then there will be only a single password for the entire system.

Someone could hack the system and steal your files. Although there is a finite probability of this, it is less likely than the same thing happening on many current campus computers, and far less likely that someone breaking into your office and stealing your computer.

Faculty and students would be empowered (actual empowerment, not "Dilbert empowerment"). No, this doesn't sound like a disadvantage to me, either, but we live in strange times.

The system doesn't require 100% buy-in from the faculty, and the faculty who don't buy in aren't forced to participate. (Again, that seems like an advantage to me.)

It will cost a bundle (actually, the costs are modest as such things go; it just seems like a lot to those of us who are happy to get $100 for supplies to teach a course). But it will save far more than it costs, in reduced needs for mass storage, separate servers, and other duplicated hardware.

It cries for ethernet connections and decent computers on every faculty desktop. But then we need this anyway. Lev Gonick's leased-computer plan has the potential to integrate quite well (although it is certainly not necessary for the system to work).

It requires actual planning and vision for information technology on campus.

Things that are actual disadvantages

It does require 100% buy-in from the administration. IMHO it is the fundamental part of a computing infrastructure for our "University of the Twenty-first Century" (or whatever it is), but administrative waffling will take it down faster than pulling the plug on a T3 line into a router. :-) And it has continued vulnerability. In an ideal universe (or ideal University), such a system would evolve over time with new technology to meet new needs. In the Cal Poly universe, it could be replaced overnight like an old voice-mail system. If the system works the way it's supposed to, you'll back up your local files onto the DFS. If it works the way it could in my worst nightmare, you back up your DFS files onto your local disk.

What things could imperil the Intranet Project?

  1. Major University initiatives that are not directly compatible, such as changes to the campus database that use a separate authentication system. This is the easiest way to undermine it, since its whole premise is that it is the campus information infrastructure.
  2. Major investments for future projects that duplicate intranet services. If I were to go out and get a grant to set up a departmental server, with e-mail, web service, ftp, netnews, etc., that would be really counterproductive. Ironically I was considering just that a year ago, because it seemed at the time like the only solution.
And that's the basic problem. As long as the Intranet Project is an idea and a prototype, every solution we take now is at best duplicative and at worst wasteful of a future that may not come to pass. We can either do what we need to do, and take the chance that we are weakening the best chance we'll have for a long time to have an integrated functioning information infrastructure on campus, or else we can wait, and do nothing, and hope it comes to pass. (We could also squabble about the details while we're at it.)

But none of this is an issue if the Intranet Project is real. I say let's do it.


This page Copyright © 1996 by Curtis Clark. Last revision Saturday, January 25, 1997.

Space for this page is provided by California State Polytechnic University, Pomona. Although it is intended to further the educational mission of the University, the opinions expressed here are those of Curtis Clark, and do not represent official policy of the University.