CHI 96 Workshop: HCI and the Web, Position Papers

Keith Instone

Bowling Green State University
instone@cs.bgsu.edu

This is my position, even though I do not REALLY have to write up a paper in order to get in the workshop. This was written hastily and at the very last minute.

What is the World Wide Web?

It is easy to just talk about "the Web". That's fine for the 6 o'clock news or the newspapers. But we are professionals, and we should be using more accurate terms. Stolen directly from Roy Fielding's IRUS presentation on World Wide Web Software: An Insider's View, here are some different things that people mean when they say "the web" (italicized stuff added by me):

The Project: Collaboration on a global scale.
The Protocols: URI, HTML, HTTP.
The Applications: Clients and servers. Browsers, agents, proxies, gateways. Authoring tools.
The Information/Content: For every topic, there is at least one fanatic.

Let's try to use more accurate terms. So, for example, if we start talking about navigation problems, let's be sure to isolate what problems stem from the current state of HTML, which are caused by the lower level protocol (HTTP), what could be fixed by added functionality to servers, and what could be done to the browsers to help alleviate the problems.

SIGCHI as a mover and shaker

I think SIGCHI should incorpate Web services into its member benefits. This currently is access to the SIGCHI Bulletin online. This could also be member-only discussions, and maybe member-only access to other really tasty bits of online information. SIGCHI sites shold be held up to the utmost standards of design and implementation.

SIGCHI should also be maintainers of lots of PUBLIC information about HCI for "the rest of the world" to have access to. SIGCHI should organize a specially-made web on "what is HCI and how you can use it on your web site". Include: SIGCHI's guide to good design, pointers to where to learn more (CHI tutorials, duh), links to active fields of research that could be applied to the Web, issues about users that should not be ignored, etc. I already maintain the HCI Virtual Library and am willing to give SIGCHI control over the name and my content, but SIGCHI would have to take up managing it (I want to retire!). Hans De Graaff (who does HCI Index) and Mikael Ericsson (who does HCI Resources) are both interested in working more formally with SIGCHI to provide an improved HCI Virtual Library.

SIGCHI (through ACM) should be members of the World Wide Consortium (W3C). The idea of W3C is to be a central "standards" body and to bring all of the Web vendors together (like Netscape and Microsoft) so they can agree on HTML entensions and the like. SIGCHI (and other ACM agencies) should be there to steer the vendors toward considering USER needs and other agendas that they are apt to miss in the rush for the almighty dollar. SIGCHI should be there to show them that USABILITY is one way to get that dollar. Steven Pemberton's CWI is a member and Steven has already had an impact on W3C's efforts for style sheets.

What about a UIST-like symposium on Web usability? The original idea to use UIST as a model for an ACM/Web event belongs to Rick Furuta, who thought it would be a good way to bridge the gap between SIGLINK and the Web community. I think this could work for SIGCHI as well, with perhaps even the two SIGs working together on the symposium.

HCI individuals as movers and shakers

An individual can make a difference as well! Use the Internet to make contacts and suggestions. Years ago I recall sending email to the Mosaic developers, saying that I thought it would be useful to have ALT text displayed for unloaded graphics. Later, they did it and I immediately thanked them. I do not know if I was the first to sugest it, or if I was one of many who complained, or if they thought it up themselves first, but they implemented, which all that I really care out. lately, I've struck a dialog with Hagan Heller, a GUI designer at Netscape. Hagan will be at CHI (!), presenting a wonderful tutorial at the same time as this workshop. I hope to "grab her ear" about some of the usability problems with their browser. (And I have her tutorial notes, which I will let your browse at the workshop.) Also, Darrell Sano has worked with Jakob Nielsen and is now at Netscape, so there are important people at these developers that HCI people can talk to.

Although we are mainly "HCI folks" here, some very influential businesses are represented here. I hope these representatives can act as liaisons between the HCI community and the company Web groups so that HCI people can get their voices heard easier. Robin Jeffries's position paper is very encouraging in that way because I see an HCI person getting involved with the Java developers. The fact that Java demos all have that silly swirling coffee, despite Jakob Nielsen's remarks, is a bad sign, however.

So, as an "enlightened" and informed person, you can accept some responsibility for these bad designs and user interfaces on the Web. We you see a problem, point it out and offer constructive criticisms. Find the important people and let them know what you think. Write about usability Web issues. Jakob Nielsen's Alertbox has some raised some much-needed awareness. Other, "non-HCI" folks are writing about it now too. We need to better document our own develpoments so others can learn. I need to write up a detailed description of how I developed the CHI 96 web site (the organization of the user interface, how it is usable by blind people, etc), so I am not an angel either. We need more things like Sun's home page analysis to stand up as examples of "how to do it right".

So, what about getting these academic activities to have an impact? I organized another workshop, at Hypertext '96, about hypermedia research and the Web. We had two "Web experts" (Dan Connolly/HTML and Roy Fielding/HTTP: be sure to read their position papers for that workshop) who helped illuminate us academians and what is bad about the web today and what is bad about its inherenet design. It turns out the Web could do much more today (typed links, for example), but no one has really implemented it all yet.

The outshot of that workshop will be (hopefully!) a report on "opportunities for improving the Web": a list of things that can be improved. There will be lots of usability suggestions as well os some lower-level ones. There will be lots of references to the literature and many implementation details. We will work hard to keep it from being just a bunch of "bitchings about the state of the Web". The audience will be W3C members, so we hope to get these people interested so that they will read it and then there will be a chance that it could really make a difference. I had hoped to have a rough draft done in time for the CHI workshop, but I doubt that will happen.

After the CHI workshop, there is a BCS HCI symposium on a related topic. We are trying to keep these 3 workshops "in tune" with each other, so any results from this workshop will be passed on to the organizers of that one. We even tried to get a panel on usability at the next Web conference, but they rejected us.

Conclusion

Now is the time to get involved. Hopefully at this workshop we can make some progress on getting HCI people more involved in the development of the Web.

CHI 96 Workshop: HCI and the Web, Position Papers